OUP user menu


DOI: http://dx.doi.org/10.1093/alcalc/agh159 269-273 First published online: 3 May 2005


Aims: To investigate the relationship of different patterns of alcohol intake to various types of trauma. Methods: We examined the associations of alcohol consumption in a series of 385 consecutive trauma admissions (278 men, 107 women, age range 16–49 years). Patients underwent clinical examinations, structured interviews on the amount and pattern of alcohol intake, and measurements of blood alcohol concentration (BAC). Results: On admission, 51% of the patients had alcohol in their blood. Binge drinking was the predominant (78%) drinking pattern of alcohol intake. Assaults, falls and biking accidents were the most frequent causes of trauma. Dependent alcohol drinking and binge drinking were found to be significantly more common among patients with head trauma than in those with other types of trauma (77% vs 59%, OR = 2.38; 95% CI 1.50 to 3.77). The OR for sustaining head injury increased sharply with increasing BAC: 1–99 mg/dl (1.24; 95% CI 0.55–2.01), 100–149 mg/dl 1.64; 95% CI 0.71–3.77), 150–199 mg/dl (3.20; 95% CI 1.57–6.53) and >199 mg/dl (9.23; 95% CI 4.79–17.79). Conclusions: Binge drinking is a major risk factor for head trauma among trauma patients. Assaults, falls and biking accidents are the commonest causes for such injuries. The relative risk for head injury markedly increases with increasing blood alcohol levels. Alcohol control measures should feature in policies aiming at the prevention of trauma-related morbidity and mortality.

(Received 30 September 2004; first review notified 28 October 2004; in revised form 3 January 2005; accepted 6 January 2005)


Excessive alcohol consumption is a major cause of accidents throughout the world. Although population studies have found no direct relationship between per capita alcohol consumption and injury mortality per se, a high frequency of alcohol misuse has been noted previously in several studies on patients with trauma (Rivara et al., 1993, Rivara et al., 1997; Porter, 2000). In studies which have specifically addressed head injuries, alcohol misuse has also been found to be a common cause of morbidity and mortality (Gurney et al., 1992; Rönty et al., 1993; Corrigan, 1995).

Binge drinking is currently a growing problem in most Western countries, and recent surveys in general hospitals have indicated that it may actually be a more prevalent cause of medical hazards than chronic drinking (Chikritzhs et al., 2001; Mäkelä et al., 2001; Naimi et al., 2003). Binge drinking has also been linked with an increased prevalence of trauma in several studies (Honkanen et al., 1976; Cherpitel et al., 1995; Rivara et al., 1997; McLeod et al., 1999; Li et al., 2001; Borges et al., 2004). McLeod et al. (1999) reported that the consumption of ≥5 drinks on one occasion produces an odds ratio (OR) of 3.4 (95% confidence interval (CI) 1.8–6.4) for sustaining an injury. Despite the large body of evidence of alcohol abuse as a risk factor for trauma, little attention has been paid to the comparisons of the types and mechanisms of injury that are associated with different patterns of alcohol drinking and to actual blood alcohol levels at the time of injury.

The aim of this study was to examine the types and mechanisms of injury in consecutive trauma patients who underwent detailed interviews on the amount and pattern of drinking preceding the injury.



The study was conducted at Oulu University Hospital, which treats all trauma patients with or without referral from primary health care in a city of 120 000 inhabitants. A total of 385 consecutive trauma admissions between June 1998 and July 2000 among patients in the age range of 16–49 years and showing evidence of physical trauma, requiring assessment and treatment by the emergency room physician, were included. The study protocol was approved by the ethical committee of the hospital, and informed consent was obtained from all the patients or their relatives before the study.

After immediate critical care, all clinical data, including the cause, type and extent of injury were carefully recorded by trained emergency department physicians on a structured checklist designed for this purpose. The causes of injury were classified as follows: motor vehicle crash, fall on the ground, fall from height, assault, sport injury and unspecified injury (Table 1). Motor vehicle crashes (40) included both motor vehicle occupants (32) and motorcycle riders (8). The class of unspecified (113) injuries included both blunt (62) and penetrating/cutting (51) injuries. In each case, the injured body parts were divided into six categories as follows: head, spine, thorax, abdomen and upper and lower extremities. The presence of injury was recorded when there was a distinct physical evidence of trauma as assessed by the emergency room physician. Additionally, the injury severity score (ISS) was used as an index of trauma severity (Baker et al., 1974; Copes et al., 1988). In the case of patients with head injury, the Glasgow coma scale (GCS) score was also determined, and the patients were graded as having mild (GCS 13–15), moderate (GCS 9–12) or severe (GCS 3–8) brain injury. The emergency room physician decided the additional examinations (imaging) that were needed, and patients were subsequently treated according to the hospital routine protocols. All patients with head trauma were followed up for possible signs of neurological deterioration. The patients included in this study did not report cancer, multiple sclerosis, stroke, treated epilepsy or any other neurological diseases of known aetiology.

View this table:
Table 1.

Main clinical characteristics

No. of patients278 (72%)107 (28%)
Age, years (mean ± SD)31 ± 1031 ± 11
Injury severity score (ISS), mean ± SD (range)4 ± 7 (1–75)3 ± 3 (1–25)
Trauma mechanism, n (%)
    Fall on the ground47 (17)31 (29)
    Fall from height20 (7)1 (1)
    Assault53 (19)11 (10)
    Biking accident28 (10)18 (17)
    Motor vehicle crash31 (11)9 (8)
    Sport injury17 (6)6 (6)
    Not specified82 (30)31 (29)
Surgical operation, n (%)38 (14)9 (8)
Hospital stay >2 days, n (%)63 (23)17 (16)
Dead, n (%)3 (1)0

Alcohol data

Blood alcohol concentration (BAC) was determined either from exhaled air (BrAC) (n = 184) or from serum samples (BLAC) (n = 165) as described previously (Savola et al., 2004). The amount and pattern of alcohol consumption was recorded by one of the authors (O.S.) blinded to the data on blood alcohol levels. Interviews on alcohol consumption were carried out using a structured interview protocol during a follow-up visit, which took place within 6 weeks after the injury. The history of alcohol consumption included the following information: how many drinks of alcohol (standard drink = 12 g of ethyl alcohol) the patient had consumed during the (i): 24 h and (ii): 1 week preceding the injury. The interviews on alcohol consumption were based on the time-line follow back method (Sobell and Sobell., 1995). Based on the data, the patients were classified into groups as follows: dependent drinkers (‘alcoholics’), binge drinkers, light-to-moderate drinkers and non-drinkers. Dependent drinkers were persons who showed clinical evidence of pathological alcohol use, social impairment and tolerance/withdrawal. The daily alcohol consumption by these individuals had exceeded a mean of 80 g. Binge drinking was defined as an ethanol intake of ≥6 (men) or ≥4 (women) standard drinks of alcohol in one session. Binge drinkers were further divided into two groups: frequent binge drinkers reported binge type drinking more than once a month. Infrequent binge drinkers reported binging 1–11 times/year. Light-to-moderate drinkers did not drink for intoxication, but consumed 1–2 standard drinks/day either daily or less frequently. Non-drinkers had not drunk any alcohol during the year preceding the injury. They included both life-long abstainers and ex-drinkers. Dependent drinkers and frequent binge drinkers together made up the group referred to here as hazardous alcohol drinkers.

Laboratory procedures

Venous blood samples were obtained upon admission, and blood alcohol concentrations were determined on a Vitros 250 clinical chemistry analyzer (Johnson & Johnson Clinical Diagnostics, Rochester, NY). ALCO-SENSOR III (Intoximeters, Inc, St Louis, MI) was used for the breath analyses.

Statistical methods

Continuous variables were compared between the groups by using Student's t-test with SPSS, version 10.0, for Windows. OR and 95% CI and differences between proportions and 95% CIs were calculated by using the CIA statistical software version 1.0. OR estimates and 95% CIs after adjustments for age and sex were calculated by logistic regression with SPSS.


The distributions of BAC in the trauma patients on admission are shown in Table 2. Fifty-one per cent of the patients had alcohol in their blood on admission, and most (86%) of these had reached the level of 100 mg/dl (22 mmol/l). Women were more often without any alcohol in blood, whereas men were more frequently severely intoxicated (BAC ≥200 mg/dl).

View this table:
Table 2.

Distributions of BACs in trauma patients on admission as divided by gender

AllMenWomenDifferences between proportions (95% CI)*
BACn = 385n = 278n = 107
0 mg/dl, %187 (49)124 (45)63 (59)14 (3–25)
1–49 mg/dl12 (3)10 (4)2 (2)2 (−3–5)
50–99 mg/dl15 (4)11 (4)4 (4)0 (−6–4)
100–199 mg/dl70 (18)51 (18)19 (17)1 (−9–8)
≥200 mg/dl73 (19)82 (30)19 (18)12 (2–20)
  • 100 mg/dl = 22 mmol/l.

  • * Difference between men and women (percentage).

When the patients were classified according to the types of trauma, the incidence of BAC-positive patients was significantly higher among patients with head trauma than among those with other types of trauma (65% vs 32%, OR = 3.92; 95% CI 2.55–6.03). The OR for sustaining head injury increased sharply with increasing BAC: at and above the level of 150 mg/dl, the risk for head trauma was significantly higher than that for other types of trauma (Fig. 1). The ORs (95% CI) were 0.51 (0.42–0.63), 0.67 (0.32–1.38), 0.84 (0.40–1.76), 1.71 (0.93–3.17) and 4.87 (2.82–8.40) for 0 mg/dl, 1–99 mg/dl, 100–149 mg/dl, 150–199 mg/dl and >199 mg/dl, respectively. After adjustment for age and sex, the corresponding ORs were 0 mg/dl (reference), 1–99 mg/dl (1.24; 95% CI 0.55–2.01), 100–149 mg/dl (1.64; 95% CI 0.71–3.77), 150–199 mg/dl (3.20; 95% CI 1.57–6.53) and >199 mg/dl (9.23; 95% CI 4.79–17.79).

Fig. 1.

ORs for head injury at various levels of BAC. The risk for head injury was greater than the risk for other types of injury at >150 mg/dl of BAC. The bars indicate the 95% CIs at different BACs, and if the bar does not cross the level one, it is statistically significant.

Table 3 summarizes the association between BAC and the cause and type of injury. Ninety-four per cent of the assault victims were BAC-positive on admission. Most of these patients (69%) showed blood alcohol levels at or > 150 mg/dl, and most (92%) of their traumas were head injuries. The patients injured by falls on the ground and biking accidents also frequently had alcohol in their blood, 60 and 61%, respectively. The BAC-positive patients in these groups also typically sustained head traumas. Seventy-seven per cent of those with head trauma owing to a fall on the ground were BAC-positive, whereas only 48% of those who were BAC-negative had head trauma (OR = 3.50; 95% CI 1.32–9.26). The likelihood of sustaining head trauma owing to falls on the ground increased with increasing BAC. Eighty-nine per cent of those with BAC at or >150 mg/dl had head trauma. Ninety-three per cent of the BAC-positive bicyclists had head trauma, whereas the corresponding percentage for the BAC-negative cases was 61% (OR = 8.26; 95% CI 1.48–45.45). Instead, the BAC-negative (50%) bicyclists had injuries in their extremities significantly more often than the BAC-positives (7%) (OR = 12.99; 95% CI 2.35–71.43). However, there was no significant association between the severity of head trauma and BAC on admission either in the total group of patients with head trauma or in any of the subgroups with different causes of trauma.

View this table:
Table 3.

Specific causes of injury and presence or absence of head trauma (HT) and findings on BAC on admission

Cause of injuryTotalHT+HT−TotalHT+HT−
Assault6055 (92)5 (8)44 (100)0
Fall on the ground4736 (77)11 (23)3115 (48)16 (52)
Biking accident2826 (93)2 (7)1811 (61)7 (39)
Motor vehicle crash1817 (94)1 (6)2218 (82)4 (18)
Fall from height99 (100)0129 (75)3 (25)
Sport injury101 (100)228 (36)14 (64)
  • BAC+, positive blood alcohol concentration; BAC−, negative blood alcohol concentration.

The patterns of alcohol drinking among the trauma patients are shown in Table 4. A total of 78% of the patients reported binge-type drinking, whereas only 8% were alcoholics. Another 8% of the patients were light-to-moderate drinkers, and 6% were non-drinkers. The predominant drinking pattern was binge drinking both for men and for women, but men were more often frequent binge drinkers than women. In contrast, women were more often infrequent binge drinkers and light-to-moderate drinkers than men. Men were hazardous drinkers (i.e. dependent drinkers or frequent binge drinkers) significantly more often than women (75% vs 53%, OR = 2.71; 95% CI 1.66–4.42).

View this table:
Table 4.

Drinking patterns of the 349 trauma patients who were interviewed about their alcohol consumption

Pattern of drinkingAllMenWomenDifference between proportions (95% CI)*
n = 349n = 252n = 97
Dependent drinkers, (%)26 (8)26 (10)010 (5–15)
Frequent binge drinkers214 (61)163 (64)51 (53)11 (1–24)
Infrequent binge drinkers59 (17)34 (14)25 (26)12 (3–23)
Light-to-moderate drinkers28 (8)14 (6)14 (14)9 (2–18)
Non-drinkers22 (6)15 (6)7 (7)1 ( 9–4)
  • * Difference between men and women (percentage).

The patterns of alcohol drinking in the groups classified according to the different types of trauma are shown in Table 5. Dependent alcohol drinking and frequent binge drinking were found to be the most common patterns among the patients with head trauma. These drinking patterns occurred significantly more often in the patients with head trauma than in those with other types of trauma (77% vs 59%, OR = 2.38; 95% CI 1.50–3.77).

View this table:
Table 5.

Drinking patterns of the interviewed trauma patients (n = 349) according to the type of trauma

Pattern of drinkersAllHead trauma(%)Other trauma(%)Both head and other trauma (%)
Dependent drinkers2623 (88)2 (8)1 (4)
Frequent binge drinkers21490 (42)85 (42)39 (16)
Infrequent binge drinkers5921 (36)31 (54)7 (10)
Light-to-moderate drinkers286 (21)20 (71)2 (7)
Non-drinkers226 (27)12 (59)4 (14)

The analyses on the patterns of alcohol drinking according to the cause of injury showed that the patients injured by assaults included dependent drinkers and frequent binge drinkers, but other drinking patterns were absent in this group (data not shown). All the dependent drinkers had been hit on the head, whereas 5 frequent binge drinkers had not (the difference was not statistically significant). Forty-two of the 64 patients (66%) injured by falls on the ground were dependent drinkers or frequent binge drinkers, and 22 patients (34%) were infrequent binge drinkers or light-to-moderate drinkers. There were no non-drinkers in this group. Sixty-nine per cent of hazardous drinkers (i.e. dependent drinkers and frequent binge drinkers) had head traumas, whereas the corresponding percentage was 46% in the group of non-hazardous drinkers (i.e. infrequent binge drinkers and light-to-moderate drinkers) (OR = 2.67; 95% CI 0.92–7.75). Of the injured bicyclists, 82% (37/45) were dependent drinkers, frequent binge drinkers, or infrequent binge drinkers. Eighty-seven per cent of them were injured on the head, whereas only 50% of the group of light-to-moderate drinkers/non-drinkers were injured on the head (OR = 6.40; 95% CI 1.20–34.20). The dependent drinkers, frequent binge drinkers and infrequent binge drinkers with head injury had higher BACs on admission than those without head injury after both falls on the ground (mean ± SD, 205 ± 149 vs 70 ± 87 mg/dl; P < 0.01) and bicycle accidents (mean ± SD, 179 ± 111 mg/dl vs 47 ± 65 mg/dl; P < 0.01).


The present study in a cohort of young adults and working aged individuals shows that binge-type drinking is the most typical pattern of alcohol drinking in trauma patients. The data further shows, that high BACs are more frequent among the victims of head trauma than in those with other types of trauma. The data are consistent with the view that alcohol intoxication specifically increases the risk for head trauma.

These findings emphasize the adverse consequences of binge drinking, which may also manifest as severe intoxications in individuals who are not regular drinkers and who do not have increased tolerance to alcohol. Naimi et al. (2003) reported recently that the overall prevalence of binge drinking is 14% among US adults and found that 73% of all binge drinkers can be classified as moderate drinkers in terms of their total alcohol consumption. The number and proportion of binge drinkers in Finland is known to be relatively high. Mäkelä et al. (2001) found that the mean frequency of binging in Finland is ∼11 times a year. At the time of the present study the total amount of drinking in the area (Northern Finland), in which the study was conducted (7.9 l of pure ethanol/capita) has been somewhat lower than the average in Finland (8.8 l). Recent national statistics have, however, shown that only 8% of adult men and 9% of women are abstainers, whereas a total of 51, 31, 20 and 15% of men between 15–19, 20–29, 30–49 and 50–69 years report binge drinking at least once a month, respectively. The corresponding percentages for women are 23, 12, 5 and 3%, respectively (Metso et al., 2000). The high binging rate among the trauma patients in our study supports previous findings, indicating that the incidence of trauma is concentrated particularly in this type of population (Cherpitel et al., 1995; McLeod et al., 1999; Borges et al., 2004). Future studies should also address the possibility whether there is an impact of weekend drinking and the time of day on the occurrence of alcohol-related traumas.

In our study, we found that BACs >150 mg/dl markedly increase the relative risk of sustaining head trauma. To our knowledge, this is the first study to report how the risk for head trauma specifically increases upon increasing BAC, even though high proportions of BAC-positive patients with head trauma have also been reported earlier (Rimel et al., 1982; Brismar et al., 1983; Corrigan 1995; Dikmen et al., 1995). Our finding that 51% of all trauma patients had alcohol in their blood on admission is in good agreement with previous studies (Rivara et al., 1993). However, in studies that included only patients with head trauma, the estimated frequencies of BAC-positives have varied from 32 to 73% (Rimel et al., 1982; Brismar et al., 1983; Kraus et al., 1989; Sparadeo and Gill, 1989; Ruff et al., 1990; Porter, 2000). In our series, the corresponding percentage was high (65%), whereas the subjects with other types of injury showed a lower percentage (32%). It should be noted that in previous studies, only part of the patients have been tested for BAC, and the characteristics of study design may also have differed from ours (Rimel et al., 1982; Brismar et al., 1983; Kraus et al., 1989; Sparadeo and Gill, 1989; Ruff et al., 1990; Porter, 2000).

The causes of head trauma under the influence of alcohol were typically assaults, falls and biking accidents. Not surprisingly, nearly all those who were injured by violence were BAC-positive on admission, and all the assault victims were dependent drinkers or frequent binge drinkers. The high prevalence of hazardous alcohol drinking among assault victims has also been noted earlier (Brismar et al., 1983). In accordance with recent findings by Li et al. (2001), the risk for bicycling injury was also found to be increased among patients who were BAC-positive. In this study, bicycling under the influence of alcohol also increased the risk for head traumas, whereas the bicyclists without any alcohol in blood more often avoided head trauma and injured their extremities upon accidents. This probably reflects the deleterious effects of alcohol on psychomotor skills and the lack of preventive mechanisms to respond to situational hazards, which, in turn, may favour the occurrence of head trauma. This view was further supported by the finding that those who were also injured owing to falls and were BAC-positive were more frequently injured on the head than their sober counterparts.

Here, we did not find any association between the severity of head trauma and positive blood alcohol levels. Although some trials have indicated that alcohol could even protect from the consequences of injury, it is currently known that alcohol intoxication is associated with the occurrence of serious injuries and death from a wide variety of causes (Ward et al., 1982; Waller et al., 1986; Li et al., 2001). Our series did not include the victims who died immediately on the scene. Obviously, many accidental deaths occur before hospital admission (Waller et al., 1986) and thus, future population-based studies appear warranted to specifically address this issue. In this work, we also excluded children and those who were at the age of ≥50 years. In these groups, the occurrence of excessive alcohol consumption and injuries may be different from those observed here (Luukinen et al., 1999; Porter, 2000). Indeed, traumas are known to be major causes of morbidity, mortality and persistent functional impairment, particularly among young adults and working aged individuals (Kraus, 1993; Levin, 1993).

Interestingly, in this series of trauma patients we found a relatively low percentage of dependent drinkers (8%) compared with frequent binge drinkers (61%). Previously, the methods used to detect alcohol problems have varied greatly (Rimel et al., 1982; Brismar et al., 1983; Sparadeo and Gill, 1989; Kreutzer et al., 1990; Ruff et al., 1990; Kreutzer et al., 1991; Drubach et al., 1993; Wong et al., 1993). The lowest incidences of alcohol-related accidents, such as 16% (Rimel et al., 1982), 25% (Sparadeo and Gill, 1989) and 36% (Wong et al., 1993) have been reported in studies in which the assessments have been based on retrospective chart reviews. The highest incidences i.e. 66% (Kreutzer et al., 1991) and 58% (Kreutzer et al., 1990) have been reported from series drawn from rehabilitation centers. Approximately 45% of trauma patients have previously reported alcohol misuse based on screenings with the CAGE and SMAST questionnaires (Nilssen et al., 1994).

In conclusion, this study shows that binge drinking is the characteristic pattern of drinking among trauma patients, and that the injuries of such patients typically result from assaults, falls and biking accidents. Excessive alcohol consumption appears to specifically increase the risk for head trauma as a function of increasing BACs.


We are grateful to Risto Bloigu, University of Oulu, for his help in preparing this article. This study were supported in part by Oulu Medical Research Foundation (OS) and by the Finnish Foundation of Alcohol Studies (ON).


View Abstract