To our knowledge, this is the first study to examine the relationship between alcohol screening scores (AUDIT-C) and risk of inpatient, residential, and outpatient medical treatment for any type of trauma. This is also one of few studies to look at women’s trauma risk separate from men’s. In male VHA patients, severe alcohol misuse (AUDIT-C 9–12) was associated with a significantly increased risk for trauma, particularly fractures. Men in the oldest age group (>65
years) showed increased risk of medical care for trauma even at lower levels of alcohol misuse (AUDIT C 5–8). These findings augment prior research, which found an association between severe alcohol misuse and specific risk of fracture [17
], and between severe alcohol misuse and trauma-related hospitalization [18
The AUDIT-C was not a predictor of two-year trauma risk in female VHA patients. The absence of an association is counterintuitive, since women have been shown to be more vulnerable to the toxic effects of alcohol [30
]. Other studies of AUDIT-C scores from the SHEP survey and subsequent health outcomes had similar findings, in that associations between AUDIT-C scores and new-onset liver disease, upper gastrointestinal bleeding, and pancreatitis in the subsequent two years was associated with AUDIT-C scores in men but not in women.
Several factors may be affecting the results on for women in the present study. First, increased stigma for heavy drinking among women typically leads to underreporting drinking levels compared with men [31
]. Also, even with a sample of over 9000 women, the number of women with alcohol misuse was relatively small (400 with AUDIT-C scores of 5–8 and 97 with AUDIT-C scores of 9–12), and there were relatively small numbers of total traumas (21) and fractures (5) in the highest AUDIT-C group. In addition, it is possible that women who screened positive for alcohol misuse were less likely to have full VA benefits, which could lead to under-ascertainment of trauma in women with alcohol misuse compared with women without alcohol misuse. Furthermore, even though the survey was confidential, women may have under-reported alcohol use more frequently than men because, given their younger age, they might still have been applying for VA benefits (often from military sexual trauma), and may have worried that, if they accurately reported their drinking, they would be denied. However, these speculations do not explain results from another study with the same sample that found severe alcohol misuse (AUDIT-C 9–12) had an even more pronounced effect on mortality among women than among men [14
]. Taken together, these results highlight the importance of sex-tailored risk information as well as then need for more research to clarify these relationships. Regardless, women who screen positive for alcohol misuse should be counseled to reduce their drinking and advised of the other documented medical risks (e.g., mortality) of at-risk consumption.
Fractures comprised the third leading cause of trauma for men (2.28%), exceeded only by wounds/amputations and musculoskeletal injuries. The association between alcohol misuse and fractures may be related to the increased risk of falls due to intoxication, with potential contribution from neurologic complications of alcohol misuse such as cerebellar degeneration, Wernicke-Korsakoff syndrome, peripheral neuropathy [33
], and increased risk of motor-vehicle accidents [2
]. Furthermore, alcohol misuse has been linked to the development of osteoporosis, which also increases the risk of fractures [34
On average, older men (>65) had a heightened risk of trauma with lower levels of misuse (AUDIT-C 5–8) as well as severe misuse (AUDIT-C 9–12). This finding is consistent with two prior studies of AUDIT-C and future trauma risk [17
], both of which showed lower consumption associated with elevated risk in older men. In our study, among men >65
years, those with severe alcohol misuse (AUDIT-C 9–12) had double the risk of fracture compared with same-aged men drinking in moderation (AUDIT-C 1–4). The heightened alcohol-related risk in older men may be related to a greater propensity to fall and a greater risk of fracture from a fall.
Nondrinkers aged >50
years were at increased risk for injuries and accidents over the two-year follow-up. Previous studies found that nondrinkers were at increased risk for adverse health outcomes, including fractures and primary or secondary trauma-related hospitalization [17
] compared with people drinking at low levels. Nondrinkers were older, had greater morbidity, and had poorer health status, making them more susceptible to falls and resultant fractures. Many people reporting themselves as nondrinkers had likely stopped drinking due to medical problems that could contribute to accidents [35
]. Furthermore, Holahan et al. [26
] demonstrated that abstainers are more likely than moderate drinkers to have had prior drinking problems, to be obese, and to smoke cigarettes.
Although our data do not distinguish between former at-risk drinkers and lifetime nondrinkers, nondrinking men aged <50
years had no increased trauma risk. It is likely that, with age, some formerly heavy-drinking men migrate to the nondrinking group and potentially increase the trauma risk of nondrinkers. Consistent with this hypothesis, older nondrinking men had more comorbid illnesses than younger nondrinking men and older men with alcohol misuse. In other words, the increased risk of trauma among abstainers might be due, in part, to alcohol-related harm from former heavy drinking [26
]. Given the consistency of this finding across studies, it would be important for future studies to distinguish between former problem drinkers and lifetime nondrinkers.
These results need to be understood in light of several limitations. The sample consisted exclusively of patients of the VHA, who are predominantly male and older than the general population. Only 69% of selected patients responded to the SHEP, and only 83% of those provided complete AUDIT-C data. Although we have no reason to expect that the relationship between AUDIT-C scores and subsequent trauma is different in the observed and unobserved samples, this possibility must be considered. Also, with samples this large, some (but certainly not all) of the statistically significant differences in risk were small in absolute and relative terms.
Also, misclassification and under-ascertainment of outcomes threaten the internal validity of this study. Outpatient ICD-9 codes are not as valid as inpatient codes [36
]. Many patients in the VHA system use Medicare and do not necessarily get transferred to VA medical centers, which likely decreases the accurate detection of health-care utilization in the older age group [36
]. Others have limited VA eligibility and may seek emergency care outside the VA health-care network. As described above, the AUDIT-C assesses drinking in the past year but cannot differentiate lifetime abstainers from previous high-risk or problem drinkers. Finally, AUDIT-C data collected in the course of a mailed patient-satisfaction survey may differ in important ways from screening data obtained in the course of clinical care [37
]. Any of these limitations may have caused an over- or underestimation of the magnitude or direction of the association between AUDIT-C scores and trauma.
In summary, the AUDIT-C is a useful scaled marker of two-year risk of trauma across different treatment settings in men. Severe alcohol misuse is most strongly associated with risk of subsequent fracture in older men who are otherwise relatively healthy. Older men with moderate alcohol misuse are also at increased risk. We did not find an association between AUDIT-C scores and risk of trauma diagnoses in women, but we hypothesize that this had more to do with under-ascertainment bias than lack of an association between alcohol use and trauma in women. Our data add to the growing body of evidence in support of the AUDIT-C as a marker for future health risk among men, including trauma, medication nonadherence [11
], gastrointestinal illness [12
], and all-cause mortality [14