| | Relationships Among Premorbid Alcohol Use, Acute Intoxication, and Early Functional Status After Traumatic Brain InjuryAbstract Vickery CD, Sherer M, Nick TG, Nakase-Richardson R, Corrigan JD, Hammond F, Macciocchi S, Ripley DL, Sander A. Relationships among premorbid alcohol use, acute intoxication, and early functional status after traumatic brain injury. ObjectiveTo investigate the relationships among intoxication at time of injury, preinjury history of problem drinking, and early functional status in patients with traumatic brain injury (TBI). DesignProspective cohort study. SettingAcute inpatient TBI rehabilitation. ParticipantsParticipants were 1748 persons with TBI. InterventionsNot applicable. Main Outcome MeasuresBlood alcohol levels (BALs) were obtained at admission to the emergency department, and a history of problem drinking was obtained through interview. Study outcomes, Disability Rating Scale (DRS), and FIM instrument scores were gathered at admission to inpatient rehabilitation. ResultsMultivariate regression analysis revealed that BAL and a history of binge drinking were predictive of DRS, but not FIM, scores. A higher BAL was associated with poorer functional status on the DRS. Paradoxically, a history of binge drinking was associated with more intact functional status on the DRS. ConclusionsThe relationships among intoxication at time of injury, history of problem drinking, and early outcome after TBI were modest. Injury severity had a more significant association with TBI functional status. THE ASSOCIATION BETWEEN alcohol use and traumatic brain injury (TBI) has been commented on in numerous reports, with alcohol use being implicated as a contributing cause of the injury and as an indicator of poor long-term recovery in the postacute stage. Many studies have investigated the incidence rate of intoxication at the time of injury, as measured by the blood alcohol level (BAL) on admission to the hospital, and found that approximately one third to one half of people admitted to the hospital after a TBI show a positive BAL, with approximately one quarter showing BALs above the legally recognized intoxication level (usually .08 to .10mg/dL, depending on individual state laws during the time period of data collection).1, 2, 3, 4 In addition, evidence suggests that one half to two thirds of TBI admissions have a history of heavy alcohol use or abuse,2, 5, 6, 7, 8 whereas estimates in the general U.S. population are 7% to 8% for heavy alcohol use.9 However, for young men, who have the greatest risk of TBI, almost 50% report drinking 5 or more drinks of alcohol on at least 1 occasion during the past year.9 Several studies6, 8, 10, 11 have also documented the substantial relationship between acute intoxication and premorbid history of alcohol abuse, with estimates ranging to a high of 95% of patients with a positive admission BAL who report a history of alcohol abuse.8 Intoxication at the time of injury has been shown to complicate medical management of the TBI patient in the acute setting by increasing the risk of medical complications.7, 12, 13, 14, 15, 16, 17 As a result, several investigations have addressed the question of the relationship between alcohol-related variables and outcome. These investigations have yielded inconsistent results. Although some studies have shown that acute intoxication is related to an increased length of hospitalization,3, 8 longer duration of coma,18 and lower global cognitive outcome and independence in self-care activities,8, 19 other investigations7, 8, 20, 21, 22 have not shown these effects. Studies exploring the association of chronic alcohol abuse with early postinjury functioning have suggested that this variable is related to an increased length of hospitalization and poorer global outcome,7, 23 although there is again conflicting evidence of this effect.24 Overall, there are several lines of evidence that suggest that acute preinjury and chronic alcohol use may exacerbate TBI medical sequelae but inconsistencies as to whether alcohol variables influence early functioning and outcome. These inconsistencies may be because of differences across reports in study-sample characteristics, such as differences in injury characteristics and demographic variables, availability and variability in BAL and premorbid alcohol use assessment, and time to outcome criterion assessment. Additionally, prior studies have not taken into account the considerable overlap between BAL and premorbid history of problem drinking; the relationship between 1 of these alcohol-related variables and early functional status may be obscured or moderated by the presence of the other alcohol variable. The next step in clarifying the relationships between BAL on admission, premorbid history of problem drinking, and early functional status after TBI would be to investigate the association of each of these alcohol-related variables separately and in combination to explore if the presence of both a positive BAL and a history of problem drinking shows an additive relationship with acute functioning and an outcome above that seen with either variable in isolation. The present study sought to clarify the predictive value of intoxication at injury and a premorbid history of problem drinking on early postinjury functioning and outcome by investigating the predictive value of these variables in isolation and in conjunction to assess for a possible interaction effect. In the present study, “early” is defined as the time of admission to acute inpatient rehabilitation immediately after critical care treatment after TBI, and “functioning and outcome” is defined as the ability to perform and level of independence in self-care activities. Because previous studies24, 25 have shown a complex relationship between these variables and demographic characteristics such as age and education, the influence of these characteristics on the predictive value of admission BAL and a history of problem drinking was also investigated in addition to the association these variables have with acute hospitalization length of stay (LOS). Because there has been evidence in the literature that the potential adverse impact of acute intoxication is limited to early postinjury functioning and may be ameliorated over time,1, 3 the present study used global functioning and outcome measures at the time of acute rehabilitation hospitalization admission as the primary outcome indicators. Our specific hypotheses were that, after controlling for pertinent characteristics, (1) intoxication at the time of injury would be associated with poorer early postinjury functional status, (2) a history of binge drinking would be associated with poorer early postinjury functional status, (3) a history of drinking frequency would be related with poorer early postinjury functional status, and (4) an interaction of acute intoxication and a history of problem drinking would exist, such that these factors in combination would have an additive connection to early postinjury functional status. Methods  Participants All participants sustained a TBI and were enrolled in the Traumatic Brain Injury Model Systems (TBIMS) Research and Demonstration Program, a multicenter study of outcome after TBI funded by the National Institute on Disability and Rehabilitation Research. Inclusion criteria for the TBIMS program include the following: (1) was at least 16 years of age at time of injury, (2) arrived at a TBIMS level I trauma center within 72 hours of injury, (3) received both acute care and inpatient rehabilitation in hospitals designated as TBIMS centers, and (4) provided informed consent. As of March 31, 2003, the TBIMS database included 3893 persons who had sustained injuries since the inception of the project. Because the coding scheme for recording premorbid alcohol-use history was revised for patients injured after October 1, 1999, only patients who were injured after this date were included in subsequent analyses to maintain consistency of alcohol-use group definition. Figure 1 shows the derivation of the final study sample. Of 2169 patients available for study, 421 were excluded either because of missing 2 or more of the 3 outcome variables or missing 2 or more of the 3 drinking-related variables. This resulted in 1748 patients whose data were submitted for analysis. Seventeen centers were involved in the study, and 5 of the centers represented almost half of the subjects (49% [13% NC, 10% OH, 10% WA, 8% MS, 8% AL]). Outcome Measures Acute global outcome was assessed with the Disability Rating Scale (DRS)26 within 24 hours of the patient’s admission to acute rehabilitation. The DRS ranges from 0 to 29 points (higher scores indicate poorer outcome) and assesses abilities ranging from level of responsiveness to employment capability. Overall level of functioning in various self-care activities was measured with the subscales of the FIM instrument.27 The FIM provides ratings of the amount of assistance required for the patient to complete various activities of daily living in both physical (FIM motor) and cognitive (FIM cognitive) domains. The level of independence in each activity is rated on a 1 (complete dependence) to 7 (complete independence) scale. The total FIM motor score ranges from 7 to 91, and the total FIM cognitive score ranges from 5 to 35. DRS and FIM ratings were completed by consensus of the rehabilitation treatment team. Procedure and Analyses Informed consent was obtained from the patient or a proxy in case of patient confusion or nonresponsiveness on admission to all participating institutions. Because study participants were enrolled after admission to the inpatient rehabilitation centers, some data elements (eg, admission Glasgow Coma Scale [GCS] score, BAL) were abstracted retrospectively from acute care records, while other data elements (eg, DRS and FIM scores) were collected prospectively as part of the planned research protocol. Demographic and injury data were collected by an interview with the patient or knowledgeable source (eg, family member, significant other) and abstraction from medical records, according to the TBIMS protocol.28 BAL screening on trauma center admission was coded from available medical records and was recorded in milligrams per deciliter. Data to determine alcohol use history were based on patients’ alcohol use within the month before injury and were gathered by an interview either with the patient or a knowledgeable source (eg, family member, significant other). This “best source” method of determining alcohol-use history is considered reasonably accurate and unbiased because prior works29 have shown high levels of agreement between patient- and family-reported alcohol use. The 1748 patients who were included in subsequent analyses were divided into 4 groups based on the trauma center admitting BAL screening: (1) normal screen, no BAL detected; (2) positive BAL screen below the level of legal intoxication (<.08); (3) positive BAL screen above the level of legal intoxication but less than twice the legal limit (.08 to <.16); and (4) positive BAL twice the legal limit or greater (.16+). Problem drinking was defined in 3 ways, according to criteria proposed by Corrigan et al23 and the Dietary Guidelines for Americans.30 First, the consumption of 5 or more drinks on 1 or more occasions in the month before the injury was classified as binge drinking. Second, drinking frequency was categorized as abstinent (no alcohol), infrequent or light (<3 drinks a week), moderate (3−14 drinks a week for men, 3−7 drinks a week for women), or heavy (>14 drinks a week for men, >7 drinks a week for women). Descriptive statistics (frequencies and percentages for categorical measures and means and standard deviations [SDs] for numeric measures) were computed for the total sample. Bivariate associations were examined by using Spearman rank correlation coefficients (ρ). Simple (unadjusted analyses) and multivariate (adjusted analyses) linear regression models were used to test hypotheses regarding prediction of outcomes at rehabilitation admission. BAL, binge drinking, and drinking frequency were entered as predictor variables, in addition to sex, age, educational level, LOS at the hospital at which the patient received acute care immediately following injury (LOS), and emergency department admission GCS score in a series of multivariate regression analyses with DRS total score, FIM cognitive, and FIM motor scores at admission as the criterion variables. To address possible confounding, comparisons were made between the unadjusted and adjusted analyses to evaluate the unique effects of the predictors. For all multivariate linear regression models, effects for continuous variables were shown as interquartile coefficients. Interquartile coefficients indicate the change in the value of the outcome (dependent variable) as the value of the predictor changes from the 25th to the 75th percentile. Although the number of subjects in the study sample was 1748, the number of subjects for the bivariate correlations and univariate statistics ranged from 1299 to 1748 depending on the association and variable studied. However, the number of subjects in each regression model included all subjects (N=1748) by using statistical imputation (ie, for the 3 regression models, any value missing on the predictors were plugged in [imputed] by using individual predictive models using the interrelationships among the other predictors and outcomes). A multiple-imputation strategy was used to take into account the uncertainty of the imputations, and 50 datasets were generated. The results from the 50 analyses (coefficients, confidence intervals [CIs], P values) were combined to produce the findings reported here. This approach generates stable datasets and reliable regression models.31 All calculations were performed by using S-PLUS software a and the Hmisc and Design libraries of Harrell.32 All statistical tests used 2-sided tests of significance with a level of .05. Results Descriptive Statistics and Bivariate Associations Demographic characteristics for the 1748 patients are presented in table 1. Most patients were white (73%), were men (72%), and had a mean age of 38 and a range of 16 to 94 years. Thirty-one percent graduated high school or obtained an equivalence degree (General Educational Development [GED] diploma), and 36% had a lower level of educational attainment. Sixty percent of the sample sustained a severe TBI, as evidenced by an emergency department admission GCS score between 3 and 8. Patients had a mean LOS in the acute-care setting of 20 days and a mean emergency department admission GCS score of 8. As compared with the study sample, the 421 patients who were excluded tended to be more impaired, as evidenced by a greater proportion of patients with lower admission GCS scores, longer length of acute care hospitalization, and poorer functional status ratings on rehabilitation admission. | | |  | Categorical Predictors | N=1748 | n=421 | P | |
|---|
| Missing | n | Missing | n | |
|---|
| Sex | 0 (0) | | 0(0) | | .09 | | | Male | | 1261(72) | | 321(76) | | | | Female | | 487(28) | | 100(24) | | | | Education | 48(3) | | 42(10) | | .95 | | | Grades <9 | | 128(8) | | 23(6) | | | | Grades 9–11 | | 467(28) | | 106(28) | | | | GED, high school degree | | 525(31) | | 126(33) | | | | Some college | | 367(22) | | 78(21) | | | | College degree | | 213(13) | | 46(12) | | | | GCS score | 74(4) | | 28(7) | | .01 | | | 3−8 | | 998(60) | | 263(67) | | | | 9−12 | | 190(11) | | 37(9) | | | | 13−15 | | 486(29) | | 93(24) | | | | Binge (5+ drinks) | 46(3) | 548(32) | 361(86) | 17(28) | .53 | | | Drinking frequency | 22(1) | | 55(13) | | .85 | | | Nondrinker | | 728(42) | | 18(33) | | | | Infrequent or light | | 298(17) | | 17(31) | | | | Moderate | | 377(22) | | 17(31) | | | | Heavy | | 323(19) | | 3(6) | | | | BAL | 449(26) | | 136(32) | | .96 | | | 0 | | 694(53) | | 155(54) | | | | 0.01−0.79 | | 209(16) | | 36(13) | | | | 0.80−1.59 | | 124(10) | | 32(11) | | | | ≥1.60 | | 272(21) | | 62(22) | | | | | |
| Numeric Measures | Missing | Mean ± SD | Missing | Mean ± SD | | |
|---|
| Predictors | | | | | | | | Age | 0(0) | 38±17.4 | 0(0) | 37±17.1 | .12 | | | GCS total | 74(4) | 8±4.8 | 28(7) | 7±4.5 | .13 | | | LOS acute | 0(0) | 20±15.6 | 0(0) | 23±16.9 | .003 | | | Outcomes | | | | | | | | DRS admission | 39(2) | 11.7±5.6 | 33(8) | 12.5±6.5 | .06 | | | FIM cognitive admission | 4(<1) | 16.9±8.0 | 48(11) | 15.8±8.5 | .02 | | | FIM motor admission | 137(8) | 39.3±19.5 | 75(18) | 37.4±20.5 | .04 | | | | |
Descriptive statistics by the 3 outcomes (DRS, FIM cognitive, and FIM motor at rehabilitation admission) are presented in table 2. Overall, acute global outcome and functioning status was poorer for lower education levels, higher acute LOS, lower GCS score, and not being a binge drinker. Preliminary correlational analysis indicated that acute LOS correlated negatively with GCS score (r=−.28, P<.01) but was not associated with BAL, binge drinking, or frequency of drinking (ρ range, .02−.04; P>.05). The GCS score showed significant, but small, negative correlations with all 3 alcohol-related measures (ρ range, −.08 to −.09; P<.01). BAL correlated positively with both bingeing and drinking frequency (ρ range, .49−.54; P<.01). Bingeing and drinking frequency correlated strongly and positively (ρ=.74, P<.01) but within acceptable limits in regard to multicolinearity (ie, the highest variance inflation factor for binge drinking and drinking frequency in the regression models was 2.38, well below the cutoff of 10 recommended as an indication for variable deletion33) for the adjusted regression models described later. | | | | Predictors | DRS Admission | FIM Admission | |
|---|
| Cognitive | Motor | |
|---|
| Sex | | | | | | Male | 11.7±5.7 | 16.7±8.0 | 40.4±20.1 | | | Female | 11.5±5.5 | 17.3±8.1 | 39.3±19.5 | | | Education | | | | | | Grades <9 | 12.3±5.5 | 16.0±7.7 | 36.2±16.7 | | | Grades 9−11 | 12.1±5.6 | 16.4±7.9 | 39.1±19.6 | | | GED, high school degree | 11.4±5.7 | 17.2±7.9 | 39.7±19.5 | | | Some college | 11.2±5.6 | 17.4±8.2 | 40.8±20.1 | | | College degree | 11.2±5.5 | 17.6±8.5 | 39.6±20.0 | | | Age (y) | | | | | | <23 (1st quartile) | 11.9±6.1 | 16.7±8.2 | 39.9±20.9 | | | 23−35 (2nd quartile) | 11.8±5.7 | 17.0±8.1 | 40.9±20.6 | | | 36−48 (3rd quartile) | 11.5±5.3 | 16.6±7.9 | 39.7±18.5 | | | 49+ (4th quartile) | 11.4±5.4 | 17.1±7.8 | 36.7±17.6 | | | LOS acute (d) | | | | | | <10 (1st quartile) | 8.7±3.9 | 20.0±6.9 | 49.9±16.5 | | | 10−16 (2nd quartile) | 10.8±5.0 | 17.7±7.5 | 42.4±18.4 | | | 17−27 (3rd quartile) | 13.0±5.6 | 15.2±8.3 | 35.1±19.0 | | | 28+ (4th quartile) | 14.5±6.2 | 14.2±8.1 | 28.4±17.5 | | | GCS score | | | | | | 3−8 | 13.0±5.9 | 15.1±7.9 | 36.1±19.8 | | | 9−12 | 10.5±5.0 | 18.1±7.6 | 44.5±19.4 | | | 13−15 | 9.5±4.5 | 20.0±7.4 | 43.7±17.9 | | | Binge (5+ drinks) | | | | | | No | 11.8±5.8 | 16.7±8.1 | 38.0±19.3 | | | Yes | 11.3±5.3 | 17.1±7.8 | 42.0±19.9 | | | Drinking frequency | | | | | | Nondrinker | 11.9±6.0 | 16.7±8.1 | 37.3±19.1 | | | Infrequent or light | 11.4±5.2 | 17.0±8.1 | 39.5±19.7 | | | Moderate | 11.2±5.4 | 17.4±8.0 | 42.4±20.1 | | | Heavy | 11.8±5.4 | 16.4±7.6 | 40.7±19.3 | | | BAL | | | | | | 0.00 | 11.3±5.6 | 17.4±8.3 | 39.6±20.3 | | | 0.01−0.79 | 11.9±5.5 | 16.5±8.1 | 40.2±19.2 | | | 0.80−1.59 | 12.6±5.6 | 15.6±7.7 | 37.3±19.8 | | | ≥1.6 | 11.9±5.2 | 16.6±7.7 | 40.9±19.1 | | | | |
Regression Results To test for a possible interaction between BAL and alcohol-use history (bingeing, drinking frequency), preliminary multivariate regression analyses were conducted for all 3 outcomes (DRS at admission, FIM cognitive scores, FIM motor scores). No significant interaction effects were found (all P>.75). Consequently, these interaction terms were omitted from all subsequent analyses. The unadjusted (univariate linear regression) and adjusted (multivariate linear regression) effects of age, sex, education, acute LOS, GCS score at admission, and alcohol-related characteristics on DRS at rehabilitation admission are shown in table 3. For the unadjusted model shown on the left in table 3, BAL was significantly related to DRS (P<.05), as were age, education, LOS, and GCS score. The associations of frequency of drinking and bingeing with DRS were not significant. For BAL, persons over the legal limit and less than double the limit (.08−.159) had DRS admission scores that were .70 units higher (indicating greater impairment on admission) than persons with no alcohol in blood (blood alcohol, 0; 95% CI, 0.03−1.37; P=.05). The results of the multivariate linear regression analysis of DRS at admission were similar to the unadjusted analyses. Age, LOS, and GCS emerged as significant independent predictors. The BAL relationship was also significant and was stronger than in the unadjusted analyses (P=.02). A higher BAL was associated with a .85-unit increase in DRS scores (greater disability at admission). The association of binge drinking with DRS scores was stronger and now significant in the adjusted analyses (P=.02). The relationship between binge drinking and DRS scores was such that persons reporting binge drinking had DRS scores almost 1 unit less than persons who did not. The unadjusted and adjusted effects of the predictors on FIM cognitive scores at the time of admission are shown in table 4. For the unadjusted model shown on the left in the table, the alcohol-related variables were not related to FIM cognitive scores. Education, LOS, and GCS score had significant effects on FIM cognitive scores (P<.05). The results of the adjusted analyses for predicting cognitive FIM scores at admission are shown on the right of table 4, and they were similar to the unadjusted analyses, with the exception of age emerging as a significant independent predictor. Again, the alcohol-related variables did not significantly predict FIM cognitive scores. The unadjusted and adjusted effects of the predictors on FIM motor scores at admission are shown in table 5 on the left and right, respectively. For the unadjusted model, sex, age, LOS, GCS score, drinking frequency, and bingeing were related to FIM motor scores (P<.05), with higher levels of alcohol use associated with better outcomes. The results of the adjusted analyses for predicting FIM motor scores at admission again indicated that sex, age, LOS, and GCS score were significant independent predictors (P<.05). However, in the multivariate linear regression, BAL, frequency of drinking, and bingeing were not predictive of FIM motor scores. Discussion Analyses of data from the current investigation provided only limited support for our hypotheses. Multivariate regression analyses showed a modest relationship between emergency room BAL and acute global outcome on the DRS on admission to the acute rehabilitation setting, even when the relationship was adjusted for other predictors of functional capacity. A higher BAL was associated with poorer global outcome on this measure. Contrary to our hypotheses, neither BAL nor alcohol use history (binge drinking or drinking frequency) were related to level of functional independence as measured by rehabilitation admission FIM motor or cognitive scores when other known predictors of outcome were considered. These findings are generally consistent with prior investigations.25, 34 Also contrary to expectation, BAL and alcohol use history did not demonstrate an interactive relationship on any indicator of functional status. Surprisingly, a history of binge drinking was associated with a better global outcome on the DRS, even after considering other predictors of functional capacity. It is possible that persons who are more productive before injury, such as working full-time or attending school full-time, have fewer occasions to drink but tend to drink several drinks, or binge, on the few occasions that they consume alcohol. Preinjury productivity has been shown to be related to outcome, and the relationship between binge drinking and outcome in the current study may be secondary to a relationship between productivity and binge drinking. This possibility is somewhat supported by the finding that lower education was predictive of outcome in the current study, and lower levels of education are associated with lower levels of preinjury productive activity. Correlational analyses indicated that BAL, drinking frequency, and binge drinking were very weakly, albeit statistically significantly, related to injury severity at emergency department admission, with a higher BAL, a greater frequency of drinking, and a history of binge drinking being associated with lower GCS scores. This finding is consistent with the results of studies3, 4, 19, 20 showing that other associated measures of injury severity, such as duration of coma, overall level of consciousness, and length of acute hospitalization, are related to alcohol variables. But, again, it should be noted that this effect was very small and has questionable clinical significance. Additionally, the interpretation of this finding must take into consideration the possibility that intoxication at the time of injury can directly, but temporarily, impact GCS scores by decreasing a person’s ability to verbalize coherently and to follow commands. GCS scores may increase once the effect of acute intoxication wears off. The current finding of an association between preinjury alcohol history and GCS scores may be an indirect effect of the association of BALs with preinjury alcohol use because these 2 variables are highly correlated. Overall, the results of the current study suggest that alcohol use, either chronically or at the time of injury, and early postinjury functional status are very weakly associated. Although the current findings suggest modest associations between a history of alcohol use, alcohol at time of injury, and early functional status after TBI, our results did confirm prior findings regarding other factors that are related to early outcome from TBI. Consistent with many previous investigations,24, 25, 35 demographic factors were related to early outcome. A higher age was associated with poorer early functional status, whereas a higher education level was associated with a more intact early functional status. Although it may be intuitively compelling that a higher age is associated with greater vulnerability to the effects of injury, it is more difficult to give a clear explanation of why higher levels of education should be beneficial to very early functional status outcomes including physical independence. Also consistent with many previous investigations, injury severity is a powerful predictor of early functional outcome. The initial level of responsiveness (GCS score) was predictive of early independence, cognitive, and motor status even when adjusted for acute-care LOS. One factor that may partially account for the lack of relationship between the alcohol-related variables and functional status in this study concerns the demographic makeup of the study sample. The sample was comprised mostly of younger subjects who typically do not show as pronounced effects of misuse because of the relatively shorter time they have engaged in the behavior. Although the current sample differed from the general adult population in preinjury alcohol consumption (one quarter of adults binge at least once in 30 days and one tenth are heavy drinkers in the general population36 vs one half binge and one third heavy drinkers in current sample), the present sample overlaps with a demographic that has higher rates of alcohol use (younger men), as previously noted.9 The relatively young age (median age, 35.5y) of the current sample may have mitigated the relationship between alcohol misuse and early functional status. Persons with longer histories of high levels of alcohol use may show a greater vulnerability to the effects of TBI on functional status. Study Limitations The results of the present study should be viewed in light of several potential weaknesses. The outcome measures used in the current study were limited to global measures of functional status assessed at the time of admission to inpatient rehabilitation. It is possible that preinjury alcohol consumption is related to functional status only through subtler mediating variables, such as higher-level cognitive functioning. The relationship between alcohol and cognitive functioning is seen in prior studies37, 38 with persons with a history of alcohol abuse. Higher-level cognitive deficits may not necessarily be reflected in global ratings on the DRS and FIM. Additionally, the measures used to represent preinjury alcohol history in the current study (frequency of drinking and binge drinking) are not synonymous with alcohol abuse or dependence. It is possible that alcohol consumption potentiates the effect of TBI, thus having a stronger association with global functioning and outcome, only if the person has passed a certain threshold of consumption, moving from heavy or binge drinking to alcohol abuse or dependence. This study is also subject to the typical weaknesses inherent in a multicenter database study, such as variations in data collection between centers, a significant portion of missing data in the database and how this affects the statistical analysis, unequal contributions among centers to the database, and cultural differences between different areas of the country. Additionally, the 421 excluded patients with available outcome data tended to be more impaired than the study sample, thus possibly biasing the results toward patients with less severe impairment. The extent to which patients receiving rehabilitation at TBIMS centers are similar to the general population of patients with TBI is unknown, although the wide geographic representation of TBIMS sites provides some reassurance about possible generalization of findings. This study included only those individuals who survived the injury and progressed to admission to inpatient rehabilitation. This study did not include those people who expired early after injury, improved beyond requiring inpatient rehabilitation, or were discharged from acute care to an alternative level of care such as a nursing home. Prior studies of the TBIMS national database have revealed that variables found predictive of brain injury outcome in trauma series may not be predictive of outcome for a rehabilitation only sample. For example, GCS score and Revised Trauma Score, which are strong predictors of morbidity and mortality in acute care, were found not to have strong relationships to early functional outcome in a rehabilitation sample.39, 40 The findings of the current investigation leave questions unanswered regarding the complex relationships among a history of alcohol use, alcohol at the time of injury, and early functional status after TBI. Future studies should use both measures of alcohol consumption and structured diagnostic interviews that assess criteria for alcohol abuse and dependence from the Diagnostic and Statistical Manual of Mental Disorders, 4th Edition,41 to investigate if abuse or dependence potentiates the effect of TBI, thus strengthening any potential association with global functioning or outcome. Given the association between a poorer higher-level cognitive functioning and a history of alcohol abuse,37, 38 future studies including more specific measures of cognitive outcome may be helpful to explore if diminished cognitive functioning mediates the relationship between alcohol use and functional status after TBI. It is important to note that this study investigated early outcomes at the time of rehabilitation admission. The relationship of alcohol-use history with longer-term outcomes was not examined. We do not yet know whether or not chronic premorbid alcohol use is associated with poorer long-term outcomes. Conclusions The present investigation found only very limited and weak associations between alcohol use and early outcome after TBI. It appears that the predominantly young, male sample was resilient to any early negative impact of acute intoxication, history of binge drinking, or history of frequent alcohol use. Supplier Acknowledgments Data for subjects reported on in this study were contributed by the TBIMS programs at University of Alabama, Birmingham, AL; Santa Clara Valley Medical Center, San Jose, CA; Craig Hospital, Englewood, CO; Emory University/Shepherd Center, Atlanta, GA; Spaulding Rehabilitation Hospital, Boston, MA; Mayo Clinic, Rochester, MN; Rehabilitation Institute of Michigan, Detroit, MI; Methodist Rehabilitation Center, Jackson, MS; University of Missouri, Columbia, MO; Kessler Medical Rehabilitation Research and Education Corp, West Orange, NJ; Carolinas Rehab, Charlotte, NC, Ohio State University, Columbus, OH; Oregon Health Sciences University, Portland, OR; Moss Rehabilitation Research Institute, Philadelphia, PA; The Institute for Rehabilitation and Research, Houston, TX; and Virginia Commonwealth University/Medical College of Virginia, Richmond, VA; University of Washington, Seattle, WA. References 1. 1Bombardier CH, Thurber CA. Blood alcohol level and early cognitive status after traumatic brain injury. Brain Inj. 1998;12:725–734. MEDLINE |
CrossRef
2. 2Corrigan JD. Substance abuse as a mediating factor in outcome from traumatic brain injury. Arch Phys Med Rehabil. 1995;76:302–309. Abstract |
Full-Text PDF (1027 KB)
|
CrossRef
3. 3Kaplan CP, Corrigan JD. Effect of blood alcohol level on recovery from severe closed head injury. Brain Inj. 1992;6:337–349. MEDLINE |
CrossRef
4. 4Tate PS, Freed DM, Bombardier CH, Harter SL, Brinkman S. Traumatic brain injury: influence of blood alcohol level on post-acute cognitive function. Brain Inj. 1999;13:767–784. MEDLINE |
CrossRef
5. 5Kreutzer JS, Doherty KR, Harris JA, Zasler ND. Alcohol use among persons with traumatic brain injury. J Head Trauma Rehabil. 1990;5:9–20.
CrossRef
6. 6Kreutzer JS, Witol AD, Sander AM, Cifu DX, Marwitz JH, Delmonico R. A prospective longitudinal multicenter analysis of alcohol use patterns among persons with traumatic brain injury. J Head Trauma Rehabil. 1996;11:58–69.
CrossRef
7. 7Ruff RM, Marshall LF, Klauber MR, et al. Alcohol abuse and neurological outcome of the severely head injured. J Head Trauma Rehabil. 1990;5:21–31.
CrossRef
8. 8Sparadeo FR, Gill D. Effects of prior alcohol use on head injury recovery. J Head Trauma Rehabil. 1989;4:75–82.
CrossRef
9. 9National Center for Health Statistics. Health, United States, 2006 with chartbook on trends in the health of Americans. Hyattsville: NCHS; 2006;. 10. 10Bombardier CH. Alcohol use and traumatic brain injury. West J Med. 1995;162:150–151. MEDLINE 11. 11Dikmen SS, Machamer JE, Donovan DM, Winn HR, Temkin NR. Alcohol use before and after traumatic head injury. Ann Emerg Med. 1995;26:167–176. Abstract | Full Text |
Full-Text PDF (1306 KB)
|
CrossRef
12. 12Gurney JG, Rivara FP, Mueller BA, Gurney JG, Rivara FP, Mueller BA. The effect of alcohol intoxication on the initial treatment and hospital course of patients with acute brain injury. J Trauma. 1992;33:709–713. MEDLINE 13. 13Jurkovich GJ, Rivara FP, Gurney JG, Seguin D, Fligner CL, Copass M. Effects of alcohol intoxication on the initial assessment of trauma patients. Ann Emerg Med. 1992;21:704–708. Abstract |
Full-Text PDF (479 KB)
|
CrossRef
14. 14Steinbok P, Thompson G. Metabolic disturbances after head injury: abnormalities of sodium and water balance with special reference to the effects of alcohol intoxication. Neurosurgery. 1978;3:9–15. 15. 15Leonard BE. The involvement of neuromediation in alcohol abuse and alcoholism. Neurochem Int. 1995;26:343–346. MEDLINE |
CrossRef
16. 16Nevo I, Hamon M. Neurotransmitter and neuromodulatory mechanisms involved in alcohol abuse and alcoholism. Neurochem Int. 1995;26:305–336. MEDLINE |
CrossRef
17. 17Markwalder T. Chronic subdural hematomas: a review. J Neurosurg. 1981;54:637–645. MEDLINE |
CrossRef
18. 18Kraus JF, Morgenstern H, Fife D, Conroy C, Nourjah P. Blood alcohol tests, prevalence of involvement, and outcomes following brain injury. Am J Public Health. 1989;79:294–299. MEDLINE |
CrossRef
19. 19Drubach DA, Kelly MP, Winslow MM, Flynn JP. Substance abuse as a factor in the causality, severity, and recurrence rate of traumatic brain injury. Md Med J. 1993;42:989–993. MEDLINE 20. 20Edna T. Alcohol influence and head injury. Acta Chir Scand. 1982;148:209–212. MEDLINE 21. 21Pories SE, Gamelli RL, Vacek P, Goodwin G, Shinozaki T, Harris F. Intoxication and injury. J Trauma. 1992;32:60–64. MEDLINE 22. 22Lehmkuhl LD, Hall KM, Mann N, Gordon WA. Factors that influence costs and length of stay of persons with traumatic brain injury in acute care and inpatient rehabilitation. J Head Trauma Rehabil. 1993;8:88–100.
CrossRef
23. 23Corrigan J, Bogner J, Lamb-Hart G, et al. Problematic substance use identified in the TBI Model Systems national database (The Center for Outcome Measurement in Brain Injury). http://tbims.org/combi/subst/index.html2005;. 24. 24Brooks N, Symington C, Beattie A, Campsie L, Bryden J, McKinlay W. Alcohol and other predictors of cognitive recovery after severe head injury. Brain Inj. 1989;3:235–246. MEDLINE |
CrossRef
25. 25Kelly MP, Johnson CT, Knoller N, Drubach DA, Winslow MM. Substance abuse, traumatic brain injury and neuropsychological outcome. Brain Inj. 1997;11:391–402. MEDLINE |
CrossRef
26. 26Rappaport M, Hall KM, Hopkins K, Belleza T, Cope DN. Disability rating scale for severe head trauma: coma to community. Arch Phys Med Rehabil. 1982;63:118–123. MEDLINE 27. 27Wright J. The FIM (The Center for Outcome Measurement in Brain Injury). http://www.tbims.org/combi/FIM2002;. 28. 28Traumatic Brain Injury Model Systems National Data and Statistical Center: on-line syllabus. Available at: http://www.tbindsc.org/Syllabus/live/live.asp. Accessed February 12, 2007. 29. 29Sander AM, Witol AD, Kreutzer JS. Alcohol use after traumatic brain injury: concordance of patients’ and relatives’ reports. Arch Phys Med Rehabil. 1997;78:138–142. Abstract |
Full-Text PDF (689 KB)
|
CrossRef
30. 30U.S. Department of Health and Human Services. Dietary guidelines for Americans. http://www.healthierus.gov/dietaryguidelines/2005;. 31. 31Harel O, Zhou XH. Multiple imputation: review of theory, implementation and software. Stat Med. 2007;26:3057–3077. MEDLINE |
CrossRef
32. 32Harrell FE. Hmisc and Design Libraries. http://biostat.mc.vanderbilt.edu/twiki/bin/view/Main/RS2006;. 33. 33Myers R. Classical and modern regression with applications. 2nd ed.. Boston: Duxbury Pr; 1990;. 34. 34Turner AP, Kivlahan DR, Rimmele CT, Bombardier CH. Does preinjury alcohol use or blood alcohol level influence cognitive functioning after traumatic brain injury?. Rehabil Psychol. 2006;51:78–86. 35. 35Bush BA, Novack TA, Malec JF, Stringer AY, Millis SR, Madan A. Validation of a model for evaluating outcome after traumatic brain injury. Arch Phys Med Rehabil. 2003;84:1803–1807. Abstract | Full Text |
Full-Text PDF (67 KB)
|
CrossRef
36. 36Centers for Disease Control and Prevention. Alcohol & public health. Available at: http://www.cdc.gov/alcohol. Accessed June 21, 2006. 37. 37Grant I. Alcohol and the brain: neuropsychological correlates. J Consult Clin Psychol. 1987;55:310–324.
CrossRef
38. 38Parsons OA, Nixon SJ. Neurobehavioral sequelae of alcoholism. Neurol Clin. 1993;11:205–218. MEDLINE 39. 39Zafonte RD, Hammond FM, Mann NR, Wood DL, Black KL, Millis SR. Relationship between Glasgow coma scale and functional outcome. Am J Phys Med Rehabil. 1996;75:364–369. MEDLINE |
CrossRef
40. 40Zafonte RD, Hammond FM, Mann NR, Wood DL, Millis SR, Black KL. Revised trauma score: an additive predictor in traumatic brain injury disability. Am J Phys Med Rehabil. 1996;75:456–461. MEDLINE |
CrossRef
41. 41American Psychiatric Association. Diagnostic and statistical manual of mental disorders. 4th ed.. Washington (DC): APA; 1994;. a Methodist Rehabilitation Center, Jackson, MS b University of Mississippi Medical Center, Jackson, MS c Memorial Hermann TIRR, Houston, TX d Center for Epidemiology & Biostatistics, Cincinnati Children’s Hospital Medical Center, University of Cincinnati College of Medicine, Cincinnati, OH e Department of Physical Medicine and Rehabilitation, Ohio State University, Columbus, OH f Carolinas Rehabilitation, Charlotte, NC g Shepherd Center, Atlanta, GA h Craig Hospital, Englewood, CO i Department of Physical Medicine and Rehabilitation, Baylor College of Medicine/Harris County Hospital District, Houston, TX.  Reprint requests to Chad D. Vickery, PhD, Methodist Rehabilitation Center, 1350 E Woodrow Wilson, Jackson, MS 39216
Supported by the National Institute on Disability and Rehabilitation Research (grant no. H133A0205514). No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit upon the author(s) or upon any organization with which the author(s) is/are associated. PII: S0003-9993(07)01592-4 doi:10.1016/j.apmr.2007.07.047 © 2008 American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation. Published by Elsevier Inc. All rights reserved. | |
|
FULL TEXT01592-4/journalimage?img=PIIS0003999307015924.gr1.lrg.gif&fig=fig1&kwhquery=&issn=0003-9993&ishighres=false&allhighres=false&free=yes','journalimage');)
