Patterns of alcohol and substance use and abuse in persons with spinal cord injury: Risk factors and correlates¹

Article Outline

• Abstract
• Methods
  • Participants
  • Procedures
  • Measures
    • Alcohol- and substance-related questions
    • CAGE questionnaire
    • Outcome measures
  • Statistical analyses
• Results
  • Demographic and injury-related characteristics of the total sample
  • Alcohol consumption and abuse and substance use after SCI
    • Intensity and frequency of drinking patterns
    • Alcohol abuse
    • Substance use
  • Characteristics of SCI drinkers and abstainers
  • Characteristics of substance users and nonusers
  • Differences in medical and psychosocial outcomes
    • Medical outcomes: pain and pressure ulcers
    • Psychosocial outcomes after SCI: satisfaction with life, participation in activities, and occupation
• Discussion
  • Limitations of study
• Conclusions
• Acknowledgements
• References
• Copyright

Abstract 

Tate DG, Forchheimer MB, Krause JS, Meade MA, Bombardier CH. Patterns of alcohol use and abuse in persons with spinal cord injury: risk factors and correlates. Arch Phys Med Rehabil 2004;85:1837–47.

Objective

To investigate patterns of alcohol consumption and abuse and substance use among persons with spinal cord injury (SCI), relating these patterns to demographic and injury-related characteristics, as well as to key medical and psychosocial outcomes.

Design

Retrospective cross-sectional.

Participants

Subjects with traumatic SCI (N=3041) with dates of injury between June 6, 1975, and June 23, 2002, who were interviewed between November 2000 and March 2003.

Setting

Sixteen Model Spinal Cord Injury Systems participating in this collaborative study during the 2000–2005 grant cycle.

Interventions

Not applicable.

Main outcome measures

Alcohol consumption, substance use, CAGE questionnaire, Satisfaction With Life Scale, Craig Handicap Assessment Reporting Technique, and pain. Data were analyzed using chi-square tests, analysis of variance, analysis of covariance, and logistic regression models.

Results

Fourteen percent of the subjects were classified as likely to have an alcohol abuse issue, based on the CAGE, and 11% reported using illegal drugs or prescription medications for nonmedical reasons. Demographic and injury characteristics were associated with alcohol consumption patterns, abuse, and substance use. At-risk drinkers and substance users tended to be younger, single, male, and less educated. Those who were CAGE positive and substance users reported more pain and lower satisfaction with life. Persons who drank without indication of problem drinking had superior occupation outcomes. Pressure ulcers were associated with substance use.

Conclusions

Alcohol abuse and substance use were related to a number of adverse outcomes. The specific role of drinking with increased work activity deserves further exploration.

Keywords:  Alcohol abuse, Alcohol consumption, Rehabilitation, Spinal cord injuries, Substance abuse

ALCOHOL AND SUBSTANCE ABUSE have been identified as risk factors for spinal cord injury (SCI) and the link between use of illicit substances and occurrence of traumatic injury has often been reported in the literature.1, 2, 3 Alcohol abuse has been cited as an obstacle to rehabilitation, correlated with longer lengths of stays, poor rehabilitation outcomes, decreased life satisfaction, depression, anger, anxiety and increased risk for seizures, pressure ulcers, urinary tract infections, and reinjury.4, 5, 6, 7, 8, 9 Furthermore, alcohol abuse has been associated with poor ratings of health and impaired self-care activities up to 18 months after the onset of disability, and a preinjury history of substance abuse has been significantly related to an increased rate of pressure ulcers 30 months after the onset of SCI.4, 8, 10

A number of research articles11, 12 have documented the prevalence of substance abuse among persons with disabilities, including SCI. Other studies11, 13 identified the devastating effects of substance abuse and the challenges of treating persons with disabilities who also have substance and alcohol abuse histories. Although several studies include both alcohol and drug use in a single category of “substance abuse,” others speak of these as separate issues. This latter approach is used in our study.

Estimates of alcohol and substance use at the time of injury are reported to be anywhere from 17% to 62%.7, 11, 14 These estimates are even higher for persons with SCI living in the community.12 Heinemann et al11 reported that alcohol use prevalence was 90% for 18- to 25-year-olds and 60% for those older than 25 years. Heinemann also conducted a study to examine exposure to and recent use of 10 types of substances, including alcohol and marijuana, in a sample of 103 recently injured persons with SCI. Heinemann compared study data to the general population and concluded that the SCI sample reported greater use of all substances except inhalants. Marijuana was used by 51% of younger subjects and by 30% of older ones. Other researchers12, 14 investigating illicit drug use at the time of injury have reported that more than 30% of persons admitted to hospitals with a diagnosis of SCI tested positive for illicit drugs.

Misuse of prescription drugs, alcohol, marijuana, and cocaine among persons with SCI was the focus of a study in 1991.2 The prevalence of marijuana use was found to be 26% for this sample of persons with SCI. Substance use has also been related to SCI etiology. Compared with those with nonviolent etiologies, persons with violence-related SCI (gunshot wound, assault) were significantly more likely to have positive admission toxicology screens, drug screens, and intoxication screens.7

Although most studies evaluated the effect of substance and alcohol use at the time of injury, some have focused on preinjury patterns, and a few have examined postinjury patterns. Persons with histories of excessive alcohol use before injury have higher mortality rates, are more likely to sustain serious brain abnormalities, and have a greater chance for physical deterioration after injury.15, 16, 17 Likewise, preinjury alcohol consumption is among the factors statistically predictive of suicide after SCI.18 In most cases, alcohol abuse post-SCI is a continuation of an earlier pattern of problem drinking.19, 20 In this respect, preinjury substance abuse is an important predictor of postinjury substance abuse.

Patterns of alcohol use before and after SCI appear to be strongly related. Elliot et al21 found that up to two thirds of SCI patients resume drinking within 18 months of injury. Other authors have described the rate of alcohol use among persons with SCI as being lower than the general population at 18 months postinjury; yet among SCI heavy drinkers—those with preinjury history of alcohol abuse—the rate was as high as 49%.18, 22 High rates of heavy drinking episodes (5+ drinks per occasion) were noted by Krause et al6 when studying the American Indian population with SCI. The average number of drinks per drinking occasion was higher for American Indian men with SCI than for those without (4.9 drinks vs with 4.0, P<.001), based on nationwide data from the Behavioral Risk Factor Surveillance System.

Only a few studies have addressed alcohol abuse problems of persons with SCI immediately after their injuries. Using the “window of opportunity” concept, Bombardier and Rimmele10 found a significant association between “readiness to change” and drinking habits in SCI. Those with a positive history of alcohol abuse and higher daily consumption patterns were more likely to report readiness to change.

Studies describing patterns of alcohol and substance use utilizing large national samples of persons with SCI may provide further information about drinking behaviors before and after injury and specific recommendations on treatment guidelines. Two objectives guided the present investigation conducted with a large sample of participants with SCI from 16 Model Spinal Cord Injury Systems (MSCIS) across the United States: (1) to identify persons with SCI at risk by describing their patterns of alcohol and substance use and abuse in relation to their individual characteristics, and (2) to determine how these patterns relate to medical and psychosocial outcomes after SCI. Such information can assist rehabilitation professionals in addressing the special and unique alcohol- and substance-related needs of those with SCI.

Based on our clinical knowledge and the literature reviewed, the 2 hypotheses were tested: (1) differences in patterns of alcohol consumption, abuse, and substance use are related to demographic characteristics, such as age, gender, marital status, education, etiology, time postinjury, and injury severity; and (2) alcohol consumption, abuse, and substance use after SCI are associated with key medical and psychosocial outcomes (ie, the experience of pain, pressure ulcers, satisfaction with life, occupational activity). When controlling for key moderating factors, those participants with SCI classified as at-risk drinkers, those having alcohol abuse problems, and those who report being substance users will have the worst outcomes after injury.

Back to Article Outline

Methods 

Our cross-sectional study uses research data collected during subjects’ annual follow-up interviews at 1 of the 16 current (2000–2005) MSCIS. All data are electronically submitted and stored at the National Spinal Cord Injury Statistical Center (NSCISC) in Birmingham, AL.

Participants 

Patients were eligible for study inclusion if they had a new traumatic SCI at the first admission to the MSCIS; had dates of injury from 1975 to 2002; were ages 18 and older; and received their inpatient acute medical and surgical care, inpatient rehabilitation, an organized program of outpatient rehabilitation, and/or systematic day-hospitalization rehabilitation at one of the MSCIS. Those who completed an organized rehabilitation program before Systems admission were ineligible, as were persons who were not US citizens or who did not have a clear degree of neurologic impairment after traumatic injury. All subjects included were interviewed at least once after their discharge from initial inpatient rehabilitation during the current grant cycle. Although most subjects have only had 1 follow-up interview during the current grant cycle, some subjects were interviewed more than once. To avoid bias from nonindependent observations, this sampling frame was limited to 1 case per subject. Thus, only the first observation per subject was included in the data set.

Procedures 

Each MSCIS center uses common procedures for data collection, as described in the MSCIS syllabus. Each obtained approval from its own institutional review board, thus following the ethical standards for responsible conduct of research, ensuring subjects’ safety and data confidentiality in accordance with the new regulations of the Health Insurance Portability and Accountability Act.

For our study, alcohol abuse was defined, according to the Diagnostic and Statistical Manual of Mental Disorders, 4th Edition (DSM-IV), of the American Psychiatric Association, as recurrent alcohol use despite significant adverse consequences of drinking, such as problems with work, law, health, or family life.17 The term “substance use” was chosen in lieu of “abuse,” to be consistent with other SCI studies’ methodologies and to simply indicate the improper use of illegal drugs, substances, and/or prescription medications with potential harmful consequences to the individual and society. The term “alcohol consumption” refers to the quantity and frequency of alcohol use. Binge drinking refers to drinking 5 or more drinks on an occasion at least once during a month. To classify alcohol consumption, researchers combined subjects’ responses to 4 drinking questions using the 2000 standards recommended by the National Institute on Alcohol Abuse and Alcoholism23 (NIAAA). NIAAA recommends that people aged 65 years and older limit their consumption to 1 drink a day. The NIAAA definitions for alcohol use are described in table 1. The CAGE was selected as the measure of alcohol abuse. Participants were also asked about their use of substances (yes, no) and, more specifically, the type of substances used.

Table 1. NIAAA Definitions for Classification of Alcohol Use23

Because the level of physical impairment associated with SCI may affect outcome findings, participants’ neurologic injuries were classified using the International Standards for Neurological Classification of Spinal Cord Injury, developed by the American Spinal Injury Association24 (ASIA). Thus, tetraplegia refers to impairment or loss of motor and/or sensory function in the cervical segments of the spinal cord, whereas paraplegia refers to impairment or loss of motor and/or sensory function in the thoracic, lumbar, or sacral segments of the spinal cord. The ASIA Impairment Scale is briefly summarized as follows: ASIA grade A (complete injury with no sensory or motor function preserved in the sacral segments of S4-5); ASIA grade B (incomplete injury with sensory but no motor function preserved below the neurologic level; includes the sacral segments above); ASIA grade C (incomplete injury with motor function preserved below the neurologic level; up to half of key muscles below the neurologic level preserved); ASIA grade D (incomplete injury with motor function preserved below the neurologic level; more than half of key muscles preserved); and ASIA grade E (normal sensory and motor function). Participants classified as ASIA grade E were excluded from the study.

Measures 

Standardized alcohol and substance use questions were part of the regular annual follow-up of MSCIS participants. Alcohol use was assessed in 2 ways: (1) 4 items regarding frequency and quantity of drinking and (2) the CAGE questionnaire. The first question in the survey protocol concerning alcohol use and abuse was: “Do you drink any alcoholic beverages, such as beer, wine, wine coolers, or liquor?” The response choices to this question were: “No, never ever drank alcohol” and “Yes, currently drink alcohol or did drink alcohol in the past.” A skip pattern is included in the survey, with subjects who responded negatively to this first question not being asked any further alcohol-related questions. For those who responded yes, 3 additional questions were asked: (1) “During the past month, how many days per week did you drink any alcoholic beverages, on average?” (2) “On the days you drank, about how many drinks did you drink, on average?” and (3) “Considering all types of alcoholic beverages, how many times during the past month did you have 5 or more drinks on an occasion?” With regard to substance use, participants were asked about their use of up to 6 substances or prescribed medications used for nonmedical purposes. More specifically, they were asked about their use of specific drugs, such as crack/cocaine, pot/marijuana, LSD/hallucinogens, heroin/opiates, and speed/stimulants, as well as medications prescribed to them or to someone else or any other type of drug that is unknown.

Participants also completed the CAGE (an acronym for cut down, annoyed, guilty, and eye opener), as a way to assess severity of drinking problem and dependence. The CAGE requires yes or no responses to 4 questions: (1) Have you ever felt you should cut down on your drinking? (2) Have people annoyed you by criticizing your drinking? (3) Have you ever felt bad or guilty about your drinking? and (4) Have you ever taken a drink first thing in the morning (eye opener) to steady up your nerves or get rid of a hangover? These 4 questions were placed in order, to permit the use of the mnemonic CAGE. CAGE scores vary from 0 to 4 points. Participants who answered any of the questions positively were classified as CAGE positive.25 The CAGE has been recognized as among the most efficient and effective screening devices for alcohol abuse.25 Validation interviews conducted by Ewing,25 using formal diagnostic criteria for alcohol abuse, yielded a 97% true positive rate for this diagnosis. This rate compared favorably with the rate of 81% found in a similar validation study completed by another author.26 A positive response to the CAGE questionnaire, although it may not completely determine the diagnosis of alcohol abuse on its own, can suggest the high likelihood of the presence of this diagnosis. Using a DSM-IV diagnosis of alcohol abuse as a criterion, CAGE positive scores have a sensitivity of 77% and a specificity of 99%.27

Two medical outcomes were selected on the basis of their reported associations with alcohol consumption and abuse and substance use among persons with SCI. Information regarding pain and pressure ulcers was derived from the NSCISC database. Pressure ulcer data were gathered by visual inspection in the clinic at the time of the participant annual evaluation.

Pain data were collected by interviewing subjects and asking them about pain existence, severity, and the degree of interference with normal activities during the past month. To be able to conduct a singular, comprehensive assessment of pain, we combined pain data from these questions into a singular score. This is the approach also taken by the Medical Outcomes Study 36-Item Short-Form Health Survey (SF-36). The SF-3628 contains 8 distinct scales devised to assess mental and physical health and well-being among persons across the population spectrum, from the general population to those with mental and physical conditions of varying degrees of severity. The interference question was taken verbatim from the pain scale of the SF-36. The pain intensity question (asked of MSCIS participants) mirrors the SF-36 pain intensity question; however, it uses an 11-point scale rather than a 6-point one, as the SF-36 does. For both the MSCIS and the SF-36 pain intensity question, the base response is no pain. To transform subjects’ affirmative responses, these answers were grouped by rating levels accordingly: ratings of 1 and 2 were classified as 1 (very mild pain); 3 and 4 as 2 (mild pain), etc. After these rating transformations, scale scores for the SF-36 pain scale were calculated using the standard SF-36 algorithm.28

Data on pressure ulcers were coded with respect to whether participants had any ulcers and, if so, whether they had surgical procedures to treat the ulcers (ie, flap procedures). From these questions, subjects’ responses were coded as (1) not having an ulcer, (2) having an ulcer without procedure, and (3) having an ulcer that required a surgical procedure.

The 2 psychosocial outcomes selected reflect the participants’ subjective opinion of their quality of life (QOL) and their degree of participation in society, as measured by engagement in occupation and productive activities. More specifically, the 2 psychosocial outcomes selected were satisfaction with life and occupational activity. Both measures are currently part of the MSCIS database. The Satisfaction With Life Scale (SWLS) is a 5-item, self-report measure that was used to measure global life satisfaction.29 These statements are rated on a Likert scale from 1 (strongly disagree) to 7 (strongly agree). Administration is brief and easy. Reliability of the scale has repeatedly exceeded expectations.30 Test-retest reliability has been moderately high (average, .82). Most important, the SWLS has been normed on a number of populations, including people with SCI. When used with MSCIS participants, SWLS scores were found to be directly associated with access to the environment at both 1 year and 2 years postinjury.30

The Occupation scale of the Craig Handicap Assessment Reporting Technique (CHART) was used to assess SCI subjects’ participation in occupational activities. The CHART Short Form includes 19 items, 6 dimensions, and a total score and was developed on the basis of the World Health Organization’s conceptual scheme to assess the degree to which physical impairments may result in limitations for the SCI person in relation to their participation in the environment.31 The CHART’s test-retest reliability coefficients have been reported to range from .80 to .95, whereas good discriminant validity was obtained when using the test with contrasting groups.31

The CHART’s 6 dimensions are physical independence, cognitive independence, mobility, occupation, social integration, and economic self-sufficiency. Participants were asked about how much personal assistance they received with these tasks to meet their physical needs, such as dressing or bathing (physical independence); how much assistance they received in and outside the home and with cognitive tasks, such as remembering, making decisions, attention, and concentration (cognitive independence); whether they could physically move within their environment (mobility); how they occupied their time in socially beneficial activities, such as work, school, housekeeping, or parenting (occupation); and how they interacted with others (social integration). They were also asked about their household income (economic self-sufficiency).

Although scores from the CHART Occupation scale were used as an outcome measure for our study, CHART Mobility and Physical Independence scores were used as moderating variables, when determining the association of alcohol and substance patterns with outcomes other than occupation. The rationale for choosing these 2 subscales was to best account for individual differences in physical functioning in relation to the environment, thus going beyond the level of neurologic impairment assessed by ASIA levels.

Statistical analyses 

Descriptive statistics, including means, standard deviations (SDs), and percentages, were computed for all relevant variables. Reported response rates varied slightly according to the variables selected. Participants were grouped according to a classification scheme described earlier for meaningful comparisons. The quality and quantity of data available specified the type of analyses performed. Most analyses used alcohol-related data due to possible underreporting and the small sample sizes associated with the responses to the substance use question. To test the first hypothesis, differential statistics were used including chi-square tests and analysis of variance. The Bonferroni adjustment for multiple tests was used to determine significant comparisons.

The second hypothesis was tested using multivariate models to control for moderating factors. A set of parallel analysis of covariance (ANCOVA) models were specified in order to test the relationship of alcohol consumption (NIAAA), alcohol abuse (CAGE), and substance use, to study outcomes of global satisfaction with life (measured by the SWLS), pain-related QOL (measured by the SF-36 pain scale), and occupational activity (measured by the CHART Occupation scale). Logistic models were specified to test these relationships for prevalence of pressure ulcers due to the dichotomous nature of this outcome. Although models with greater predictive power could likely have been specified by including additional predictors, this was not done because the goal of these multivariate analyses was to explain the unique contributing role of the alcohol and substance use rather than to minimize the variance in the study’s outcomes. The following moderators were included in the multivariate specifications: severity of neurologic impairment on discharge from initial rehabilitation, gender, race, highest level of education completed, etiology, current age, and time since injury. In addition, the CHART Physical Independence and Mobility scale scores were included, as were subjects’ occupation status. The latter were excluded for analyses of the CHART Occupation scale.

Back to Article Outline

Results 

Demographic and injury-related characteristics of the total sample 

The total sample included 3041 eligible participants with SCI. Table 2 illustrates their characteristics. Participants, on average, had been injured 9.7 earlier and most were classified as having complete neurologic injuries, with 41.2% as paraplegia ABC and 38.1% as tetraplegia ABC.

Table 2. Demographic Characteristics for the SCI Sample From the MSCIS Program (N=3041)∗

Abbreviations: GED, General Education Development diploma; MVC, motor vehicle collision.

Alcohol consumption and abuse and substance use after SCI 

Of the entire sample, 59.8% (1820) of participants with SCI responded positively to the question of drinking any alcoholic beverages; 40.2% (1221) reported not drinking. In terms of average frequency of alcohol consumption, 59.5% of the entire sample said that they either did not drink or drank less than 1 day a week; 14.7% reported drinking once per week; 10.7% reported drinking on 2 days a week; and 12.3% reported drinking on more than 2 days a week. An additional 2.8% reported that they drank but that they did not know how frequently. With respect to number of drinks consumed on the days when they drank, 13.9% of drinkers reported drinking an average of 1 drink; 13.7%, 2 drinks; 7%, 3 drinks; 3.6%, 4 drinks; and 7.2%, 5 or more drinks a day. Among drinkers, the average number of reported drinks consumed per week reported was 5.7; 15% reported drinking 5 or more drinks on at least 1 occasion per month. A drink was defined as 1 can or bottle of beer, 1 glass of wine or wine cooler, 1 cocktail, or 1 shot of liquor.

Using the CAGE, 14.2% (429) of all subjects were classified as CAGE positive (answering affirmatively to at least 1 question), which suggests the presence of alcohol abuse. Another 45.3% (1369) of the respondents were classified as CAGE negative, which indicates that they drink but responded negatively to all 4 CAGE questions. Among those subjects reporting any current or prior drinking, 23.9% were CAGE positive. Among CAGE positive subjects, the mean score ± SD was 1.78±0.91. The CAGE score distribution for these subjects was as follows: 50% scored 1; 27% scored 2; 18% scored 3, and 5% (21 persons) scored 4, answering yes to all CAGE questions. Table 3 illustrates the percentages of affirmative responses by item.

Table 3. Distribution of Affirmative Responses by CAGE Item Among CAGE-Positive Respondents

Most of the sample (89%) indicated no use of illegal drugs or prescribed medications for nonmedical reasons during the past year. Of the remaining, 10.2% reported using 1 drug, and fewer than 1% (.8%) reported using more than 1 drug. Of those reporting substance use, 73.3% (235) stated using marijuana; 14.4% (42) used crack or cocaine; and 12.3% (29) used a number of other drugs, including heroin, speed, and medications.

Characteristics of SCI drinkers and abstainers 

Significant differences in terms of both alcohol consumption, assessed using NIAAA criteria for safe and moderate drinking, and CAGE scores, assessing potential alcohol abuse, were found in terms of age (at injury and current) (P<.001). Those who drank more heavily were significantly younger than were moderate drinkers and persons who abstained. Younger participants with SCI reported drinking more than older ones (F=31.9, P<.001). At-risk drinkers (those engaged in binge drinking or drinking >14 drinks a week) were significantly younger (37.5y) than moderates and abstainers, who did not differ in age from each other. Participants classified as CAGE positive were also younger (38.8y) than CAGE negatives (40.9y) and abstainers (43.0y). Equally significant were consumption differences reported by time postinjury, with abstainers (8.1y) reporting less time postinjury than others (F=44.3, P<.001). Table 4 describes these data.

Table 4. Current Age, Time Postinjury, and Alcohol Consumption and Abuse

Abbreviations: Neg, negative; Pos, positive.

Alcohol consumption and abuse patterns varied when participants with SCI were classified by neurologic groupings (table 5). Abstaining was most likely to occur among those classified as having tetraplegia ABC (44.5%) and least likely among those with ASIA grade D injuries (35.5%). Those ASIA grade D or sensory and motor incomplete impairments most frequently reported abusive drinking (19.4% vs 17.0% among the entire sample). No other significant differences in alcohol consumption or abuse were observed across levels of neurologic classification.

Table 5. Alcohol Consumption and Abuse by Neurologic Classification

Significant differences were also noted in terms of other demographic characteristics, including gender, etiology, race, marital status, education, and occupation (table 6). All statistical comparisons were significant at P less than .001. Both men and single persons were disproportionately more likely to be at-risk drinkers, as defined by the NIAAA. Single persons were also more likely to be CAGE positive. The relationship between level of education and the CAGE classification system was a curious one. Subjects with a bachelor’s or postgraduate degree were less likely to report abstaining than were those with lower levels of education, with 26.9% of them saying that they abstained, compared with 45.6% of other subjects. Subjects with higher levels of education were also the most likely to be CAGE negative—59.5% of them classified in this manner. No association was present between level of education and prevalence of CAGE positive reports. Reports of abstaining were most prevalent among African Americans (57.6%), and CAGE negative classifications were most prevalent among whites (50%).

Table 6. Demographic Characteristics Associated With Alcohol Consumption and Abuse

NOTE. Values are n (%).

Motor vehicle collisions (MVCs) were the most frequent cause of injury across all groups (50.7%). Participants with sports-related injuries were more likely to be at-risk drinkers than were others (27.1% vs 15.8%) and they were less likely to be abstainers (25.1% vs 43.2%). They were also likely to be CAGE positive (19.6%), which indicates potential alcohol abuse problems. With respect to occupation status at time of interview, 49.1% of respondents were unemployed and 29.9% were working. Abstinence rates were highest among retirees (49.0%) and homemakers (47.5%) and lowest among those employed (26.0%) and students (39.4%).

Characteristics of substance users and nonusers 

Substance users were significantly younger than were nonusers at the time of assessment (42.2y vs 36.0y) (P<.001). Users and nonusers did not differ in terms of time postinjury. They also did not differ in terms of neurologic classification between substance users and nonusers. Table 7 gives these data.

Table 7. Demographic and Injury Characteristics and Substance Use

NOTE. N=3017. Variables with fewer responses are noted accordingly.

Substance use was associated with gender, marital status, and education level. Men reported significantly higher rates of substance use than women (12.5% vs 5.8%). Similarly, the rate was much higher among persons who were single (15.8%) than among those who were married (5.3%). Persons with lower levels of education had higher prevalence rates of substance use; among those whose highest level of education completed was high school or less, the rate was 12.8%, compared with 8.1% among those who completed some postsecondary educational program. No differences were observed along racial and ethnic lines.

Etiology was associated with substance use. As was the case with alcohol consumption, those injured in sports were significantly more likely to use substances than was the sample as a whole (19.7% vs 11.0%). Substance use was most prevalent among students (13.7%) and unemployed persons (13.6%) and least prevalent among homemakers (3.4%) and retirees (5.7%).

Differences in medical and psychosocial outcomes 

Findings with respect to the second hypothesis were mixed. Abstainers reported less pain than did moderate and at-risk drinkers (P<.05), but there were no differences in pain between the 2 drinking groups (moderate, at-risk). In contrast, participants with SCI classified as CAGE positive, suggesting alcohol abuse, reported significantly worse pain outcomes than did both abstainers and CAGE negatives (P<.001). Such findings are indeed supportive of our second hypothesis. Similarly, those reporting use of illegal substances also scored lower on pain items (indicating worse pain) than did those who reported not using substances (P<.001) (table 8). Three multivariate models were specified to explain pain as an outcome. Each model focused on a different predictor: NIAAA alcohol consumption, CAGE score, and substance use. All models contained the following additional factors to control for their effects on pain: gender, education, neurologic level, race, etiology, time postinjury, CHART Occupation, and CHART Mobility. Pain was best predicted by the alcohol and substance use variables in conjunction with factors such as gender, neurologic level, mobility, and occupation, with significance levels varying from P less than .05 to .001.

Table 8. Outcomes by Alcohol Consumption and Abuse and Substance Use∗

The prevalence of pressure ulcers during participants follow-up annual evaluations was 32.2% for the entire sample. Of those, 27.6% had an ulcer without surgical procedure or closure and only 4.6% had a skin closure procedure. Among those reporting that they drank alcoholic beverages, 29% were classified as having pressure ulcers, compared with 48% of those who reported that they abstained (P<.005). These differences, however, were not significant when controlling for moderating factors, as explained below. Reported substance use was associated with prevalence of pressure ulcers, with those reporting drug use having higher prevalence rates: 38% of these persons had pressure ulcers, compared with 32% of those who did not report engaging in substance use (P<.05).

Globally, the 3 logistic models that were developed to explain the presence of ulcers were quite similar: their chi-square tests were all highly significant (P<.000), and they all predicted 71% of the subjects correctly in terms of prevalence of ulcers during the past year, with a sensitivity of 60% and a specificity of 73%. There was 1 major difference among the models, however. Neither risky alcohol consumption nor alcohol abuse was associated with the presence of ulcers while substance use was; participants with SCI reporting that they used drugs were more likely to have pressure ulcers (P<.001) (table 9).

Table 9. Logistic Regressions Models for Pressure Ulcers Based on Alcohol Consumption, Abuse, and Substance Use

Abbreviation: CI, confidence interval.

Psychosocial findings based on alcohol consumption differed substantially from those based on abuse. There were no significant differences in self-reported satisfaction with life among participants with SCI in any of the 3 groups based on frequency and amount of alcohol consumption: abstainers, moderate, and at-risk drinkers. In contrast, significant differences were found when comparing CAGE positives with negatives and abstainers (P<.001) (table 8). Participants who were CAGE positive reported the lowest satisfaction with life (score, 19.7). These findings were supportive of our second hypothesis. It appears from our data that using a diagnostic criterion of alcohol abuse, such as CAGE, instead of noting the frequency and amount of drinking, was a better and more sensitive way to determine satisfaction with life differences among participants with SCI. As noted in table 3, among the participants who drank, most (79%) of shared the common perception or belief that they needed to curb their drinking.

Consistent with CAGE data, those reporting use of substances also reported lower satisfaction with life (P<.001). Significant factors influencing satisfaction with life, in addition to alcohol abuse and substance use, were age, gender, occupation, etiology, mobility, and time since injury. Neurologic level of injury was not significant in determining satisfaction with life differences in relation to substance use.

Our results with respect to the second hypothesis were unclear regarding occupation outcomes. When considering alcohol consumption, abstainers reported lower occupation scores than did both moderate and at-risk drinkers (P<.001) (table 8). In this case, those who reported drinking also appeared to be more involved with work-related activities than did those who did not drink. In contrast, when investigating alcohol abuse instead of consumption, there were no differences between abstainers or potential abusers (CAGE positive) in terms of their scores on the CHART Occupation scale. Both had lower scores than those participants classified as CAGE negatives. CAGE negatives (suggesting absence of alcohol abuse) reported the highest scores for the CHART Occupation scale. There were no differences in CHART Occupation scale scores among those who reported using and not using substances. Based on the multivariate models, differences in occupation were significantly related to CAGE scores and all moderating factors with the exception of etiology.

Back to Article Outline

Discussion 

Alcohol consumption, abuse, and substance use problems deserve special attention among professionals treating persons with SCI. The numbers of persons with SCI with these problems are higher than in the general population, as shown by the studies reviewed and as confirmed by our findings. Statistics for the general population released by NIAAA indicate that 7.41% of persons aged 18 years and older meet standard diagnostic criteria for alcohol abuse and dependence.23 In contrast, our findings showed that 17% of participants with SCI engaged in at-risk drinking behaviors based on the NIAAA criteria. Similarly, in our study, the rate found for alcohol abuse based on frequency of CAGE-positive responses is 14.2%.

In our study, 11% of participants reported using illegal substances, mainly marijuana. Fewer than 1% reported using other illegal drugs. These figures appear to be low compared with data from other SCI studies. Young et al12 reported a 16% rate of marijuana use in persons with SCI living in the community, whereas Heinemann et al2 found this rate to be 42%. It is possible that circumstances surrounding interviews for MSCIS participants, including the type, order, and format of questions, may have inhibited subject responses, thereby accounting for some underreporting. Subjects may have been apprehensive about disclosing such sensitive and private information to their health care providers for fear of being denied further services or benefits. Although there are numerous legitimate reasons for variance in prevalence rates across studies, the large discrepancy between the findings of Young and of Heinemann brings into question possible problems with the reliability of data regarding substance use in general. Because of potential underreporting of these problems, findings concerning substance use should be interpreted with some caution.

Most of our findings supported those described by NIAAA for the general population.23 NIAAA reports that alcohol abuse is higher among men than among women and is highest in the youngest age cohort (18–29y); young non-African-American men are almost twice as likely as young African-American men to have alcohol use disorder; and a higher percentage of African Americans than whites abstain from using alcohol.8, 13 In our study, participants with SCI classified as risk drinkers were also likely to be younger (age at injury, 26.7y), white, male, single, and less educated. Similar to the NIAAA report, among African-American participants with SCI, a higher proportion reported abstaining than did other drinking groups (table 6). It is interesting to note that not only were SCI subjects classified as at-risk and CAGE positive younger than abstainers, but they were also significantly younger at the times of their injuries (table 4).

Alcohol is among the leading causes of death among injuries caused by MVCs. In our sample, MVCs accounted for the highest percentage of injuries across both nondrinkers and all types of drinkers. Among at-risk drinkers, MVCs accounted for 51.2% of the SCI cases. These results are supportive of those from other studies that report MVCs to be the main cause of SCI among at-risk drinkers. Among both at-risk drinkers and substance users, a disproportionately high number acquired their SCI in sports-related events. These findings suggest that in addition to driving safety programs, it is important that education programs be developed that can be used in both schools and work sites, focusing on prevention and safety for younger persons interested in engaging in sports-related activities.

Researchers investigating traumatic brain injuries have found no relationship between blood alcohol levels and neurologic outcomes.1 Although our data did not focus on blood alcohol level at the time of injury, persons with ASIA grade D impairments ratings were more likely to be classified as at-risk drinkers (23.8%) or to report alcohol abuse (24%) than were members of other neurologic groups. In contrast, substance use was not associated with neurologic level of injury, which suggests that the use of substances by persons with SCI is independent of the degree of physical limitations.

Worse medical outcomes, such as pain and pressure ulcers, were linked to patterns of alcohol consumption, abuse, and substance use. A bivariate assessment of the data suggested that pressure ulcers were not associated with intensity of alcohol use, alcohol abuse, or drug use. Similar findings are reported in the literature by Heinemann and Hawkins,5 which indicates that people with SCI with preinjury alcohol problems who abstained from alcohol after injury were at increased risk for developing pressure ulcers. When we controlled for moderating factors, such as neurologic level of injury, gender, race, and age, those with substance use problems were at a higher risk for pressure ulcers. Our pressure ulcer findings are also indicative of the complexity of factors associated with their development, only one of which may be the use of illegal substances and/or inappropriate use of prescription medicine. Not surprisingly, pressure ulcers occurred more often among substance users with complete injuries. This is likely because these people lack sensation and therefore do inadequate pressure relief. It is possible that a combination of other factors, such as more active lifestyle among those with paraplegia and high incidence of other medical complications like spasticity, may also explain some of these findings.

Greater attention must be paid to pain management in SCI, because unresolved pain symptoms may lead to a higher incidence of alcohol abuse and substance use, and this may result in a downward cycle. In our study, participants classified as having potential alcohol abuse problems (CAGE positive) also experienced the worst pain. These findings confirm the importance of having physicians, psychologists, and/or social workers discuss their substance and alcohol dependency issues with patients, during admission and initial evaluation assessments, when SCI subjects first enter a designated system of care. Unidentified dependency problems are likely to be exacerbated when prescribed pain treatments are implemented, which may often prove to be ineffective, during the course of rehabilitation. Follow-up assessments after discharge from the rehabilitation program are equally important to prevent the recurrence of dependency problems. The outpatient team needs to be well informed about these issues, in order to coordinate patient care that promotes healthy lifestyle behaviors in the community.

With respect to satisfaction with life, our results varied. Although there were no differences when comparing participants with SCI by the NIAAA patterns of alcohol consumption, those classified as CAGE positive and those who reported using substances reported lower satisfaction with life. This suggests that persons with SCI may use alcohol and substances as a preferred escape and avoidance coping mechanisms to address what they perceive as life’s stressful situations after injury. Unable to cope more effectively with these problems, they are likely to experience a sense of loss of control over their lives and decreased life satisfaction, as well as to be at a greater risk for depression. Although beyond the purview of our study, similar relationships between alcoholism and depression have been noted elsewhere.32 Beyond these behavioral issues, from a physiologic perspective as central nervous system depressants, both alcohol and substances may further negatively impact one’s perception and self-report of his/her satisfaction with life.

The complexity of these issues (ie, behavioral, physiologic, environmental) is further illustrated by our mixed findings about occupational outcomes. The direction of these findings seems dependent on a specific classification criterion. Those persons who abstained from drinking had lower scores on the CHART Occupation scale than did persons who drank alcohol, as classified by the NIAAA system. In contrast, persons who were CAGE negative (suggesting drinking in moderation) had significantly higher scores than did those who abstained. No differences between substance users and nonusers were observed in terms of occupational outcomes. Several factors may have accounted for these unexpected findings, such as the high number of unemployed persons in our sample and the relatively limited sensitivity of the CHART Occupation scale to appropriately detect group differences. Occupation outcomes included the number of hours a week one is involved in a paid job, school and study, homemaking, home maintenance, and recreation. The quality of this involvement is not accounted for by the current version of the CHART.

As a measure, the CAGE appears to reliably identify participants with SCI with alcohol abuse problems. Those participants identified as CAGE positive reported the lowest satisfaction with life. Thus, the CAGE as a brief screening measure is a reliable and valid indicator of alcohol abuse problems among persons with SCI receiving their care at an MSCIS program.

It is unclear why the NIAAA scheme of assessing risk was limited in its ability to screen for adverse outcomes. Its purpose differs from that of the CAGE in that it is not intended to be used as a diagnostic screen, but rather, to indicate that above its intermediate level, defined as “moderate drinking” in the general population, persons are at risk for alcohol-related problems. The precision of this threshold is unclear in the SCI population. Persons with SCI are not familiar with the NIAAA guidelines for moderate and at-risk drinking (table 1), nor do they often know the implications of drinking at elevated levels. What levels of drinking are considered “at-risk” in the SCI population is currently unknown. The results of our study do not suggest that the current NIAAA screen is inherently invalid or inappropriate to use in relation to persons with SCI, but rather that, at this time, insufficient information is available to determine what thresholds are appropriate and for what outcomes.

Limitations of study 

Several caveats concerning the representativeness of the sample are warranted in attempting to generalize our findings to the national SCI population. First, the sample may not fully reflect the characteristics of the population-sampling frame, because the MSCIS database is not population-based.33 Therefore, our data are limited in their ability to most appropriately determine important epidemiologic information, such as the incidence and prevalence of alcohol and substance use. The data presented, however, illustrate trends in proportions, which ultimately can affect incidence. Nonetheless, the MSCIS national database provides the largest existing pool of information on alcohol and substance use in persons with SCI.

A possible problem for any cohort study is incomplete data reflected by missing data in certain variables and/or data underreporting, thus limiting the validity of conclusions. Although missing data did not affect the study’s statistical power, it may have indirectly influenced our findings by including differential sample bias related to underreporting problems. For these reasons, caution must be taken when comparing our findings with past and future studies of substance use in persons with SCI and in extrapolating the findings to all persons with SCI. Similarly, because alcohol and substance use data are not collected from persons under the age of 18 by the MSCIS, it is important that conclusions not be generalized to include this younger subpopulation.

Back to Article Outline

Conclusions 

Extremely little is known about the additional risks attributable to alcohol use or abuse that occurs after SCI. There are several needs for future research. First, given the strong likelihood of underreporting of illicit drug use and the potential of underreporting of alcohol use, research is needed to identify the means to enhance the accuracy of this reporting. This may require more labor-intensive methods of collecting the data, such as separating collection of this type of data from the primary MSCIS data collection. Second, alternative types of substance use need to be considered, particularly use of prescription medications that may have psychotropic effects. Investigation of this type of usage pattern may produce more accurate reporting, because these medications have been prescribed and do not carry either the stigma or the legal ramifications of illicit drug use. It is also possible that the lower-than-expected reports of illicit drug use may reflect the easy and inexpensive access to prescription medications. Third and perhaps most important, as an MSCIS of care, we need to go beyond documenting patterns and consequences of misuse and focus on both prevention strategies implemented shortly after SCI onset and intervention studies for patients who have established maladaptive usage patterns in the community.

Longitudinal studies are needed that examine the differential impact of preinjury alcohol abuse, intoxication at the time of injury, and postinjury alcohol use on outcomes. The effects of alcohol on health maintenance behaviors, cognition, complications, and community integration also merit further investigation when addressing the needs of persons with SCI. It is only through continued multidisciplinary, collaborative, and longitudinal research that we will be able to help people with SCI to overcome the adverse consequences of alcohol and substance misuse.

Back to Article Outline

Acknowledgements 

We acknowledge the support of all 16 MSCIS funded by the National Institute on Disability and Rehabilitation Research, Office of Special Education and Rehabilitative Services, US Department of Education, and of the National Spinal Cord Injury Statistics Center at the University of Alabama, Birmingham, for their assistance and sponsorship of this work.

Back to Article Outline

References 

1. Bombardier CH, Temkin NR, Machamer J, Dikmen S. The natural history of drinking and alcohol-related problems after traumatic brain injury. Arch Phys Med Rehabil. 2003;84:185–191
   • View In Article
   • Abstract
   • Full-Text PDF (103 KB)
   • CrossRef
2. Heinemann AW, Doll MD, Kevin JA, Schnoll S, Yarkony G. Substance use and receipt of treatment by persons with long-term spinal cord injuries. Arch Phys Med Rehabil. 1991;72:482–487
   • View In Article
   • MEDLINE
3. Richards JS, Kewman DG, Pierce CA. In:  Frank RG,  Elliott TR editor. Handbook of rehabilitation psychology. Washington (DC): American Psychological Association; 2000;p. 11–26
   • View In Article
4. Hawkins DA, Heinemann AW. Substance abuse and medical complications following spinal cord injury. Rehabil Psychol. 1998;43:219–231
   • View In Article
   • CrossRef
5. Heinemann AW, Hawkins DA. Substance abuse and medical complications following spinal cord injury. Rehabil Psychol. 1995;40:125–140
   • View In Article
   • CrossRef
6. Krause JS, Coker J, Charlifue S, Whiteneck GG. Health behaviors among American Indians with spinal cord injury (comparison with data from 1996 Behavioral Risk Factor Surveillance System). Arch Phys Med Rehabil. 1999;80:1435–1440
   • View In Article
   • Abstract
   • Full-Text PDF (2308 KB)
   • CrossRef
7. McKinley WO, Kolakowsky SA, Kreutzer JS. Substance abuse violence and outcome after traumatic spinal cord injury. Am J Phys Med Rehabil. 1999;78:306–312
   • View In Article
   • MEDLINE
   • CrossRef
8. Tate DG. Alcohol use among spinal cord injured patients. Am J Phys Med Rehabil. 1993;72:192–195
   • View In Article
   • MEDLINE
   • CrossRef
9. Tate DG. The use of the CAGE questionnaire to assess alcohol abuse among spinal cord injury persons. J Rehabil. 1994;60:31–35
   • View In Article
10. Bombardier CH, Rimmele C. Alcohol use and readiness to change after spinal cord injury. Arch Phys Med Rehabil. 1998;79:1110–1115
    • View In Article
    • Abstract
    • Full-Text PDF (903 KB)
    • CrossRef
11. Heinemann AW, Mamott BD, Schnoll S. Substance use by persons with recent spinal cord injuries. Rehabil Psychol. 1990;35:217–238
    • View In Article
    • CrossRef
12. Young ME, Rintala DH, Rossi CD, Hart KA, Fuhrer MJ. Alcohol and marijuana use in a community-based sample of persons with spinal cord injury. Arch Phys Med Rehabil. 1995;76:525–532
    • View In Article
    • Abstract
    • Full-Text PDF (972 KB)
    • CrossRef
13. Fullerton DT, Harvey RF, Klein MH, Howell T. Psychiatric disorders in patients with spinal cord injuries. Arch Gen Psychiatry. 1981;38:1369–1371
    • View In Article
    • MEDLINE
14. O’Donnell JJ, Cooper JE, Gessner JE, Shehan I, Ashley J. Alcohol, drugs, and spinal cord injury. Alcohol Health Res World. 1981;82:27–29
    • View In Article
15. Ruff RM, Marshall LF, Klauber MR. Alcohol abuse and neurological outcome of the severely head injured. J Head Trauma Rehabil. 1990;51(3):21–31
    • View In Article
16. Dunlop TW, Uduarhelyi GB, Stedem AF, et al.  Comparison of patients with and without emotional/behavioral deterioration during the first year after traumatic brain injury. J Neuropsychiatry Clin Neurosci. 1991;3:150–156
    • View In Article
    • MEDLINE
17. In: Diagnostic and statistical manual of mental disorders (DSM-IV). 4th ed.. Washington (DC): American Psychiatric Association; 1994;p. 175–204
    • View In Article
18. Kolakowsky-Hayner SA, Gourley EV, Kreutzer JS, Marwitz JH, Cifu DX, McKinley WO. Pre-injury substance abuse among persons with brain injury and spinal cord injury. Brain Inj. 1999;13:571–581
    • View In Article
    • MEDLINE
    • CrossRef
19. Heinemann AW, Keen M, Donahue R. Alcohol use by persons with recent spinal cord injury. Arch Phys Med Rehabil. 1988;69:619–624
    • View In Article
    • MEDLINE
20. Charlifue SW, Gerhart K. Behavioral and demographic predictors of suicide after traumatic spinal cord injury. Arch Phys Med Rehabil. 1991;72:488–492
    • View In Article
    • MEDLINE
21. Elliot TR, Kurylo M, Chen C, Hicken B. Alcohol abuse history and adjustment following spinal cord injury. Rehabil Psychol. 2002;47:278–290
    • View In Article
    • CrossRef
22. Kolakowsky-Hayner SA, Gourley EV, Kreutzer JS, Marwitz JH, Meade MA, Cifu DX. Post-injury substance abuse among persons with brain injury and persons with spinal cord injury. Brain Inj. 2002;16:583–592
    • View In Article
    • MEDLINE
    • CrossRef
23. National Institute on Alcohol Abuse and Alcoholism . 10th Special report to the US Congress on alcohol and health: highlights from current research from the Secretary of Health and Human Services. Washington (DC): US Department of Health and Human Services, Public Health Service, National Institutes of Health; 2000; Publication no. 00-1583
    • View In Article
24. American Spinal Injury Association and International Medical Society of Paraplegia . In: International standards for neurological classification of spinal cord injury. 6th ed.. Chicago: ASIA; 2000;p. 1–24
    • View In Article
25. Ewing JA. Detecting alcoholism (the CAGE questionnaire). JAMA. 1984;252:1905–1907
    • View In Article
    • MEDLINE
26. Mayfield D, McLeod G, Hall P. The CAGE questionnaire (validation of a new alcoholism instrument). Am J Psychiatry. 1974;131:1121–1123
    • View In Article
    • MEDLINE
27. Hearne R, Connolly A, Sheehan J. Alcohol abuse (prevalence and detection in a general hospital). J R Soc Med. 2002;95:84–87
    • View In Article
    • MEDLINE
    • CrossRef
28. Ware JE, Snow KK, Kosinski M, Gandek B. SF-36 health survey (manual and interpretation guide). Boston: The Health Institute, New England Medical Center; 1993;
    • View In Article
29. Diener E, Emmons RA, Larson RJ, Griffin S. The satisfaction with life scale. J Pers Assess. 1985;49:71–74
    • View In Article
    • MEDLINE
    • CrossRef
30. Richards JS, Bombardier CH, Tate DG, et al.  Access to the environment and life satisfaction after spinal cord injury. Arch Phys Med Rehabil. 1999;80:1501–1506
    • View In Article
    • Abstract
    • Full-Text PDF (758 KB)
    • CrossRef
31. Whiteneck G, Tate DG, Charlifue S. Predicting community reintegration after spinal cord injury from demographic and injury characteristics. Arch Phys Med Rehabil. 1999;80:1485–1491
    • View In Article
    • Abstract
    • Full-Text PDF (757 KB)
    • CrossRef
32. Elliot T, Frank RG. Depression following spinal cord injury. Arch Phys Med Rehabil. 1996;77:816–823
    • View In Article
    • Abstract
    • Full-Text PDF (1324 KB)
    • CrossRef
33. Stover SL, DeVivo MJ, Go BK. History, implementation and current status of the National Spinal Cord Injury Database. Arch Phys Med Rehabil. 1999;80:1365–1371
    • View In Article
    • Abstract
    • Full-Text PDF (930 KB)
    • CrossRef

• 1 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.