Prevalence of Drinking within Low-Risk Guidelines during the First Two Years Following Inpatient Rehabilitation for Moderate or Severe Traumatic Brain Injury

Abstract

The objective of this retrospective, longitudinal study was to investigate the prevalence of drinking within the recommended limits (i.e., low-risk drinking) following moderate/severe traumatic brain injury (TBI). Data were drawn from the National Institute on Disability, Independent Living, and Rehabilitation Research TBI Model Systems National Database (TBIMS), a longitudinal dataset closely representative of the U.S. adult population requiring inpatient rehabilitation for TBI. The sample included 6,348 adults with moderate or severe TBI (injured October 2006 – May 2016) who received inpatient rehabilitation at a civilian TBIMS center and completed the alcohol consumption items for pre-injury, and 1- and 2-year post-injury. National Institute on Alcohol Abuse and Alcoholism (NIAAA) guidelines define low-risk drinking as no more than 4 drinks per day for men or 3 drinks per day for women, and no more than 14 drinks per week for men, or no more than 7 drinks per week for women. Low-risk drinking was common both before and after TBI, with more than 30% drinking in the low-risk level pre- injury, and more than 25% at 1- and 2-years post-injury. Post-injury, the majority of drinkers consumed alcohol in the low-risk level regardless of pre-injury drinking level. Definitive research on the long-term outcomes of low-risk alcohol consumption following more severe TBI should be a high priority.

Introduction

The public health burden of traumatic brain injury (TBI) is significant, with accumulating evidence that many people with moderate or severe TBI experience deterioration in the years following injury.¹ Corrigan and colleagues postulated several causes for this long-term decline, including that damage to the frontal lobes endemic to TBI causes changes in behavioral self-regulation that lead to risky behaviors, including alcohol use.² While the potential harms associated with excessive alcohol use are well-documented for persons with TBI,^3–6 there is no data on the prevalence nor long-term consequences, if any, of drinking within low-risk guidelines after a more severe TBI. The absence of such data has contributed to confusion among rehabilitation professionals regarding the advice to give their patients regarding the safety of drinking after TBI. Patients may interpret ambiguity as license.

Longitudinal studies have shown a decrease in consumption immediately post-injury, followed by an escalation in alcohol consumption as time increases post-injury.^7–11 While there have been numerous studies documenting the prevalence of excessive drinking (e.g., binge drinking) in the years following TBI,^{4,5,10,12–16} there has been limited research on the prevalence of drinking within the guidelines; a level of drinking considered low-risk according to National Institute on Alcohol Abuse and Alcoholism (NIAAA) guidelines. NIAAA’s low-risk drinking guidelines are derived from sex-specific consumption limits for weekly and daily drinking.¹⁷ NIAAA estimates that the risk of developing alcohol use disorder is 1% among low-risk drinkers.¹⁷

Pagulayan and colleagues¹¹ examined patterns of post-injury consumption among 170 adults treated for mild to severe TBI at a Level I Trauma Center and found that 10% reported “light drinking” 3–5 years post-injury, defined as 1–2 drinks up to twice a week.^15,18 While this study¹¹ also reported prevalence for drinking that exceeded NIAAA’s daily drinking guidelines (i.e., binge drinking), it did not examine drinking within NIAAA’s low-risk guidelines based on both gender-specific daily and weekly limits.

Multiple studies have found that excessive alcohol consumption after more severe TBI has marked effects on long-term outcomes, including increased risk for suicide, seizure, and premature death.^3–6 Yet, to our knowledge, there have been no studies examining the long-term effects of low-risk drinking following moderate/severe TBI. Lacking adequate empirical evidence, rehabilitation professionals do not have an evidence base for making recommendations regarding the potential safety or harm of low-risk drinking following more severe TBI. In turn, patients may receive no or conflicting information from their providers, and form their own conclusions about how much alcohol is too much. While NIAAA’s drinking guidelines are developed to capture risk for development of alcohol use disorder, drinking within the low-risk guidelines may not mitigate alcohol-related harms. After more severe TBI, drinking in the low-risk level may increase the risk for adverse health outcomes (e.g., falls, reinjury, seizures).

For persons with moderate/severe TBI, the gap in knowledge is twofold: 1) minimal data on the prevalence of drinking within “low-risk” guidelines, and 2) a lack of empirical evidence about the long-term effects of drinking within NIAAA low-risk guidelines. The purpose of this brief report is to address the first gap in knowledge. If low-risk drinking is common, the importance of the second gap in knowledge increases. To address prevalence, we examined low-risk drinking at pre-injury and 1- and 2-years post-injury among a large cohort of adults in the U.S. who received inpatient rehabilitation for moderate/severe TBI. While this information will not provide guidance to rehabilitation clinicians who are routinely asked about the safety or harms of drinking following TBI, it will inform the field regarding the importance of addressing that question.

Methods

This is a secondary analysis of data drawn from the National Institute on Disability, Independent Living, and Rehabilitation Research TBI Model Systems (TBIMS) National Database, a longitudinal dataset representative of the U.S. population of late adolescents and adults requiring inpatient rehabilitation for TBI.^19,20 The TBIMS program is the largest longitudinal study in the world following individuals with TBI who received inpatient rehabilitation, allowing for examination of recovery and outcomes. The dataset has an attrition rate of less than 20%.²¹ The analytic sample for the current study included persons with TBI injured October 1, 2006 through May 31, 2016 who received inpatient rehabilitation at a civilian TBIMS center, were 18 years or older when enrolled, and completed the alcohol items to assess for pre-injury, and 1-year and 2-year post-injury alcohol consumption (n=6,348).

TBIMS participants were asked the following items about their alcohol consumption pre-injury (i.e., assessed during inpatient rehabilitation regarding the month before injury), and at 1-year and 2-years post-injury: 1) have you had at least 1 drink of any alcoholic beverage such as beer, wine, wine coolers, or liquor; 2) how many days per week or per month did you drink any alcoholic beverages on average; 3) on the days when you drank, about how many drinks did you drink on the average; and 4) how many times did you have 5 or more drinks on an occasion (4 or more for females) – [i.e., binge drinking]? Based on responses to these items, we constructed the following alcohol use risk groups consistent with NIAAA definitions of increasing risk for developing alcohol use disorder (i.e., abstainer, low-risk, at-risk, high-risk).¹⁷ Drinking within the recommended guidelines, or low-risk drinking, is defined as no more than 4 drinks on any day for men or 3 drinks per day for women (i.e., daily limits), and no more than 14 drinks per week for men or no more than 7 drinks per week for women (i.e., weekly limits).¹⁷ At-risk drinking is defined by exceeding either daily or weekly limits, but not both.¹⁷ High-risk drinking is defined as exceeding both daily and weekly limits.¹⁷ According to NIAAA, the risk of developing alcohol use disorder is 1% for the low-risk group, 28% for the at-risk group, and 50% for the high-risk group.¹⁷

Until 2017, TBIMS did not assess binge drinking using NIAAA recommended gender-specific cut-points; 5+ was used for both males and females. Therefore, if females reported drinking 4+ drinks on the item, “on the days when you drank, about how many drinks did you drink on average?,” or if they met the binge drinking definition, we coded them as yes to binge drinking.

Utilizing a retrospective cohort design, we computed the prevalence of the drinking risk groups at pre-injury, and 1- and 2-years post-injury. To examine how drinking risk changed from pre- to post-injury, we stratified the sample by pre-injury drinking levels and calculated the proportion whose drinking fit in each category at 1- and 2-years post-injury. Generalized linear mixed models (assuming a multinomial distribution and a generalized logit link) using PROC GLIMMIX in SAS were utilized to model changes in alcohol distributions over time, overall and within pre-injury risk groups.

Original data collection required written informed consent and was approved by each TBIMS Center’s Institutional Review Board for the conduct of research in accordance with the Declaration of the World Medical Association. Data reported were provided in aggregate by the National Data and Statistical Center. This study conforms to all STROBE guidelines and reports the required information accordingly (see Supplementary Checklist).

Results

Table 1 displays the prevalence of drinking within the NIAAA risk groups before and after injury among TBIMS participants. There were significant changes in the distribution of drinking behaviors from pre-injury to 1-year and 2-years post-injury, and between 1- and 2-years post-injury (p-values < 0.001). Low-risk drinking was common both before and after receiving inpatient rehabilitation for TBI, with more than 30% drinking in the low-risk level prior to injury, and more than one-fourth at 1- and 2-years post-injury. Even though the majority of persons with moderate/severe TBI abstained from drinking in the first two years post-injury, among those who reported drinking, the majority consumed alcohol in the low-risk level.

Table 1.

Distribution of Alcohol Risk among Persons with TBI at Pre-injury, 1- and 2-Years Post-Injury (n=6348)

	Pre-Injury		1-Year Post-Injury		2-Years Post-Injury
	n	%	n	%	n	%
Abstainer	2467	38.9%	4009	63.2%	3640	57.3%
Low-Risk	1958	30.8%	1599	25.2%	1784	28.1%
At-Risk	1153	18.2%	548	8.6%	655	10.3%
High-Risk	770	12.1%	192	3.0%	269	4.2%

Note: There were significant changes in the distribution of drinking behaviors from pre-injury to 1-year and 2-years post-injury, and between 1- and 2-years post-injury (p-values < 0.001).

Figure 1 stratifies the sample by pre-injury drinking risk group to reveal the longitudinal patterns of drinking at 1-year and 2-years post-injury by pre-injury group. Even though approximately half of pre-injury drinkers reported abstaining from alcohol at 1-year post-injury, ranging from 51.7% of pre-injury low-risk drinkers to 46.5% of highest-risk pre-injury drinkers, the overall trajectory was an increase in alcohol use as persons moved further away from their inpatient rehabilitation. There were significant changes in drinking behaviors from 1-year to 2-years within the pre-injury abstaining, low-risk, and increased risk groups (p-values < 0.001), but not within the pre-injury highest risk group (p-value = 0.29). Among all pre-injury drinking risk groups, the highest prevalence of drinking at 1- and 2-years post-injury was within the low-risk category. That is, regardless of pre-injury drinking level, if persons drank alcohol at 1-year or 2-year post-injury, they were most likely to consume alcohol within the low-risk risk group.

Figure 1.

1-Year and 2-Year Post-Injury Alcohol Risk, by Pre-Injury Drinking Level. Figure 1 stratifies the sample by pre-injury drinking risk group to reveal the longitudinal patterns of drinking at 1-year and 2-years post-injury according to the National Institute on Alcohol Abuse and Alcoholism risk categories (i.e., abstainer, low-risk, increased risk, and highest risk).

When examining the trajectory of post-injury alcohol use among the pre-injury low-risk drinkers, we found that almost 40% continued to consume alcohol within the same low-risk level one-year post-injury. Additionally, 8% of pre-injury low-risk drinkers increased their alcohol use risk at 1-year post-injury. At 2-years post-injury, almost 45% of pre-injury low-risk drinkers continued to engage in low-risk drinking, and 11% increased their drinking risk behavior.

Discussion

Using the TBIMS database of adults who received inpatient rehabilitation for a moderate/severe TBI, we found that more than one-third of persons with TBI (36.8%) reported drinking at 1-year post-injury; increasing to 42.7% at 2-years post-injury. Even among pre-injury abstainers, 20% initiated drinking by 2-years post-injury. Our findings support previous research demonstrating an immediate drop in alcohol use following TBI, followed by an increase as time post-injury increases.^11,22,23

Our study addresses the prevalence of drinking within NIAAA’s low-risk guidelines among persons who received inpatient rehabilitation for moderate/severe TBI. Within the entire sample, over a quarter reported drinking at the low-risk level at 1-year (25.2%) and 2-years (28.1%) post-injury. In fact, among all pre-injury drinking groups, the most common level of drinking at both 1-year and 2-years post-injury was drinking in the low-risk level. Pagulayan and colleagues¹¹ found a lower prevalence (10%) of “light drinking” at 3–5 years post-TBI, using a definition that diverged from the NIAAA definition. While their study¹¹ had a small convenience sample of persons with a broader severity of TBI, our study sample has been found to be representative of the U.S. population of persons requiring inpatient rehabilitation for moderate/severe TBI.^19,20 Our findings underscore the importance of future research to determine potential risks associated with drinking within low-risk guidelines following inpatient rehabilitation for TBI.

The long-term consequences of experiencing a TBI requiring inpatient rehabilitation are significant and require disease management approaches to reduce future morbidity and mortality.²⁴ Eliminating or limiting the use of alcohol is a key aspect of preventive elements for disease management of most neurological conditions.^25–27 TBI rehabilitation has been limited by a lack of clarity about the safety of low-risk drinking, despite numerous studies identifying drinking as a risk factor for TBI,^9,28 and evidence of an immediate decrease in drinking post-injury followed by an increase.^22,23

It remains unknown if drinking within low-risk guidelines is actually low-risk for persons who received rehabilitation for moderate/severe TBI, and if guidance for the general population confers the same risk for persons with more severe TBI. For instance, a male in our sample who drinks 4 drinks per night on 3 days per week, would fall into NIAAA’s low-risk drinking level (because he did not exceed either the weekly or daily limits). However, we speculate that drinking 4 drinks on 3 days per week may contribute to increased risk for persistent TBI morbidity (e.g., cognitive impairment, behavioral dysregulation, poor psychological health), adverse medical consequences (e.g., seizures, falls, high rates of reinjury, rehospitalizations) or decreased life expectancy. If future studies find that drinking within the recommended limits has a negative effect on long-term outcomes for persons with moderate/severe TBI, it would be important for rehabilitation professionals to communicate this knowledge as a substantial portion of this population consumes an amount of alcohol that would no longer be considered low-risk. This knowledge would have implications for how rehabilitation professionals communicate with their patients about safety or risks of drinking following TBI.

The appropriateness of the low-risk drinking guidelines in the US and other high-income countries has been criticized recently for being too high; a systematic review in the Lancet supports limits for alcohol consumption that are lower than those recommended in current guidelines to reduce risk for all-cause mortality and cardiovascular disease.^29,30 If low-risk drinking is being re-considered in the general population, then it certainly deserves scrutiny among persons with more severe TBI. Indeed, NIAAA states that clinicians seeing patients with specific health comorbidities, or health conditions exacerbated by alcohol, may need to recommend lower-limits of alcohol use or abstinence.¹⁷

Our findings may be limited by the self-report of alcohol consumption as opposed to structured diagnostic assessment. Studies conducted over the last 40 years on bias in self-report of alcohol consumption consistently conclude that context can affect report; however, when drinking behavior is elicited in a non-judgmental manner with no apparent consequence for reporting use, the reports are generally accurate, especially when assessed in aggregate for a group.³¹ Studies that employ objective means for evaluating self-report have supported the overall accuracy of self-report.^32,33 The sample was restricted to participants who completed the alcohol consumption items at all three time periods, yet missing alcohol data was only 2–4% at each assessment. Because the TBIMS study did not assess binge drinking using NIAAA recommended gender-specific cut-points until 2017, we anticipate that some women may be coded as low-risk when they actually would have met the female-specific binge drinking definition (moving them into the increased risk level). To improve upon this bias, females who reported drinking 4+ drinks on the item “on the days when you drank, about how many drinks did you drink on average?” were coded as yes to binge drinking, reducing this bias.

Conclusions

While many persons who received inpatient rehabilitation for TBI abstained from drinking at 1- and 2-years post-injury, the majority of those who continued drinking did so within the NIAAA guidelines (i.e., low-risk level). Definitive research on the long-term outcomes of low-risk alcohol consumption following more severe TBI would be beneficial so that there is an evidence-base to determine if individuals with moderate/severe TBI should be advised to drink less than current low-risk guidelines or not at all. Medical and rehabilitation professionals need this information to be able to provide evidence-based advice.