In the period 2000 to 2016, the percentage of drinkers in the world has decreased by almost 5%. However, since the world’s population has increased, the decrease in the prevalence of drinking did not translate into fewer current drinkers [3]. In the period 2006 to 2014, the number of acute and chronic alcohol-related emergency department (ED) visits increased significantly [15]. Furthermore, between 2001 - 2002 and 2012 - 2013, 12-month alcohol use, high-risk drinking, and alcohol use disorder increased significantly. The increase in these outcomes were greatest among women, older adults, and racial/ethnic minorities [2].

The incidence and the admission rates of AWD increased significantly over the studied period. There was also a trend of worsening clinical outcomes (inpatient mortality, LOS, and THC) in our analysis. Dose-response relationships could be quantified for all alcohol-related disease categories, with the relative risk increasing with increased level of alcohol consumption for most diseases [16]. It has been found that the likelihood of experiencing withdrawal symptoms increased with increasing alcohol consumption [17]. These trends likely reflect the changes in alcohol consumption observed in the last two decades.

Our analysis showed that middle aged adults and elderly patients had higher mortality rates, LOS, and THC when compared to young adults. Alcohol consumption has been increasing in the elderly population particularly in women [18]. In addition, there has been a recent increase in binge drinking in this population [19]. Evidence suggests that elderly have higher mortality from alcoholic liver disease compared to younger patients [20-23]. Liskow et al found that elderly patients had more severe withdrawal and required higher doses of benzodiazepines [24]. Wonjar et al found the elderly patients had longer hospital stays for AWD and more electrolytes abnormalities compared to younger patients [25]. Similarly, Brower et al found that elderly patients had more withdrawal symptoms for longer duration [26]. The etiology of the worse outcomes in this population is multifactorial. Elderly patients with alcohol use disorder have higher rates of respiratory, endocrine, cardiovascular and central nervous system (CNS) diseases, and dementia compared to young patients [27]. In addition, older patients are at higher risk for cognitive and functional impairment during withdrawal [28]. Older adults have increased sensitivity to benzodiazepines and decreased metabolism of long-acting agents. Benzodiazepines are associated with increased risk of delirium and cognitive impairment in older adults [29]. Furthermore, many older adults drink alcohol and take medications that may interact with alcohol. Some of these interactions are due to age-related changes in the absorption, distribution, and metabolism of alcohol and medications [30]. Exacerbation of therapeutic and adverse effects of medications when combined with alcohol likely contributes to the observed outcomes in our study. Admissions for substance use including cocaine/crack, marijuana/hashish, heroin, non-prescription methadone, and other opiates have increased in the elderly population in the last decade [31, 32].

Our study showed that females had better clinical outcomes in terms of inpatient mortality, LOS, and THC when compared to males, a finding that is consistent with the available evidence [33, 34]. It has been found that withdrawal seizures were significantly more frequent among men than among women [35]. Canales et al found that men with AWD had higher rates of sepsis and longer LOS compared to women [36]. Benedict et al found that male gender was associated with 2.6-fold increase in the risk of benzodiazepine resistant withdrawal [37]. Drinking patterns differ between males and females. According to the global status report on alcohol and health, females are less often current drinkers than males and they drink less on average and engage less often in heavy episodic drinking [3]. Women prefer mild types of alcoholic beverages such as wine and beer [38-40]. They become intoxicated less frequently and are less likely to abandon personal control while drinking [41]. On the other hand, women tend to present with more severe liver disease, particularly alcoholic hepatitis, and do so after a shorter period of excessive drinking and at a lower daily alcohol intake [42, 43]. They also have higher mortality from alcoholic liver cirrhosis compared to men [44]. Interestingly, Goodson et al found that the presence of liver disease was not predictive of alcohol withdrawal [45].

Glutamate-N-methyl-D-aspartic acid (NMDA), γ-aminobutyric acid (GABA), and many other receptors are involved in the adaptive changes in the brain during chronic ethanol exposure [46]. Evidence from animal studies suggests the alcohol induces gender-selective adaptations at the molecular level. It has been found that alcohol-dependent females did not show the same alterations in gene expression for GABA or NMDA receptors as did alcohol-dependent males [47]. Female rats recover more quickly from withdrawal than male rats, even with a greater volume per weight alcohol intake [48, 49]. Differences at the hormonal level also play a role; Devaud et al suggested that the differences between men and women in AWD may be mediated by allopregnanolone [50], an endogenous GABA receptor ligand with anxiolytic and anticonvulsant properties [51].

Our analysis showed that hospitalizations for Blacks were associated with lower mortality and THC, however, they had longer LOS. Caetano et al found that the prevalence of withdrawal symptoms was generally lower among Blacks than among Whites and Hispanics [17]. It has been found that Caucasian race was 3.6 times more likely to be develop benzodiazepine resistant withdrawal [37]. Chan et al found that White patients have twice the odds of having severe alcohol withdrawal when compared to non-White patients, and Black patients have smaller odds of having severe alcohol withdrawal when compared to non-Black patients [52].

The prevalence of alcohol abuse is greater among Whites than among Blacks [53]. In addition, Naimi et al and Herd et al found that Blacks consistently had the lowest rates of binge drinking episodes compared to other ethnicities [54, 55]. Evidence suggests a racial variation in the toxic effects of alcohol. Stewart et al found that Black non-Hispanic and Mexican Americans were more likely to have a 2-fold elevation in aspartate aminotransferase levels when compared with White non-Hispanic Americans [56].

The reason for the longer LOS observed in Blacks is unclear. It has been found that Black patients were more likely to be cocaine dependent than alcohol dependent and were more likely to experience cocaine withdrawal symptoms rather than alcohol withdrawal symptoms [17]. Whether withdrawal from other drugs have led to the observed higher LOS in this population needs further investigations. The lower mortality rate in this population might also contribute to the observed longer LOS.

Our study showed an increase in the incidence and the admission rates for AWD in the period 2010 to 2019. Mortality, LOS, and THC increased over the studied period. Younger age, female gender, and Black race were associated with better clinical outcomes. Our study has some important limitations. The NIS reports information on hospitalizations rather than individual patients, as a result, one patient can be included more than once in the analysis. The severity of AWD could not be determined from the NIS. We could not determine what proportion of patients were managed in the intensive care setting (which may impact outcomes). Therapeutic interventions, medications given, laboratory results, and imaging are not provided in NIS. Finally, the study used ICD codes; therefore, the database may contain errors related to miscoding.