Sex difference in alcohol withdrawal syndrome: a scoping review of clinical studies

Background We conducted a review of all studies comparing clinical aspects of alcohol withdrawal syndrome (AWS) between men and women. Methods Five databases (PubMed, Cochrane, EMBASE, Scopus and Clinical Trials) were searched for clinical studies using the keywords “alcohol withdrawal syndrome” or “delirium tremens” limited to “sex” or “gender” or “sex difference” or “gender difference.” The search was conducted on May 19, 2023. Two reviewers selected studies including both male and female patients with AWS, and they compared males and females in type of AWS symptoms, clinical course, complications, and treatment outcome. Results Thirty-five observational studies were included with a total of 318,730 participants of which 75,346 had AWS. In twenty of the studies, the number of patients presenting with or developing AWS was separated by sex, resulting in a total of 8,159 (12.5%) female patients and a total of 56,928 (87.5%) male patients. Despite inconsistent results, males were more likely than females to develop complicated AWS [delirium tremens (DT) and AW seizures, collective DT in Males vs. females: 1,792 (85.4%) vs. 307 (14.6%), and collective seizures in males vs. females: 294 (78%) vs. 82 (22%)]. The rates of ICU admissions and hospital length of stay did not show sex differences. Although variable across studies, compared to females, males received benzodiazepine treatment at higher frequency and dose. One study reported that the time from first hospitalization for AWS to death was approximately 1.5 years shorter for males and males had higher mortality rate [19.5% (197/1,016)] compared to females [16% (26/163)]. Conclusion Despite the significant heterogeneity of the studies selected and the lack of a focus on investigating potential sex differences, this review of clinical studies on AWS suggests that men and women exhibit different AWS manifestations. Large-scale studies focusing specifically on investigating sex difference in AWS are needed.


Introduction
Alcohol withdrawal (AW) is a common medical condition with characteristic clinical manifestations that take place a few hours to a few days after cessation or significant reduction in heavy and prolonged alcohol consumption.These manifestations range from mild anxiety, nausea, shaking, and agitation to seizures, delirium tremens (DT), and death (1)(2)(3)(4).According to DSM-5 TR, less than 10% of individuals with alcohol withdrawal syndrome (AWS) will ever develop severe autonomic hyperactivity and alcohol withdrawal DT, and less than 3% of individuals will experience tonic-clonic seizures during alcohol withdrawal (5).Nonetheless, AWS is a major public health problem in the U.S. with approximately 500,000 episodes per year sufficiently severe to require pharmacological treatment (6).Approximately 50% of middle-class, highly functioning individuals with alcohol use disorder (AUD) experience AWS, and more than 80% of hospitalized AUD patients may experience alcohol withdrawal (5).Even though AWS is highly prevalent and causes significant morbidity and mortality, there is a dearth of information about its prevalence among women or sex-difference in clinical manifestations, treatment response or outcome (7,8).
Until recently, men surpassed women by a wide margin in social and problematic alcohol drinking patterns (9, 10), and most of our clinical knowledge about AWS came from studies that enrolled predominantly (11)(12)(13)(14)(15)(16)(17) or only men (18,19).Over the past two decades, a robust increase in alcohol consumption (20), hazardous drinking (21)(22)(23), alcohol-related emergency room visits (24), and AWS (25) have been observed among women, especially among adolescents (23,(26)(27)(28)(29). Current management of AWS in women assumes that women exhibit clinical manifestations, respond to treatment, and develop AWS complications similar to men.However, gaining a deeper understanding of the role of sex in AWS is crucial to both recognizing and effectively managing this complex syndrome.
Animal studies have shown evidence of sex differences in AWS (30) with male rats had greater withdrawal seizure susceptibility than female rats (31, 32) and only male mice experienced increased seizure risk following repeated alcohol withdrawal episodes (33).In addition, male rats exhibited increased anxiety like behaviors during alcohol withdrawal which was demonstrated both with enhanced acoustic startle responses, elevated plus maze and suppressed social activity (32,34,35).One study showed a significantly lower alcohol withdrawal severity in female mice than male mice (36).Despite these wellestablish sex differences in animal AWS models, there are inconsistencies between clinical AWS studies; some suggest that men tend to experience more severe AWS than women (37)(38)(39)(40)(41), while others demonstrate that women can also experience AWS as severe as men (42)(43)(44)(45)(46)(47).It is essential to recognize the significance of sex differences in AWS, as evidenced by both animal and clinical studies, highlighting the urgency to better understand and address these differences.
To establish the findings on sex differences in AWS in the current literature and address disparities, we conducted a comprehensive scoping review of all clinical studies related to AWS.In this review, we compare outcomes including AWS symptoms, AWS complications, hospital length of stay (LOS), ICU admission rates and LOS, laboratory values, clinical course, and treatment plans between males and females.We hypothesize that women can also present with complicated AWS, because it is well-established that women are more vulnerable to developing alcohol-related complications at lower-level drinking and after a shorter duration of alcohol consumption (time from first use to dependence) compared to men (48)(49)(50).However, we expect to see that males will be more likely to suffer from alcohol withdrawal seizures, as consistently shown in animal studies.

Methods
We conducted a scoping review of the literature by searching five databases, PubMed, Cochrane, EMBASE, Scopus and Clinical Trials.The search strategy was designed and conducted by an experienced librarian.Controlled vocabulary with keywords was used to search for studies describing alcohol withdrawal and sex difference.Supplementary file 1 provides an outline the search strategy listing all the search terms used and how they are combined.This review was registered prospectively with PROSPERO (CRD42023394108).

Article selection and quality assessment
The search was conducted on May 19, 2023, and yielded 796 records.An abstract review identified articles addressing clinical symptoms and their management; these were shortlisted.From all original articles (open-label or double-blind trials, prospective or retrospective observational studies, and cohort or cross-sectional studies) written in English, we included those that met the following criteria: (1) study participants consisted of both male and female patients experiencing alcohol withdrawal syndrome, (2) results compared males and females in at least one of the following outcomes: (i) AWS symptoms, (ii) AWS complications (i.e., delirium tremens, seizures, hallucinations, or mortality), (iii) ICU admission rates, (iv) hospital length of stay, (v) laboratory values and clinical course, or (vi) treatment plan.We excluded pre-clinical studies, articles not in English language, articles not in full text, studies that do not compare at least one outcome of interest between males and females, reviews, commentaries, or letters to the editor.All authors agreed on the inclusion and exclusion criteria.The inclusion or exclusion of individual studies was discussed between the two lead authors (HU and MMM).Articles with non-agreement were discussed with the senior author (OAA).The full texts of the shortlisted articles were reviewed.Cross-references were searched from selected studies and relevant articles were also evaluated for inclusion (Figure 1).
The quality of each study was independently evaluated by the two lead authors (HU and MMM).Cohort studies were assessed using the Newcastle Ottawa Assessment Scale (NOAS) (51).This scale evaluates various aspects, including sample selection (representativeness of the target population, sample size, comparability between respondents and non-respondents, and outcome ascertainment), comparability (comparability between subjects in different outcome groups), and outcomes (method of outcome measurement and statistical tests employed).A maximum of four stars can be given to a study under the category of selection, two stars under the comparability category, and three stars for the outcome category.The adapted for cross sectional studies NOAS was used to assess cross sectional studies (52).This scale evaluates sample selection (representativeness of the cases, sample size, non-response rate, and ascertainment of screening/ surveillance tool), comparability (potential confounders), and outcome (outcome assessment, and statistical test).Sample selection can be awarded a maximum of 5 stars, comparability can be awarded a maximum of one star, and outcome can be awarded a maximum of 3 stars.Finally, observational studies were assessed based on the observational studies NOAS (53).This scale evaluates sample selection (representativeness of the exposed cohort, selection of the non-exposed cohort), exposure (ascertainment of exposure, exposure dose, retrospective/prospective dose ascertainment), comparability (confounding), outcome assessment, and follow up (period and adequacy).Sample selection and follow-up assessment can be awarded a maximum of two stars; comparability and outcome assessment can be rewarded a maximum of one star each; ascertainment of exposure can be awarded 4 stars maximum.Discrepancies in quality assessment were resolved through discussions involving a senior author (OAA) until a consensus was reached.The results of the quality assessment for all included studies can be found in Supplementary file 2. The data were synthesized, and the relevant findings are discussed below.

AWS severity assessment scales
Among studies included, three studied used the Clinical Institute Withdrawal Assessment of Alcohol Scale, revised (CIWA-Ar) (54)(55)(56) to measure the severity of AWS.Additionally, two studies used the Clinical Institute Withdrawal Assessment of Alcohol (CIWA-A) (57,58), and one used the Alcohol Withdrawal Scale (AWS) (59).

Results
Thirty-five observational studies met the inclusion/exclusion criteria.Almost half of the included studies were conducted in the U.S. (n = 15), followed by Spain (n = 8), Germany (n = 6), including a study conducted in both Germany and Nigeria, and Poland (n = 2).The other 4 studies came from Denmark, Croatia, Sweden, and Australia.More than a third (n = 13) were published prior to the end of 2000.The details of these studies are summarized in Table 1.These 35 studies included a total of 318,730 participants of which 75,346 presented with or developed AWS.Twenty of the studies (n = 65,087) separated the number of AWS patients by sex, resulting in a total of 8,159 (12.5%) female patients and a total of 56,928 (87.5%) male patients.(55) showed a non-significant trend for male sex as a risk factor for AWS [Relative risk (RR) = 2.2, 95% CI: 0.9-5.1,p = 0.06].Steel et al. 's study with inpatient veterans showed that male sex was associated with an increased probability of inpatient AWS (p < 0.001) (74).Among a cohort of males and females with alcohol dependence, Deshmukh et al. (66) observed higher endorsement of AW criterion among men (p < 0.025).Similarly, Jarque-López et al. reported major AWS was more common among men (224/278, p < 0.001) (37).Interestingly, one study in adolescents reported that the mean number of AWS symptoms (i.e., irritability, tremor, nausea and/or vomiting) was higher among females [Mean (SD) for males vs. females = 9.31 (5.6) vs. 11.1 (5.7), p < 0.05] (75).

Sex difference in clinical characteristics
Wojnar et al. (43) reported that women with AWS were non-significantly older [males (n = 1,016) vs. females (n = 163): 41.2 ± 11 vs. 39.8 ± 9.8 years] and drank significantly less alcohol than men (p < 0.0001).Also, the time from intensive drinking to the first withdrawal episode was 5 years shorter in women (p < 0.0001).Of note, this study did not define intensive drinking.O'Connor et al. (47) reported no significant difference in age [males (n = 145) vs. females (n = 34): 38 ± 10 vs. 36 ± 10 years] or in the amount of alcohol intake.In their study, the duration of alcohol misuse before developing AWS was found to be 3 years less in women than in men, though the finding was not significant (19.1 vs. 16.1 years, p = 0.08).Canales et al. ( 42) recently reported that among patients with AWS, women were younger than men [males (n = 1,372) vs. females (n = 118): 45.6 vs. 43.9years, p = 0.009].Schuckit et al. (39) compared 160 men and 51 women with severe AWS, the two groups were not significantly different in the age of onset of alcohol dependence (males vs. females: 24.5 ± 8.9 vs. 23.5 ± 8.8 years), years of heavy drinking (14.2 ± 10.0 vs. 11.0 ± 8.5), frequency of drinking (6.0 ± 1.8 vs. 5.4 ± 2.1 days per week), and total number of alcohol use problems (33.3 ± 6.2 vs. 32.5 ± 6.2).Family history of AUD was found to be a risk factor for AWS in women (OR = 2.85, 95% CI: 1.07-7.54)(8) while amount of alcohol consumption was a risk factor for AWS in both men and women, although it was more likely in women (75).Earlier onset of alcohol consumption (OR for every 5 years = 1.89, 95% CI:1.69-2.08)was associated with increased probability of AWS in men (8).

Sex difference in AWS medical and psychiatric comorbidities
Wojnar et al. found that personality, anxiety, and depressive disorders, benzodiazepine and barbiturate misuse were more common among women with AWS (p < 0.0001) (43).In contrast, Schuckit et al. (39) reported that there was no sex difference regarding a number of different substances used (males vs. females: 2.1 ± 1.9 vs. 2.4 ± 1.8), the pattern of exposure to various drugs, history of medical conditions, and psychiatric symptoms.Women tended to report more past depression (females vs. males: 94% vs. 81%); however, there was no statistically significant difference after Bonferroni corrections.Comorbid cannabis use (OR = 2.8, 95% CI:1.04-7.7)was associated with increased probability of AWS in men (8)

Discussion
The results of this review highlight heterogenous methodology and inconsistent findings on potential clinical differences between men and women during AWS.Because of this heterogeneity, we were unable to conduct a meta-analysis.Despite hundreds of clinical studies on different aspects of alcohol withdrawal, several studies included only male patients (18, 19,79,80).Others included female patients, but the results were merged for both sexes (18, 81-84) making drawing a conclusion on sex difference impossible.While most of the literature on AWS focuses on treatment, the studies did not examine potential differences in treatment outcome between men and women (7,(13)(14)(15)(16)(17)(85)(86)(87)(88)(89)(90)(91)(92)(93)(94).As such, our understanding and treatment of AWS in females follow the assumption that there are no sex differences in response to treatment, which may not be the case.
This dearth of sex-specific information is surprising given that a significant proportion of general hospital (8%) and ICU (20%) patients experience signs and symptoms of AWS (95), and still women were underrepresented in clinical studies.In our review, only 13% of all AWS patients were females.It is true that about a third of these studies were conducted before the year 2000, while the rapid increase in drinking among women has been witnessed over the past decade or two (20)(21)(22)(23)(25)(26)(27)(28)(29).The observed increase in alcohol consumption rates, particularly among young women, could be related with menstrual cycle changes and associated distress (96)(97)(98).However, we did not include studies in our scoping review based on this variable, but rather on sex differences in alcohol withdrawal manifestations.Future studies are needed to highlight this aspect.
Our careful inspection of the current literature shows that while both men and women suffer from several medical and psychiatric comorbidities, extended hospital stay, high rates of ICU admission, complications during acute alcohol withdrawal, and high rates of mortality after discharge, there are several reported sex differences.Specifically, males appear to have higher rates of AWS-associated DT and mortality.Nonetheless, it is still a challenge to identify sex differences in many clinical aspects of AWS because of the heterogeneity of studies, small number of women included, and the inconsistent methodology.Large scale studies are urgently needed to examine potential differences between males and females in various aspects of the AWS.

Patients characteristics and risk factors
Two studies (43,47) showed the time between alcohol misuse and first withdrawal of alcohol was shorter in women than men which supports telescoping phenomena.Telescoping refers to the pattern of females' accelerated progression from the onset of alcohol drinking to alcohol-related problems and treatment seeking, when compared to males (47,99).However, a more recent study suggested that the telescoping effect is not evident in the general population (48).
Several studies have found multiple factors that increase the risk and severity of AWS, including previous AWS episodes, higher drinking levels, concurrent illness, and abnormal clinical and laboratory findings (44,81,83,84,100,101).When these risk factors are considered in terms of sex differences, the studies have revealed inconsistent results.Goodson  however, they included studies with predominantly male patients and also concluded that further research is needed to evaluate the effect of sex on the course of AWS.Similarly, Wood et al. (101) concluded in their review that having severe AWS was not more likely in men than in women [Likelihood ratio (LR) for men 1.3, 95% CI: 1.0-1.7].Some studies report that male sex is a risk factor of developing severe AWS (40, 55), while others do not (46, 64).

Laboratory findings
Previous studies showed that thrombocytopenia (100, 102), low serum potassium, (100, 103), higher initial alanine aminotransferase (ALT) (100,103), were observed among patients with severe AWS.In our review, Wojnar et al. (43) was the only study that gave laboratory results separately based on sexes and showed that anemia and low serum potassium level were observed more commonly among women, while increased ALT, and hypoproteinemia were observed among men.It is known that anemia is a more common finding in women (104).Likewise, a study evaluating comorbidities in AUD patients observed that women experience significantly more anemia (105).Therefore, it is unclear whether Wojnar's finding of anemia more commonly in women with AWS simply reflects the more the finding across female population (43).Hypokalemia, reported by the literature as one of the most prominent risk factors for developing severe AWS was observed more frequently in women in Wojnar's study.While total body potassium is lower in women, serum potassium concentration is sex-independent (106), therefore, the hypokalemia observed might be related with AWS.Male sex is related to higher serum ALT levels (107), and findings of higher ALT levels in male AWS patients (43) might be independent of AWS itself.However, as higher ALT levels were associated with severe AWS, higher ALT levels in males might put male patients at risk of developing more severe AWS.As such, it is possible to speculate that male patients with AUD and alcoholassociated liver disease are more likely to develop AWS, compared to females.Berggren et al. (65) reported that thrombocytopenia, a finding associated more commonly with severe AWS, was also observed more in male patients.However, it is known that men have lower platelet counts compared to women regardless of AWS (108), therefore observing thrombocytopenia more commonly in male AWS patients might not be related to AWS.As studies demonstrated that thrombocytopenia is a predictor of severe AWS (65, 100, 102), the normally lower platelet value in men can be interpreted as men may be more susceptible to developing severe AWS.

Medical and psychiatric comorbidities
Prior studies have shown that women are more sensitive to the toxic effects of alcohol (78,109), have lower activity of alcohol dehydrogenase enzyme (110, 111), higher total body fat, less body weight and lower liver mass compared to men (5, 112).Therefore, women have higher blood alcohol concentrations for the same amount of alcohol than men (113).Women may be at higher risk of alcohol associated liver disease than men at any level of alcohol intake (114).Liver diseases and cirrhosis occur in women with shorter and lesser amounts of alcohol use compared to men (115).Two studies in our review support these findings (64, 78).
Wojnar et al. (43) reported that comorbid psychiatric disorders and substance use disorder were more common among males, while Schuckit et al. (39) concluded that there was no sex difference regarding these comorbidities.While Wojnar et al. retrospectively reviewed the records of patients hospitalized for AWS, Schuckit et al. conducted a structured interview with alcohol-dependent patients.Differences in methodology may explain, at least in part, these contrasting results.Also, the different regions and populations where the studies were conducted (Poland vs. USA) may also have confounded the results, as cultural drinking habits, availability of other addictive substances, and genetic factors may vary.

Hospital length of stay, ICU admissions
Wojnar et al. (43) reported that women needed longer course of hospital stay, while Canales et al. (42) reported that men needed longer hospital stay and ICU stay.The difference between the time periods (1997 vs. 2022) and the regions of the studies (USA vs. Europe) could contribute to the difference in findings.In addition, the difference in patients' comorbidities may have contributed to these contradicting results, as Canales et al. (42) demonstrated that men experienced more sepsis and pneumonia.Ring et al. (71) did not observe sex difference regarding length of hospital stay; however, they found that comorbid pneumonia prolonged the hospital stay.As can be seen from this study, pneumonia itself might be the reason for longer hospital and ICU stay for males in Canales's (42) study.Of note, women were older in Wojnar's et al. (43) study population, while men were older in the Canales's study.This difference in age across the two studies may be another factor that could have played a role in their different findings.
Overall the male to female ratio for all alcohol-related ICU admissions was 4.2:1 in the literature (116).Among the publications selected in our review, Canales et al. (42) was the only study that evaluated sex differences in ICU admissions in AWS and observed that the percentage of ICU admissions was similar (16%) for men and women.This finding can be interpreted as indicating that the severity of AWS does not differ between sexes, but more studies are needed to draw a solid conclusion.

AWS complications
One study (43) demonstrated that women were significantly more likely to develop DT, while most of the studies (37,38,40,41,57,78) reported significantly higher number of males or no significant sex differences (42, 47) in DT cases.Two studies (43,57) reported significantly higher number of AW seizures among males, while others (42, 73) did not find any significant difference.There were also studies (44)(45)(46)(47)64) that did not find sex differences in AWS complications.The studies that did provide sex-specific statistics on alcohol withdrawal seizures did not evaluate potential sex differences in phenotypes of alcohol withdrawal seizures and other seizure-related parameters such as seizure onset age, status epilepticus development, and treatment response.Similarly, studies did not evaluate characteristics of hallucinations in terms of sex difference.Male predominance seen in AWS complications might be related to underrepresentation of female frontiersin.orgparticipants or indeed male patients being more prone to complications of AWS.To understand this, studies examining the incidence of AWS complications in men and women separately are needed.

AWS treatment
Females were found to exhibit more treatment side effects (57), while there was a disagreement among studies regarding medication doses by sex (42,43).These findings might be related to participant characteristics (comorbidities, age, severity of AWS).As with treatment of AUD, AWS treatment may require sex-specific individualization for optimal care (117).Due to the sparsity of literature on the topic, no reliable conclusion can be drawn on differences in management of AWS between men and women.

AWS mortality
Patients with AUD have increased mortality rates compared to the general population, and the mortality rate for men is greater than that for women (118).Similarly, Lewis et al. (41) found significantly shorter time from hospitalization to mortality among males with AUD compared to females.Moreover, men were found to have higher medical complications and more severe AUD than women.Consistent with these findings, one study (43) reported higher mortality rate and shorter time between first hospitalization and mortality in males, while two study did not report sex difference in AWS mortality (2, 70).

Limitations
The major limitation of our review of sex differences in AWS is the scarcity of research specifically examining this topic in a systematic manner.Most previous studies do not report separate data for males vs. females in all their results.In addition, even when reporting sex differences in results, studies reported predominance by comparing total number of males vs. total number of females for a particular outcome rather than comparing the sex-specific incidence of the outcome.The qualities of the studies were highly variable; as well, and not all studies reported p-values.In this review, not only were the studies heterogeneous based on geographical location, but also based on the treatment setting and populations.Studies focused on trauma patients (54,55), emergency room patients (25,37), inpatients who developed AWS (8, 25, 38, 44, 58, 59, 64, 68, 73-75, 77, 78), hospital patients presenting with AWS (39, 43, 45-47, 63, 67, 70, 71, 74), research center patients (39), psychiatric hospital patients (41), and detoxification unit patients (77).All study populations were mostly adults, however, only one study (75) focused on AWS in adolescents, which led to different results, as discussed previously.

Conclusion
Despite these limitations, our review highlights several significant differences in the clinical manifestations and treatment outcomes between men and women and brings more wariness to the unmet need to include more women in large-scale AWS studies and quantify sex specific differences and effects in the development, diagnosis, clinical manifestations and treatments of AWS and its complications.With the recent progressive increase in alcohol drinking among women, it is imperative for the scientific community to update the current research on AWS that has focused primarily on men and/or has not analyzed potential sex differences.

TABLE 1 (
Continued) (100).(100)concluded that sex was not a predictor of severe AWS development (DT and, or seizures);

TABLE 2
AWS complications.females (52.6%), 110/180 males (61.1%) with SAWS (seizures, disordered perceptions, or delirium) a , p value for sex difference in all complications = 0.2 26.7% females, 73.3% males b Number of female or male patients with condition/total number of female or male patients (percentage).b Percentage of female or male patients with condition from the total number of patients with condition. a