Sex difference in IgE sensitization associated with alcohol consumption in the general population

The association of alcohol consumption and immunoglobulin E (IgE) sensitization is debated. Few population-based studies have investigated whether such associations differ by sex. We explored the association of alcohol consumption with IgE sensitization in the general population, stratified by sex. We analyzed data for 1,723 adults from the 2010 Korean National Health and Nutrition Examination Survey. We divided subjects into three groups according to their self-reported alcohol consumption or serum level of gamma-glutamyltransferase (GGT), an objective marker of alcohol consumption. After adjustments, the odds ratios (ORs) of male high-risk drinkers were 2.09 (95% confidence interval [CI], 1.34–3.28) for total IgE and 1.71 (95% CI, 1.03–2.83) for Dermatophagoides farinae (DF)-specific IgE compared with male low-risk drinkers. In females, the dog-specific IgE level was associated with high-risk drinking (OR, 11.74; 95% CI, 2.04–67.24). The ORs of males in the high-serum-GGT group were 2.73 (95% CI, 1.72–4.33) for total IgE and 2.17 (95% CI, 1.35–3.47) for DF-specific IgE compared with those in the low-serum-GGT group. This study suggests a possible link between alcohol consumption and IgE sensitization, moreover, the risk of IgE sensitization was significantly higher in male high-risk drinkers. Therefore, clinicians should consider the risk of IgE sensitization possibly afflicting male high-risk drinkers.

A considerable body of evidence supports sex differences in the prevalence of alcohol-use disorders and the physical effects of alcohol 15 , as well as in alcohol consumption, metabolism, and effects on the brain and certain receptor systems 16 . Cross-cultural comparisons generally report that the biological mechanisms underpinning the sex-specific differences in the effects of alcohol and related problems are independent of cross-country and cross-generational influences 17 . Also, sex differences in IgE sensitization to aeroallergens have been reported in the general population 18 . However, no study has examined sex differences in the association of alcohol drinking with the serum IgE level.
High-risk drinkers can be identified using the Alcohol Use Disorders Identification Test (AUDIT) questionnaire, which comprehensively assesses alcohol abuse 19,20 . The AUDIT questionnaire is a reliable test to measure high-risk across sex, age, and cultures 21,22 , underscoring its appropriateness for our study. However, as AUDIT is a subjective instrument for screening high-risk drinkers, its accuracy may be limited by poor respondent memory due to high alcohol consumption. Serum gamma-glutamyl transferase (GGT) is a liver enzyme that catalyzes the transfer of gamma-glutamyl groups to other amino acids and is a useful surrogate marker for excessive alcohol consumption 23 . We used the serum level of GGT as a marker of alcohol exposure and divided the subjects into three groups according to the sex-specific GGT tertile.
To our knowledge, no prior study has investigated the association of alcohol consumption pattern and the serum GGT level with IgE sensitization in a large population, or in one stratified by sex. The purposes of this population-based study of Korean adults were to 1) investigate the association between alcohol consumption and sensitization to specific allergens and 2) assess whether these associations differ by sex.

Results
Sociodemographic and clinical characteristics. The sociodemographic characteristics of the study population (unweighted number of participants, 918 males and 805 females) are listed in Table 1. The male and female subjects were divided into low-, intermediate-, and high-risk drinkers based on their AUDIT scores. Males constituted a higher proportion of the intermediate-risk (33.1% vs. 12.9%) and high-risk (23.3% vs. 3.7%) drinkers than females (p for trend < 0.0001) ( Supplementary Fig. S1). The mean age was highest in male high-risk drinkers and female low-risk drinkers. For male groups, the proportions of high-risk drinkers were lowest in young adults and the elderly, and highest in those aged 41 to 50 years, generally showing a U-shaped distribution ( Supplementary Fig. S2). Among females, the proportions of intermediate-and high-risk drinkers tended to decrease with age. High-risk drinkers of both sexes were more likely to be employed. In addition, high-risk drinkers of both sexes tended to be current smokers. Female high-risk drinkers were most likely to be obese and have no spouse, but no such findings were evident in males. In subgroup analyses, no significant differences were found in urban residence, educational level, income, sleep duration, or serum vitamin D level among the three risk groups.
Correlation between the AUDIT score and the serum GGT level. We examined the correlation between the AUDIT score and the serum GGT level using a general linear model. The AUDIT score had positive correlation with serum GGT level (coefficient (B): 2.626, 95% confidence interval (CI): 2.260-2.992). This correlation had statistical significance (R 2 = 0.129, p < 0.0001).
Allergy related factors according to alcohol consumption pattern. The prevalence of allergic disease did not differ according to alcohol consumption pattern in either sex (Table 2). Among males, the proportion of an increased total IgE level among high-risk drinkers (55.9%) was significantly higher than that in low-(38.6%) and intermediate-risk ( prevalence of an increased total IgE level or sensitization to specific allergens among the three alcohol consumption patterns. Associations between alcohol consumption pattern and total and specific IgE levels. Table 3 shows the associations of alcohol consumption pattern with total and specific IgE levels by sex. In males, without adjustment, high-risk drinkers had a significantly increased OR for total and DF-, cockroach-specific IgE levels, and atopy compared with low-risk drinkers. After multivariable adjustment, the OR for total IgE, sensitization to DF, and atopy remained significant (Fig. 1). Without adjustment, female high-risk drinkers had a significantly increased OR for total and DF-and atopy-specific IgE levels compared with the low-risk drinkers. However, after multivariable adjustment, only the OR for dog-specific IgE level remained significant for female high-risk drinkers (Fig. 2).
Associations between serum GGT levels and total and specific IgE levels. Table 4 presents the associations of serum GGT tertiles with total and specific IgE levels by sex. Without adjustment, males in serum GGT tertile 3 showed a significantly increased OR for total and cockroach-specific IgE levels compared with those in tertile 1. After multivariable adjustment, the total and the DF-specific IgE and atopy were significantly associated with serum GGT tertile 3 but not with GGT tertile 1. The three IgE factors associated with a high serum GGT level were also associated with high-risk drinking (Fig. 3).
Without adjustment, among females, GGT tertile 3 was significantly associated with the total IgE level, and GGT tertile 2 with the cockroach-specific IgE level, compared with GGT tertile 1. However, after multivariable adjustment, no IgE level was significantly associated with GGT tertile 2 or 3 in females (Fig. 4).

Discussion
This study in the general population examined the associations of alcohol consumption and serum GGT level with IgE sensitization by sex. Male high-risk drinkers had increased total and DF-specific IgE levels and a higher prevalence of atopy, while female high-risk drinkers had a high dog-specific IgE level, after adjusting for all confounders. Furthermore, in males, the three IgE factors related to high-risk drinking were associated with a high serum GGT level. In females, no IgE was associated with a moderate or high serum GGT level. Previous studies of the association between alcohol consumption and allergic diseases and/or IgE sensitization have yielded inconsistent results. Moderately high alcohol consumption, as well as alcoholism, may be associated with high IgE levels and allergic disease 12,24,25 . Coutinho et al. reported that the serum total IgE level decreased significantly after cessation of alcohol consumption in Portuguese adults 26 . However, the association between alcohol consumption and IgE sensitization is less clear 13,27 . These inconsistent results might be caused by differences in alcohol exposure, laboratory methods, sample size, allergen sensitization profile, lifestyle, sex, and ethnicity among prior studies. Our findings suggest the importance of sex differences in the association between alcohol consumption and IgE sensitization.
The association of alcohol use with the GGT level has been demonstrated at the population level. A population-based study in Germany examined the associations of liver enzyme levels (as indicators of alcohol exposure) with the serum total IgE level. Alcohol consumption and the levels of liver enzymes, including that of GGT, were positively associated with the serum total IgE level in non-atopic and atopic subjects, consistent with our results 28 . However, the association of liver enzymes with the levels of allergen-specific IgE was not evaluated. Our findings of a significant association between both alcohol consumption pattern and the serum GGT level and the same three IgE factors (increased total and DF-specific IgE levels and the presence of atopy) strongly suggest that high-risk drinking is significantly associated with IgE sensitization in males.
Increased total IgE levels in heavy drinkers should be interpreted with caution because of the possible presence of specific IgE to cross-reactive carbohydrate determinants (CCDs) 29 . CCDs are N-glycans of plant and  www.nature.com/scientificreports www.nature.com/scientificreports/ invertebrate glycoproteins that induce widespread IgE reactivity and interfere with in vitro tests for allergy 30,31 . CCD-specific IgE cross-reacts with IgEs to other aeroallergens, such as pollens and Hymenoptera venom 32 ; its level is reportedly elevated in heavy drinkers 33 . Therefore, an increased total or specific IgE level alone is not indicative of allergy in drinkers. However, a recent study involving a large population by Vidal et al., predominantly assessing mite allergy patients, showed that CCD sensitization had no significant association with age, residential area, alcohol consumption, or smoking, and that CCD sensitization minimally interfered with IgE test results 34 . Also, CCD cross-reacts with pollens, latex allergens, Hymenoptera venoms and some food allergens, but to a lesser degree to DF, cockroach, and dog allergens 27,29 . Therefore, IgE associated with CCD likely had little influence on our results.
The mechanisms by which excessive alcohol consumption increases IgE levels are unclear 34 . Alcohol consumption impairs T-helper-type 1 (Th1) immune responses and increases T-helper-type 2 (Th2) responses in human 35,36 and mouse 37,38 . Alcohol intake induces increased intestinal permeability and facilitates access of foreign materials such as microorganisms to the immune system. Alcohol also reduces host defenses and oral tolerance, and induces sensitization 39   www.nature.com/scientificreports www.nature.com/scientificreports/ production by CD4 + T cells 38 . However, Oldenburg et al. demonstrated that alcohol exposure reduced not only airway hyper-responsiveness, but also the serum IgE level and influx of airway eosinophils in a mouse model of allergic asthma 41 . Therefore, alcohol consumption modulates immunity in a general sense, rather than simply inducing a Th2 response.
Few studies have focused on sex differences in the association of alcohol consumption with IgE sensitization to aeroallergens. Sex differences in total IgE or allergen sensitization have in turn been attributed to sex differences in alcohol consumption 36,42 . Linneberg et al. reported that males had a higher alcohol intake and total IgE level than females in a population-based study of Danish adults 36 . Sex differences in immune responses according to alcohol consumption have been investigated using mouse models. Grossman et al. reported that alcohol significantly increased the population of CD4 + T-helper cells in male, but not female, mice 43 . On the other hand, Alonso et al. reported that alcohol administration increased the serum IgE and IL-13, but not IgA or IgM, levels in mice of both sexes 37 .
Female sex hormones stimulate allergic reactions, while little is known about the effects of androgens on the immune system 18,44 . However, the mechanisms underlying sex differences in allergic reactions are poorly understood 44 . Two in vitro studies reported that estradiol induces human mast cells to release allergic mediators and enhances IgE-mediated mast-cell activation and granulation 45,46 . In a rat model of allergic lung inflammation, the estradiol receptor antagonist tamoxifen ameliorated allergic lung inflammation 47 . Importantly, estrogen and estrogenic compounds increase IgE production by mouse splenocytes in vitro, and therefore may affect allergic sensitization 48 . Chronic alcohol use decreases testosterone level and increases estrogen level in males 49 and increases testosterone and estradiol levels in females 50 . Our finding of a positive association between high-risk drinking and increased IgE levels in males may be attributed to increased estrogen levels due to high-risk drinking. Increased estrogen levels in high-risk males may lead to increased IgE production and allergic inflammation. However, because the KNHANES did not gather the data on sex hormones, we could not analyze the association between the serum levels of sex hormones and the serum IgE levels by alcohol consumption and sex. Further studies of sex differences in alcohol consumption and allergic factors, which take into consideration the serum levels of sex hormones, are warranted.
This study had several strengths. First, the KNHANES is a nationwide, population based survey, and the participants are representative of the general Korean population. Second, the serum levels of specific IgE were examined in a relatively large number of subjects. Third, we considered demographic characteristics, socioeconomic status, and comorbidities, and minimized bias by adjusting for confounding effects. Fourth, both subjective (AUDIT questionnaire) and objective (serum GGT level) measures of alcohol consumption were used. Fifth, this is the first study of sex differences in the association between alcohol consumption and IgE sensitization.
This study also had several limitations. First, a cross-sectional study cannot analyze the causal relationships between alcohol consumption and sensitization to aeroallergens. Second, parasite infection, which influences serum IgE levels, was not considered as a confounding factor, because the KNHANES does not include data on parasite infection. Third, despite the relatively large sample size, some of the female AUDIT score subgroups had few subjects (e.g. the number of female participants in the high AUDIT group = 30), resulting in wide confidence intervals. To overcome this problem, we used the sex-specific serum GGT tertiles. Fourth, based on the significant effect of both genetic variants on alcohol metabolism in populations of East Asian 51 and European 52 descent, our findings cannot be generalized to other populations. However, the ethnic homogeneity of the study population may reduce the influence of possible confounding variables. Further studies using data from the general populations of the United States and European nations are needed.
In conclusion, to our knowledge, this is the first study of the associations between alcohol consumption pattern and the serum GGT level with sensitization to specific allergens in the general population stratified by sex. www.nature.com/scientificreports www.nature.com/scientificreports/ Our findings suggest that high-risk alcohol consumption has a greater impact on IgE sensitization in males than in females. Therefore, the risk of IgE sensitization should be considered in patients who indulge in high-risk drinking, particularly in males.

Study population. The Korean National Health and Nutrition Examination Survey (KNHANES) is a
nationwide, population-based, cross-sectional health examination survey, conducted by the Division of Chronic Disease Surveillance under the auspices of the Korea Centers for Disease Control and Prevention (KCDC). The KNHANES is conducted to evaluate the health and nutritional status of non-institutionalized Korean citizens. The survey consists of three components: a health interview, a nutritional survey, and a health examination survey. The health interview and health examination are performed by trained medical staff and interviewers in mobile examination centers. The nutritional survey is conducted in the homes of subjects 1 week after the health interview 53,54 . Peripheral blood samples were collected after fasting for more than eight hours, and were immediately processed and transported to a central laboratory (Neodin Medical Institute, Seoul, South Korea). All blood samples were analyzed within 24 hours of collection 55 .
We used data collected during the 2010 KNHANES. Of the 6,740 adults (≥19 years of age) who participated in the survey, 1,723 completed the clinical examination, the health questionnaire, and measurement of total/ specific IgE (Supplementary Fig. S4). All subjects provided written informed consent for use of their data prior to survey commencement. The study was approved by the Institutional Review Board of the Korea Centers for Disease Control and Prevention (KCDC) (numbers 2010-02CON-21-C) 53 and was performed in accordance with relevant guidelines/regulations. Alcohol consumption. The AUDIT questionnaire comprises three conceptual domains: hazardous alcohol use (frequency of drinking, typical quantity, and frequency of heavy drinking), dependence symptoms (impaired control over drinking, increased salience of drinking, and morning drinking), and harmful alcohol use (guilt after drinking, blackouts, alcohol-related injuries, and other concerns about drinking). The AUDIT scores were categorized into three groups according to the World Health Organization (WHO) guidelines: low risk, 0-7 points; intermediate risk, [8][9][10][11][12][13][14][15] points; and high risk, ≥16 points 22 .
The serum GGT level is a highly sensitive metric of alcohol consumption 56 and the most widely used biomarker of alcohol abuse 23,57 . Serum GGT levels were assayed using a Hitachi Automatic Analyzer 7600 (Hitachi 7600, Tokyo, Japan). Male and female subjects were divided into tertiles based on their serum GGT level. The GGT levels of males were higher than those of females. Thus, sex-specific serum GGT tertiles were used in the analyses. In males, the cut-off points were <25 IU/L (tertile 1), 25 to 43 IU/L (tertile 2), and >43 IU/L (tertile 3). In females, the cut-off points were <14 IU/L (tertile 1), 14 to 18 IU/L (tertile 2), and >18 IU/L (tertile 3).

Measurement of total and specific IgE levels.
Serum samples were tested for total IgE and allergen-specific IgE using a 1470 WIZARD-ɣ-Counter analyzer (PerkinElmer, Turku, Finland) with ImmunoCAP100 (Phadia, Uppsala, Sweden). The Dermatophagoides farinae ([DF] house dust mite)−, Blattella germanica (German cockroach)-and dog-specific IgE levels were measured. Only these three allergen-specific IgEs were included in the KNHANES data. An increased total IgE level was defined as a serum level of ≥150 kU/L 58 . The cut-off value for serum aeroallergen-specific IgE level was 0.35 kU/L 59 .

Definitions of allergic diseases and atopy.
All subjects were asked about their history of allergic diseases.
The lifetime prevalence of allergic diseases, including asthma, allergic rhinitis, and atopic dermatitis, was assessed by a physician. Atopy was defined as a level of IgE against at least one of the three aeroallergens of 0.35 kU/L or greater 60 .

Sociodemographic characteristics and lifestyle factors.
During the health interviews, data were collected using self-reported questionnaires. The health interview evaluated educational level, occupation, household income, residential area, marital status, smoking, exercise status, and sleep duration. Smoking history was categorized as current smoker or nonsmoker. Regular exercise was defined as moderate physical activity performed for at least 30 minutes at a time at least five times a week. Height, weight, and waist circumference were measured by trained medical professionals 53 . The body mass index (BMI) was calculated as body weight (kg) divided by height in meters squared (m 2 ). Obesity was defined as a BMI ≥ 25 kg/m 2 61 . Statistical analysis. Statistical analyses were performed using the complex sample analysis procedure of PASW software (ver. 18.0; SPSS Inc., Chicago, IL, USA) to reflect the complex sampling design and sampling weights of the KNHANES, and to provide nationally representative prevalence estimates. Results are presented as percentages ± standard error for categorical variables and as estimated means ± standard error for continuous variables. To compare categorical and continuous variables, the chi-squared test and a general linear model, respectively, were used. The multivariable analysis employed logistic regression to explore the associations of alcohol consumption with IgE sensitization. We calculated the odds ratios (ORs) for IgE sensitization and atopy according to sex. Multivariable adjustment was for age, residential area, educational level, marital status, household income, occupation, smoking status, exercise status, sleep duration, obesity, vitamin D level and allergic diseases, because these variables are reportedly associated with the serum IgE levels, alcohol consumption, or both 53 . Previous presentation. This study was presented at the European Academy of Allergy and Clinical Immunology (EAACI) Congress 2018; 29 May 2018.