Serum concentrations of interleukin-8 in relation to different levels of alcohol consumption

www.elsevier.com/locate/issn/10434666 Cytokine 38 (2007) 54–60

Serum concentrations of interleukin-8 in relation to different levels of alcohol consumption Arturo Gonzalez-Quintela a,*, Joaquı´n Campos a, Francisco Gude b, Luis-Fernando Perez c, Santiago Tome´ a b

a Department of Internal Medicine, Complejo Hospitalario Universitario de Santiago, Spain Department of Clinical Epidemiology, Complejo Hospitalario Universitario de Santiago, Spain c Department of Biochemistry, Complejo Hospitalario Universitario de Santiago, Spain

Received 5 September 2006; received in revised form 8 February 2007; accepted 9 May 2007

Abstract Serum levels of interleukin-8 (IL-8) are increased in patients with alcoholic hepatitis and correlate with disease severity. The present study was aimed at investigating serum IL-8 levels in relation to different levels of alcohol consumption. Serum IL-8 was measured in (a) 459 individuals randomly selected from the general adult population, including 221 alcohol abstainers, 140 light drinkers (1–140 g/week), 53 moderate drinkers (141–280 g/week), and 45 heavy drinkers (>280 g/week), as well as (b) 137 alcoholics admitted to the hospital. The proportion of individuals with abnormally high (>10 pg/mL) IL-8 levels increased with alcohol use from 5.9% in abstainers to 10.7% in light, 13.2% in moderate, and 17.8% in heavy drinkers (P = 0.004). This proportion was exceedingly high in alcoholics admitted to the hospital (70.1%, P < 0.001 with respect to all other categories). Extremely high (>100 pg/mL) IL-8 levels were only observed among alcoholics, and were more frequent in females than in males (23.5% versus 9.7%, P = 0.03) in spite of lower alcohol consumption among the former. These data indicate that the effect of alcohol on serum IL-8 levels begins with light-to-moderate drinking and is dose-dependent. Females may be more prone than males to develop extremely high IL-8 levels after heavy alcohol intake.  2007 Elsevier Ltd. All rights reserved. Keywords: Alcohol; Alcoholic hepatitis; Interleukin-8; Female sex

1. Introduction Interleukin-8 (IL-8) is a member of the chemokine superfamily, made up of small 8–10 kDa proteins which are included among the pro-inflammatory cytokines [1,2]. The most important function of IL-8 is to act as a chemoattractant and activator of neutrophils [3,4]. Neutrophil infiltration is a hallmark of alcoholic hepatitis [5–7]. In connection with this, a body of evidence suggests that IL-8 is involved in the pathogenesis of alcoholic hepatitis [8]. Serum concentrations of IL-8 are increased in patients with alcoholic hepatitis [9–13] and tend to decrease during recovery [14,15]. Moreover, serum levels of IL-8 correlate

*

Corresponding author. Fax: +34 981 950501. E-mail address: [email protected] (A. Gonzalez-Quintela).

1043-4666/$ - see front matter  2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.cyto.2007.05.006

with the severity of alcoholic hepatitis [9,12,13,15–17]. Furthermore, high serum levels of IL-8 during alcoholic hepatitis may serve as a predictor of mortality [16]. Serum and tissue levels of IL-8 correlate with liver neutrophil infiltration in patients with alcoholic hepatitis [9,17]. Macrophages (Kupffer cells) are believed to be the major source of IL-8, but many other cell types, including hepatocytes can also produce IL-8 [18–20]. The release of IL-8 by these cells can be initiated by a wide variety of stimuli including gut-derived endotoxin and TNF-a [7,21]. Ethanol can also promote the synthesis of IL-8 in hepatocytes [19,20]. Stimulated blood mononuclear cells from heavy drinkers produce more IL-8 than those from healthy controls [22]. Most studies on serum IL-8 alterations in relationship to alcohol consumption have been performed in selected samples of alcoholics [9–17]. Also, Latvala et al. studied 14 moderate drinkers and observed that their serum IL-8

A. Gonzalez-Quintela et al. / Cytokine 38 (2007) 54–60

levels were lower than those of alcoholics with or without liver disease [23]. Thus, the possible influence of light-tomoderate alcohol use on serum levels of IL-8 has not been fully investigated. Furthermore, there are no previous reports of the distribution of serum IL-8 levels in the normal adult population. The objective of the present study was to investigate serum IL-8 levels in a large series of subjects with different levels of alcohol intake randomly selected from the general adult population. Their serum IL-8 levels were further compared with those of a series of alcoholic patients admitted to the hospital. 2. Methods

55

Department [25]. Serum samples for cytokine determinations were available for 137 individuals. The main reasons for admission were alcohol withdrawal syndrome (67 patients), complications of advanced alcoholic liver disease (35 cases), general symptoms plus liver biochemical abnormalities (19 patients), pneumonia (7 patients), acute pancreatitis (6 patients), and miscellaneous causes in the three remaining patients. The main characteristics of individuals in this group are also shown in Table 1. All patients from this group were very heavy drinkers with biochemical signs of acute liver damage (Table 1). Therefore, they were considered as a separate group. 2.2. Main determinations

2.1. Study population The total study population (n = 596) derives from two previous cross-sectional surveys in the area served by a University Hospital in northwest Spain: namely, a sample of the general population and a series of alcoholics admitted to the hospital. Detailed descriptions of these surveys have been reported elsewhere [24,25]. All individuals consented to participate. The first survey consisted of a random sample of 469 individuals in the adult population of the municipality of A-Estrada [24]. Serum samples for cytokine determinations were unavailable for nine individuals. One additional individual was excluded because he was later admitted to the hospital and was therefore included in the second group. Thus, the general population group included 459 individuals. The median age was 54 years (range 18–92 years). A total of 202 individuals (44.0%) were males. According to their habitual alcohol consumption (please see Section 2.2.1), 221 individuals (48.1%) were classified as alcohol abstainers, 140 (30.5%) as light drinkers, 53 (11.5%) as moderate drinkers, and 45 (9.8%) were classified as heavy drinkers. The main characteristics of individuals in each of these categories are shown in Table 1. The second survey consisted of a series of 138 alcoholics consecutively admitted to the Internal Medicine

2.2.1. Alcohol consumption Alcohol consumption was evaluated by the system of standard alcohol drinking units [26] which sums the number of glasses of wine (10 g), bottles of beer (10 g), and units of spirits (10 g) regularly consumed per week, as previously described [24,25]. Individuals with habitual alcohol consumption between 1 and 140 g/week were considered light drinkers, those with alcohol consumption between 141 and 280 g/week were considered moderate drinkers, and those with alcohol consumption higher than 280 g/week were considered heavy drinkers. 2.2.2. Serum IL-8 Sera for IL-8 determinations were stored frozen until testing by chemiluminescent enzyme immunoassay (Immulite, DPC, LA, CA, USA). The lower threshold for detection of IL-8 with this method is 5 pg/mL. The normal upper reference concentration is not established. For the present study, serum levels >10 pg/mL were considered abnormally high. This level approximately corresponds to the 95th percentile in abstainers from the studied population. 2.2.3. Routine tests Routine liver parameters were determined from fresh serum samples. The Advia 1650 analyzer (Bayer

Table 1 Characteristics of individuals included in the study classified according to alcohol consumption Group

No.

Sex (Male)

Age (years)

Alcohol intake (g/week)

Serum AST (U/L)a

Serum GGT (U/L)a

Serum bilirubin (mg/dL)

General population Alcohol abstainers Light drinkers Moderate drinkers Heavy drinkers Alcoholic patients admitted to the hospital

221 140 53 45 137

61 55 41 45 103

51 (33–69) 55 (41–74) 58 (43–73) 50 (41–60) 46 (39–57)

0 70 210 490 840

0.54 (0.43–0.65) 0.54 (0.46–0.65) 0.57 (0.49–0.68) 0.73 (0.54–0.96) 1.88 (1.28–3.24)

0.31 0.38 0.51 0.81 4.68

0.57 0.58 0.60 0.70 1.25

(27.6) (39.3) (77.4) (100.0) (75.2)

(68–140) (210–280) (420–560) (700–1120)

(0.22–0.45) (0.28–0.56) (0.41–0.67) (0.54–1.66) (2.23–10.4)

Figures are absolute numbers and proportions (within parentheses) or median and interquartile ranges (within parentheses). AST, aspartate aminotransferase. GGT, c-glutamyl transferase. a Expressed as the ratio to the normal upper reference level.

(0.43–0.75) (0.45–0.75) (0.51–0.69) (0.46–0.84) (0.71–2.1)

A. Gonzalez-Quintela et al. / Cytokine 38 (2007) 54–60 300

Median: 5.7 pg/mL 200

Interquartile range: <5-7.5 pg/mL Range: <5-57.5 pg/mL

100

0 05 5 >1 - 10 1 10 97 93 9 -8 85 81 77 3 -7 69 65 61 7 -5 53 49 45 41 37 33 29 5 -2 21 17 13 -9 5 <5

Diagnostics, Leverkusen, Germany) and the Olympus AU-400 analyzer (Olympus, Tokyo, Japan) were employed for individuals admitted to the hospital and individuals from the general population, respectively. In order to standardize serum transaminase values in analyses that included both groups of individuals, these were expressed as the ratio of the observed value to the corresponding normal upper reference value. The discriminant function developed by Maddrey et al. (4.6 · prothrombin time prolongation in seconds + serum bilirubin in mg/dL) was used to assess the clinical severity of liver disease in alcoholics [27]. Values above 32 were considered indicative of severe alcoholic hepatitis [27]. Serum C-reactive protein levels were measured by rate nephelometry (Beckman Coulter, Fullerton, CA, USA) as previously described [25].

Number of observations

56

2.3. Statistical analyses

3.1. General descriptive statistics of IL-8 concentrations Serum IL-8 concentrations in the whole series ranged from undetectable (<5 pg/mL) to 13,116 pg/mL. The histograms of serum IL-8 levels in subjects from the general adult population and in alcoholics admitted to the hospital are shown in Fig. 1. Serum IL-8 levels were higher in alcoholics than in subjects from the general adult population as a whole (P < 0.001).

Median: 17.8 pg/mL 20

Interquartile range: 8.9-62.1 pg/mL Range: <5-13 116 pg/mL

10

0 05 5 >1 - 10 1 10 97 93 9 -8 85 81 77 73 69 65 61 57 53 9 -4 45 41 37 33 29 5 -2 21 17 13 -9 5 <5

3. Results

Number of observations

30

Interleukin-8 concentrations were considered in categories because a number of individuals presented with undetectable levels and the distribution of detectable values was highly skewed to the right. The v2 test (with analysis of trend, when appropriate) was employed to compare proportions. The Spearman rank test was used to assess correlation. The Mann–Whitney test was used to compare numeric variables between groups. The Jonkheere–Terpstra procedure was used to test dose– response trend of numeric variables in relation to independent variables in ordered categories. Undetectable IL-8 values were considered as zero for all these nonparametric tests. In order to study the relationship between alcohol consumption and high (>10 pg/mL) IL-8 levels after adjusting for age and sex, a generalized additive logistic model was fitted to the binary response ‘high/not-high’ IL-8 concentrations, with smooth terms for alcohol consumption in g/week, a linear term for age, and a binary variable for sex [28]. Once the optimum degrees of freedom for the model were selected, the point wise estimates of odds ratios, their associated confidence intervals, and a graphical output (non-parametric dose–response curve) were obtained using the S-Plus software (MathSoft Inc., Seattle, WA) following the method described by Saez et al. [29].

Serum IL-8 (pg/mL) Fig. 1. Histogram of IL-8 concentrations. (a) Individuals from the general adult population. (b) Alcoholic patients admitted to the hospital.

3.2. Serum concentrations of IL-8 in relationship to different levels of alcohol intake The lowest proportion of individuals with detectable (>5 pg/mL) IL-8 levels was found in alcohol abstainers (61.5%), whereas the highest proportion of individuals with detectable values was observed in alcoholics admitted to the hospital (94.9%) (Fig. 2). Moreover, the majority (70.1%) of alcoholics admitted to the hospital showed high (>10 pg/mL) IL-8 levels (Fig. 2). This proportion was higher than that of any of the categories of alcohol consumption in individuals from the general population (P < 0.001 for all comparisons). Among the latter, however, the proportion of individuals with high IL-8 levels increased in parallel with alcohol consumption, from 5.9% in abstainers to 10.7% in light drinkers, 13.2% in moderate drinkers, and 17.8% in heavy drinkers (P for trend, 0.004) (Fig. 2).

A. Gonzalez-Quintela et al. / Cytokine 38 (2007) 54–60

90

Detectable IL-8

80

High IL-8

Proportion of subjects (%)

100

70 60 50 40 30 20

with high (>10 pg/mL) IL-8 levels

Proportion of subjects (%)

100

57

90

Age <50 years

80

Age ≥50 years

70 60 50 40 30 20 10

10

0

0

C

D

A

E

Fig. 2. Proportion of individuals with detectable and high levels serum levels of IL-8 in the different categories of alcohol consumption (A, abstainers; B, light drinkers; C, moderate drinkers; D, heavy drinkers; E, alcoholic patients admitted to the hospital). Empty bars represent the proportion of individuals with detectable concentrations (P5 pg/mL). Solid bars represent the proportion of individuals with high (>10 pg/mL) serum concentrations of IL-8. Both proportions tend to increase as the level of alcohol consumption increases (P < 0.001).

Age was associated with high IL-8 levels. Individuals older than 50 years had a higher proportion of increased IL-8 levels than their younger counterparts in all categories of alcohol consumption (Fig. 3). Female sex tended to be associated with higher IL-8 levels (Fig. 3). The association between light-to-moderate drinking and high IL-8 levels was still present after adjusting for age and sex (Fig. 4). In relationship to the reference category (abstainers), the age-and-sex-adjusted risk (log OR) of having high serum IL-8 levels increased from the lowest level of alcohol consumption in a nearly linear fashion. The risk was significant (P < 0.05) from 500 g/week, and a sharp increase was observed from 1000 g/week onwards (Fig. 4). Extremely high (>100 pg/mL) IL-8 levels were observed only among patients admitted to the hospital, and occurred more frequently in females than in males (8/34 cases, 23.5%, versus 10/93 cases, 9.7%, respectively, P = 0.03), in spite of lower alcohol consumption in females than in males (median 770 g/week versus median 840 g/week, respectively, P = 0.003). The female and male alcoholics were similar in age (median 46 years in both sexes) and had a similar prevalence of advanced liver disease (23.5% versus 26.2%, respectively). Individuals with advanced liver disease showed a higher proportion of increased serum IL-8 levels than patients without it (33/35 cases, 94.3%, versus 63/102 cases, 61.8%, respectively, P < 0.001). 3.3. Correlation between serum IL-8 and routine tests Among alcoholics admitted to the hospital, there was a positive correlation between serum IL-8 levels and markers

B

C

D

E

100 90

Proportion of subjects (%)

B

with high (>10 pg/mL) IL-8 levels

A

Males

80

Females

70 60 50 40 30 20 10 0

A

B

C

E

Fig. 3. Age and sex differences in the proportion of high (>10 pg/mL) serum levels of IL-8 in relation to alcohol consumption (A, abstainers; B, light drinkers; C, moderate drinkers; D, heavy drinkers; E, alcoholic patients admitted to the hospital). There were no females in the group of heavy drinkers from the general population. The proportion of individuals with high IL-8 levels increases as the level of alcohol consumption increases both in individuals <50 years (P < 0.001) and individuals P50 years (P < 0.001) with a trend towards a higher proportion among the latter (superior panel). The proportion of individuals with high IL-8 levels increases as the level of alcohol consumption increases in both males (P < 0.001) and females (P < 0.001), with a trend towards a higher proportion among the latter (inferior panel).

of alcoholic hepatitis, such as serum aspartate aminotransferase (AST) (R = 0.278, P = 0.001) and especially with serum the serum AST/ALT (alanine aminotransferase) ratio (r = 0.535, P < 0.001) and serum bilirubin (R = 0.525, P < 0.001) (Fig. 5). All 31 patients with serum bilirubin P2.5 mg/dL showed high (>10 pg/mL) serum IL-8 levels, and 15 (48.4%) showed extremely high (>100 pg/mL) serum IL-8 levels (Fig. 5). The nine patients with severe alcoholic hepatitis (as defined by a Maddrey discriminant function P32) had significantly higher serum IL-8 levels than patients with lower discriminant function

30

0.2

0.4

0.6

0.8

Serum bilirubin (mg/dL)

1.2

35

1.0

A. Gonzalez-Quintela et al. / Cytokine 38 (2007) 54–60

0.0

Log OR of high (>10 pg/mL) serum IL-8 levels

58

25 20 15 10 5 0

0

500

1000

1500

Alcohol intake (g/week)

values (median 109 pg/mL versus median 17.1 pg/mL, respectively, P = 0.008). Only one patient in this series (a 74-year-old male with a Maddrey discriminant function of 46 and a serum IL-8 value of 65 pg/mL) died during hospital admission. Serum IL-8 levels were weakly correlated with serum levels of C-reactive protein (R = 0.158, P = 0.06) and were not correlated with white blood cell counts (R = 0.084, P = 0.33). The correlation between IL-8 levels with serum c-glutamyl transpeptidase (GGT) was also low (R = 0.181, P = 0.03), and serum IL-8 tended to be negatively correlated with serum ALT levels (R = 169, P = 0.04). Among individuals from the general adult population, there was some correlation between serum IL-8 and serum AST (R = 0.141, P = 0.002), ALT (R = 0.156, P = 0.001), and GGT (R = 0.201, P < 0.001). There was also some between serum IL-8 levels and the body mass index in kg/m2 (R = 0.120, P = 0.01). There was no correlation between serum IL-8 and serum bilirubin in this group (R = 0.014, P = 0.76). 4. Discussion The present study shows that the effect of alcohol on serum IL-8 levels begins with light-to-moderate levels of drinking and is dose-dependent. The proportion of individuals with high (>10 pg/mL) IL-8 levels gradually increases as the level of alcohol consumption increases, from 5.9% in alcohol abstainers to 17.8% in heavy drinkers. Importantly, these data derive from a population-based survey. This phenomenon was not described previously, but may be in agreement with previous reports showing that serum IL-8 may increase shortly after a single alcohol intake in healthy volunteers [30]. Elevation of IL-8 in light-to-moderate drinkers supports the notion that the systemic inflamma-

7 <5

34 5-10

78 10.1-100

18 >100

7 <5

34 5-10

78 10.1-100

18 >100

8

Ratio sAST/sALT (U/L)

Fig. 4. Non-parametric dose–response curve of the age-and-sex-adjusted odds ratios (transformed into their natural logarithm) of high (>10 pg/mL) serum IL-8 levels in relationship to alcohol consumption among individuals from the general adult population. Dashed curves represent the 95% confidence limits. The horizontal line represents the null hypothesis (no effect, OR = 1, log OR = 0). The vertical line represents the reference category (alcohol abstainers).

N=

6

4

2

0 N=

Serum IL-8 (pg/mL) Fig. 5. Box plot of serum bilirubin level and serum aspartate aminotransferase (sAST)/serum alanine aminotransferase (sALT) ratios in alcoholic patients admitted to the hospital, classified according to their serum IL-8 concentration. There is a trend towards increasing serum bilirubin levels (superior panel) and a trend towards increasing sAST/ sALT ratios (inferior panel) as serum IL-8 concentrations increase (P < 0.001 for both comparisons).

tory effects of alcohol intake may begin at lower doses than previously considered. It should be noted that IL-8 is a crucial mediator in neutrophil-dependent acute inflammation and is able to activate a wide range of signalling molecules in cells other than neutrophils [4]. From the clinical standpoint, the possible significance of such an IL-8 elevation, however, remains to be elucidated. From the diagnostic point of view, the effect of light-to-moderate drinking on serum IL-8 could be of importance because IL-8 is widely used as a marker for a variety of inflammatory conditions. The effect of alcohol on serum IL-8 may be modest at lightto-moderate levels of drinking, but should be taken into account when interpreting serum IL-8 values. Alcoholics admitted to the hospital showed the highest serum IL-8 levels. Serum IL-8 in alcoholics frequently exceeded 100 pg/mL, a concentration observed in none of subjects from the general population, even in the heavy drinkers. Serum IL-8 levels in alcoholics admitted to the

A. Gonzalez-Quintela et al. / Cytokine 38 (2007) 54–60

hospital were correlated with markers of alcoholic hepatitis, particularly serum bilirubin and the serum AST/ALT ratio. Furthermore, the highest IL-8 levels were found among alcoholics with severe alcoholic hepatitis as reflected by a high Maddrey’s score [27]. This is agreement with previous reports of a correlation between serum IL-8 levels and markers of acute liver damage [9,12,16,23], histological liver injury [9,16,17], or prognosis [9,12,16] of alcoholic hepatitis. Taken together, these studies are consistent with a role for the innate immune system in the development of alcoholic liver disease [7]. Specifically, these data support the notion of a role for IL-8 as a crucial mediator in the pathogenesis of alcoholic hepatitis. Extremely high (>100 pg/mL) concentrations were more common in alcoholic females than in males. It is well-known that alcoholic liver injury is more severe, and develops more rapidly in women than men [31–33]. Specifically, females are particularly prone to alcoholic hepatitis [31,33]. The basis of this susceptibility is not entirely known. Ethanol-fed female rats show higher endotoxemia [34,35], higher production of some chemokines [35], and more severe liver injury than male rats [34,35]. This may be in agreement with the presented results, which showed that alcoholic females develop higher serum IL-8 levels than alcoholic males in spite of a lower alcohol intake. This phenomenon was not reported in previous studies, which included fewer female patients [9,10,15,16] or only males [14,23]. Further studies are needed to elucidate the possible differences in chemokine production in order to explain sex susceptibility to liver disease after heavy alcohol drinking. Acknowledgments This study was supported by Grants from the Spanish Fondo de Investigaciones Sanitarias (1306/99 and PI040920), Xunta de Galicia (PGIDIT03SAN91807R) and the SEAIC (Sociedad Espan˜ola de Alergologia e Inmunologia Clinica). J.C. was supported by a grant from the Fundacion IDICHUS (Investigacion y Desarrollo, Complejo Hospitalario Universitario de Santiago). References [1] Baggiolini M, Loetscher P, Moser B. Interleukin-8 and the chemokine family. Int J Immunopharmacol 1995;17:103–8. [2] Prieschl EE, Klumburg PA, Baumruker T. The nomenclature of chemokines. Int Arch Allergy Immunol 1995;107:475–83. [3] Baggiolini M. Chemokines in pathology and medicine. J Intern Med 2001;250:91–104. [4] Mukaida N. Interleukin-8: an expanding universe beyond neutrophil chemotaxis and activation. Int J Hematol 2000;72:391–8. [5] Bautista AP. Neutrophilic infiltration in alcoholic hepatitis. Alcohol 2002;27:17–21. [6] Jaeschke H. Neutrophil-mediated tissue injury in alcoholic hepatitis. Alcohol 2002;27:23–7. [7] Nagy LE. Recent insights into the role of the innate immune system in the development of alcoholic liver disease. Exp Biol Med (Maywood) 2003;228:882–90.

59

[8] Bird G. Interleukin-8 in alcoholic liver disease. Acta Gastroenterol Belg 1994;57:255–9. [9] Sheron N, Bird G, Koskinas J, Portmann B, Ceska M, Lindley I, et al. Circulating and tissue levels of the neutrophil chemotaxin interleukin8 are elevated in severe acute alcoholic hepatitis, and tissue levels correlate with neutrophil infiltration. Hepatology 1993;18:41–6. [10] Masumoto T, Onji M, Horiike N, Ohta Y. Assay of serum interleukin-8 levels in patients with alcoholic hepatitis. Alcohol Alcohol 1993(Suppl. 1A):99–102. [11] Gonzalez-Quintela A, Vidal C, Lojo S, Perez LF, Otero-Anton E, Gude F, et al. Serum cytokines and increased total serum IgE in alcoholics. Ann Allergy Asthma Immunol 1999;83:61–7. [12] Fujimoto M, Uemura M, Nakatani Y, Tsujita S, Hoppo K, Tamagawa T, et al. Plasma endotoxin and serum cytokine levels in patients with alcoholic hepatitis: relation to severity of liver disturbance. Alcohol Clin Exp Res 2000;24(Suppl. 4):48S–54S. [13] Fukui H. Relation of endotoxin, endotoxin binding proteins and macrophages to severe alcoholic liver injury and multiple organ failure. Alcohol Clin Exp Res 2005;29(Suppl. 11):172S–9S. [14] Hill DB, Marsano LS, McClain CJ. Increased plasma interleukin-8 concentrations in alcoholic hepatitis. Hepatology 1993;18:576–80. [15] Taieb J, Mathurin P, Elbim C, Cluzel P, Arce-Vicioso M, Bernard B, et al. Blood neutrophil functions and cytokine release in severe alcoholic hepatitis: effect of corticosteroids. J Hepatol 2000;32:579–86. [16] Huang YS, Chan CY, Wu JC, Pai CH, Chao Y, Lee SD. Serum levels of interleukin-8 in alcoholic liver disease: relationship with disease stage, biochemical parameters and survival. J Hepatol 1996;24:377–84. [17] Ishii K, Furudera S, Kumashiro R, Seo J, Koga Y, Sata M, et al. Role of serum interleukin-8 and intercellular adhesion molecule-1 in the severity of alcoholic hepatitis. Alcohol Alcohol Suppl 1994;29:81–5. [18] Bautista AP. Impact of alcohol on the ability of Kupffer cells to produce chemokines and its role in alcoholic liver disease. J Gastroenterol Hepatol 2000;15:349–56. [19] Maher JJ. Rat hepatocytes and Kupffer cells interact to produce interleukin-8 (CINC) in the setting of ethanol. Am J Physiol 1995;269:G518–23. [20] Shiratori Y, Takada H, Hikiba Y, Nakata R, Okano K, Komatsu Y, et al. Production of chemotactic factor, interleukin-8, from hepatocytes exposed to ethanol. Hepatology 1993;18:1477–82. [21] Van Deventer SJ, Hart M, van der Poll T, Hack CE, Aarden LA. Endotoxin and tumor necrosis factor-alpha-induced interleukin-8 release in humans. J Infect Dis 1993;167:461–4. [22] Gonzalez-Quintela A, Dominguez-Santalla MJ, Loidi L, Quinteiro C, Perez LF. Relation of tumor necrosis factor (TNF) gene polymorphisms with serum concentrations and in vitro production of TNFalpha and interleukin-8 in heavy drinkers. Alcohol 2004;34:273–7. [23] Latvala J, Hietala J, Koivisto H, Jarvi K, Anttila P, Niemela O. Immune responses to ethanol metabolites and cytokine profiles differentiate alcoholics with or without liver disease. Am J Gastroenterol 2005;100:1303–10. [24] Gonzalez-Quintela A, Gude F, Boquete O, Rey J, Meijide LM, Suarez F, et al. Association of alcohol consumption with total serum immunoglobulin E levels and allergic sensitization in an adult population-based survey. Clin Exp Allergy 2003;33:199–205. [25] Campos J, Gonzalez-Quintela A, Quinteiro C, Gude F, Perez LF, Torre JA, et al. The -159C(T polymorphism in the promoter region of the CD14 gene is associated with advanced liver disease and serum levels of acute-phase proteins in heavy drinkers. Alcohol Clin Exp Res 2005;29:1206–13. [26] Gual A, Martos AR, Lligona A, Llopis JJ. Does the concept of a standard drink apply to viticultural societies?. Alcohol Alcohol 1999;34:153–60. [27] Maddrey WC, Boitnott JK, Bedine MS, Weber FL, Mezey E, White RI. Corticosteroid therapy of alcoholic hepatitis. Gastroenterology 1978;75:193–9.

60

A. Gonzalez-Quintela et al. / Cytokine 38 (2007) 54–60

[28] Hastie T, Tibshirani R. Generalized additive models. London: Chapman and Hall; 1990. [29] Saez M, Cadarso-Suarez C, Figueiras A. Np.OR: an S-Plus function for pointwise nonparametric estimation of odds-ratios of continuous predictors. Comput Methods Programs Biomed 2003;71:175–9. [30] Gonzalez-Quintela A, Dominguez-Santalla MJ, Perez LF, Vidal C, Lojo S, Barrio E. Influence of acute alcohol intake and alcohol withdrawal on circulating levels of IL-6, IL-8, IL-10 and IL-12. Cytokine 2000;12:1437–40. [31] Pares A, Caballeria J, Bruguera M, Torres M, Rodes J. Histological course of alcoholic hepatitis. Influence of abstinence, sex and extent of hepatic damage. J Hepatol 1986;2:33–42. [32] Becker U, Deis A, Sorensen TI, Gronbaek M, Borch-Johnsen K, Muller CF, et al. Prediction of risk of liver disease by alcohol intake,

sex, and age: a prospective population study. Hepatology 1996;23:1025–9. [33] Barrio E, Tome S, Rodriguez I, Gude F, Sanchez-Leira J, PerezBecerra E, et al. Liver disease in heavy drinkers with and without alcohol withdrawal syndrome. Alcohol Clin Exp Res 2004;28:131–6. [34] Kono H, Wheeler MD, Rusyn I, Lin M, Seabra V, Rivera CA, et al. Gender differences in early alcohol-induced liver injury: role of CD14, NF-kappaB, and TNF-alpha. Am J Physiol Gastrointest Liver Physiol 2000;278:G652–61. [35] Nanji AA, Jokelainen K, Fotouhinia M, Rahemtulla A, Thomas P, Tipoe GL, et al. Increased severity of alcoholic liver injury in female rats: role of oxidative stress, endotoxin, and chemokines. Am J Physiol Gastrointest Liver Physiol 2001;281:G1348–56.