- Email: [email protected]
Alcohol and coffee drinking and smoking habit among subjects with HCV infection
Digestive and Liver Disease 41 (2009) 599–604
Contents lists available at ScienceDirect
Digestive and Liver Disease journal homepage: www.elsevier.com/locate/dld
Liver, Pancreas and Biliary Tract
Alcohol and coffee drinking and smoking habit among subjects with HCV infection C. Zani a , F. Donato a,∗ , M. Chiesa a , C. Baiguera b , U. Gelatti a , L. Covolo a , M.G. Antonini b , P. Nasta b , F. Gatti b , G. Orizio a , M. Puoti b a b
Department of Experimental and Applied Medicine, Institute of Hygiene, Epidemiology and Public Health, University of Brescia, Brescia, Italy Liver Unit of the Infectious Diseases Department, Spedali Civili, Brescia, Italy
a r t i c l e
i n f o
Article history: Received 28 July 2008 Accepted 20 November 2008 Available online 29 January 2009 Keywords: Alcohol drinking Coffee drinking HCV infection Tobacco smoking
a b s t r a c t Background/aims: The aims were to estimate among patients with hepatitis C virus (HCV) infection the prevalence of alcohol and coffee intake and smoking habit, the reliability of these self-reported data and the possible change of patients’ habit after their first contact with a Viral Hepatitis Service. Methods: 229 patients were initially interviewed personally at the Viral Hepatitis Service and after 6 months they were re-interviewed by phone in regard to their alcohol, coffee drinking and smoking habits. Results: Alcohol drinkers were 55.5% of males and 35.3% of females. Most subjects drank coffee daily, both men (90.0%) and women (84.9%). The proportion of current smokers was higher in males (43.6%) than females (26.9%). We found a fair to good reliability of self-reported data regarding patients’ habits, alcohol and coffee intake, and number of cigarettes smoked daily. We observed a statistically significant decrease in alcohol and coffee intake and cigarettes smoked between baseline and follow-up interviews. Conclusion: We found a fairly high proportion of HCV-infected patients who regularly drink alcohol and coffee beverages and smoke cigarettes, especially among males. The reliability of self-reported data on these habits seems satisfactory. More decisive action to modify these habits, especially alcohol intake, is required in these patients. © 2009 Editrice Gastroenterologica Italiana S.r.l. Published by Elsevier Ltd. All rights reserved.
1. Introduction Some common lifestyle habits can influence the development of liver disease. Alcohol drinking is a well-known cause of chronic liver disease [1–3] and a dose–effect relationship between alcohol intake and the risk of developing a clinically evident liver disease is reported. Tobacco smoking is a biologically plausible risk factor for liver cancer, especially among people with other risk factors for chronic liver disease, and some findings support this hypothesis [4]. On the other hand, coffee drinking may play a protective role on the liver, since an inverse relation between coffee drinking and cirrhosis or hepatocellular carcinoma (HCC) has been found in some studies [5–7]. Many subjects with hepatitis C virus (HCV) infection have a progression of chronic hepatitis C towards cirrhosis and HCC. Various viral and host factors can influence the progression of the hepatic disease [8], among them, the effect of alcohol drinking on
∗ Corresponding author at: Sezione di Igiene, Epidemiologia e Sanità Pubblica, Università di Brescia, Facoltà di Medicina e Chirurgia, Viale Europa 11, 25123 Brescia, Italy. Tel.: +39 030 3838602; fax: +39 030 370 1404. E-mail address: [email protected] (F. Donato).
progression of HCV-related disease has been investigated widely [9,10]. An interaction between alcohol intake and HCV infection in causing cirrhosis and HCC has been found [3,11,12]. Furthermore, alcohol drinking can influence the response to HCV treatment [13]. Also a possible effect of tobacco smoking in increasing, and of coffee drinking in decreasing, the risk of the progression of liver damage towards cirrhosis have been shown in some studies [5,14,15]. In spite of the role that alcohol and coffee drinking and smoking habits may play in the development of chronic liver disease due to HCV infection, few studies are available on the epidemiologic pattern of these habits in HCV-infected individuals [16]. Furthermore, as data on lifestyle habits are usually collected through interviews, their reliability is questionable, but as far as we know these behaviours have not been investigated among patients with HCV infection outside the hospital setting. The aims of this study were to investigate among patients with HCV infection: (a) the prevalence of alcohol and coffee intake and smoking habit, (b) the reliability of self-reported data on these habits and (c) the possible change of patients’ habits between before and after their first contact with a Viral Hepatitis Service.
1590-8658/$36.00 © 2009 Editrice Gastroenterologica Italiana S.r.l. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.dld.2008.11.018
600
C. Zani et al. / Digestive and Liver Disease 41 (2009) 599–604
2. Methods 2.1. Study design All subjects were outpatients admitted for the first time due to anti-HCV positivity to the Viral Hepatitis Service of the Infectious Diseases Department of the “Spedali Civili” Hospital in Brescia, North Italy, between March 2006 and December 2007. They underwent two interviews using the same questionnaire: the first was performed at the first contact with the Viral Hepatitis Service and the second on average 6 months later (range 5–12 months). First, all participants were interviewed face-to-face during their visit by a trained interviewer, and they were told they would be recalled after 6 months. In accordance with international guidelines [17], physicians advised patients to abstain totally from drinking alcohol and to stop smoking. Patients who also had a serious disease, particularly a cardiovascular disorder, were advised to moderate their coffee consumption as well. Each subject was required to sign an informed consent form prior to participation. The study was approved by the Ethics Committee of the “Spedali Civili” Hospital. The data were analysed anonymously, in accordance with Italian privacy laws concerning the use of personal health data for research purposes. The 2nd interview was conducted, using the same questionnaire, over the telephone by an interviewer who was unaware of the answers given by each individual at the 1st interview. The subjects were asked about: (a) their habits, at the time of the 1st interview, so as to evaluate the reliability of self-report data, and (b) their habits, at the time of the 2nd interview, so as to detect any changes that had occurred in the interval between the two interviews. Each interview lasted an average of 10 min. 2.2. Population All subjects were outpatients who went to the Viral Hepatitis Service of the Infectious Disease Department for the first time, following their GP’s suggestions, because of anti-HCV positivity. The patients had not received any antiviral treatment before the visit. Among them, 47.2% started antiviral treatment after the visit. They had been referred by their general practitioners because they were either discovered to have an HCV infection or found to have an increase in transaminases or other signs or symptoms suggesting progression of liver disease. Subjects were enrolled according to the following inclusion criteria: born in Italy, aged between 18 and 65 years; not having HCC, other neoplasms, severe systemic diseases such as chronic heart failure, stroke, respiratory insufficiency and chronic renal failure, or severe physical or mental handicaps. Subjects who claimed intravenous drug use were also excluded. 2.3. Questionnaire and interview The questionnaire used to collect information on alcohol and coffee consumption and smoking habit was based on one that some authors used in previous case–control studies on cancer aetiology in this area, after having tested its reliability [11,18]. Alcohol consumption was assessed by asking about the frequency and amount of alcohol consumed in a typical day or week. We considered current drinkers patients who claimed to drink alcoholic beverages at least weekly. We evaluated the total amount of alcohol intake by summing up the number of glasses of wine, cans of beer, shots of aperitif, spirits and coffee with spirits taking account of their mean alcohol content. In accordance with international guidelines, we used different categories of alcohol intake among males and females, with thresholds of 40 and 30 g/day in males and females, respectively, to separate low-moderate from medium-
high intake. Accordingly, we dichotomised current drinkers into subjects with either “low” (1–40 and 1–30 g/day in men and women, respectively) or “high” (>40 and >30 g/day) alcohol consumption. Coffee intake was assessed by summing up the cups of coffee, cappuccino and “coffee and milk” drunk in a typical day. We considered coffee drinkers patients who drank at least one cup of coffee per day. Information on tobacco smoking was also collected. Subjects were classified as current smokers if they smoked at least one cigarette per day. Information provided by patients on their current alcohol and coffee consumption and tobacco smoking at the first contact with the Viral Hepatitis Service was referred to as “before” the first contact; those concerning patients’ current habits at the time of the 2nd interview were referred to as “6–12 months after” the first contact with the Viral Hepatitis Service. To evaluate if they decreased their consumption we asked them about their habits 10 years before, and if they had changed them. If they answered positively we asked the reason for this change. In this way we could evaluate if their GP had given them advice concerning their lifestyle. 2.4. Statistical methods In order to assess reliability, we compared the answers provided by each patient at the 1st interview and those provided at the 2nd interview regarding his or her habits 6–12 months earlier, at the time of the first contact with the Viral Hepatitis Service. Agreement between the two interviews was assessed by computing Cohen’s kappa statistic (K) for dichotomous variables, and the intra-class correlation coefficient (ICC) and Spearman’s rho for ordinal or continuous variables, respectively [19,20]. The ICC was computed using a two-way analysis of variance (ANOVA) with fixed effects [19]. The strength of agreement was evaluated for each of these statistics according to Altman’s suggestions for interpreting K values [21]: poor <0.20, fair 0.21–0.40, moderate 0.41–0.60, good 0.61–0.80 and very good 0.81–1.00. Low K values were not automatically regarded as indicative of non-agreement, however, since this statistic is strongly dependent on true prevalence [22] and may be low when agreement is high due to unbalanced, symmetrical marginals [23]. At the 2nd stage, we assessed the changes that had occurred in each patient’s current habits in the 6-month interval between the 1st and the 2nd interview (before and 6–12 months after the first contact with the Viral Hepatitis Service). To this end we compared each patient’s answers regarding his or her current habits at the time of each interview, using common statistical tests for the comparison of two proportions or means. The study was designed to enroll at least 100 men and 100 women in order to achieve a reliability of 0.5 using the ICC for ordinal variables (alcohol and coffee intake and number of cigarettes smoked per day) assuming a type I error of 5%, a power of 80% and a minimally acceptable level of reliability of 0.3, for each gender separately, using the formula suggested by Walter et al. [24]. The sample size was also large enough to detect a 10 mg/day reduction of alcohol intake in current alcohol drinkers among men, from before to after the first contact with the Viral Hepatitis Service (means of alcohol intake: from 30 to 20 mg/day), given a standard deviation of 30 mg/day, using a Student’s t-test for paired data (alpha = 0.05; power = 90%). Since we expected to lose up to 30% of the patients initially recruited, we enrolled 300 subjects in the study, 50% of whom males. The statistical analyses were performed using STATA software (Stata Statistical Software release 8.0, 2003; Stata Corporation, College Station, TX).
C. Zani et al. / Digestive and Liver Disease 41 (2009) 599–604
601
Table 1 Alcohol and coffee drinking and smoking habits of 229 patients with HCV infection at their first contact with the Viral Hepatitis Service (1st interview). All patients (%) (N = 229)
Males (%) (N = 110)
Females (%) (N = 119)
Alcohol drinking Never drinker Former drinker
81 (35.4) 45 (19.6)
20 (18.2) 29 (26.3)
61 (51.3) 16 (13.4)
Current drinker Low intake High intake
103 (45.0) 72 (69.9) 31 (30.1)
61 (55.5)a 43 (70.5) 18 (29.5)
42 (35.3)a 29 (69.0) 13 (31.0)
Coffee drinking Never drinker Former drinker
25 (10.9) 4 (1.8)
8 (7.3) 3 (2.7)
17 (14.3) 1 (0.8)
200 (87.3) 80 (40.0) 80 (40.0) 40 (20.0)
99 (90.0) 27 (27.3) 40 (40.4) 32 (32.3)b
101 (84.9) 53 (52.5) 40 (39.6) 8(7.9)b
112 (48.9) 37 (16.2)
35 (31.8) 27 (24.6)
77 (64.7) 10 (8.4)
80 (34.9) 27 (33.7) 44 (55.0) 9 (11.3)
48 (43.6)c 11 (22.9) 29 (60.4) 8 (16.7)
32 (26.9)c 16 (50.0) 15 (46.9) 1 (3.1)
Current drinker 1–2 cups of coffee/day 3–4 cups of coffee/day 5+ cups of coffee/day Smoking habit Never smoker Former smoker Current smoker 1–10 cigarettes/day 11–20 cigarettes/day > 20 cigarettes/day
Percentages are computed on the column total. However, percentages of alcohol drinkers according to ethanol intake categories, of coffee drinkers according to coffee intake categories and of smokers according to cigarettes smoked categories are computed on the total number of current users, respectively. Alcohol drinking: low intake ≤ 30 g alcohol/day for females and ≤40 g alcohol/day for males; high intake: >30 g alcohol/day for females and >40 g alcohol/day for males. a p = 0.002 by the Fisher’s exact test. b p < 0.001 by the Fisher’s exact test. c p = 0.008 by the Fisher’s exact test.
3. Results A total of 301 subjects were enrolled in the study and provided a valid 1st interview on their habits regarding alcohol and coffee drinking and tobacco smoking. Of these, 229 (76.0 %) were re-interviewed after 6–12 months and 72 subjects were phoned at least three times but could not be found or refused the interview. Among them 48.0% were males and the mean age was 51.8 ± 11.9 years (range 22–77 years) (data not shown in tables). All patients had a diagnosis of chronic hepatitis, 140 (61.2%) had elevated and 89 (38.8%) normal serum transaminases (data not shown in the table). Alcohol and coffee drinking and smoking habits of patients at their first contact with the Viral Hepatitis Service are reported in Table 1. Males drank more alcohol than females: current drinkers were 55.5 and 35.3%, respectively (p = 0.002). According to the different categorizations of alcohol intake for males and females, among current drinkers 29.5% males drank >40 g and 31.0% females drank >30 g/day of ethanol per day. Most subjects drank coffee daily, with similar percentages of current
drinkers in males and females (90.0 and 84.9%), although the proportion of drinkers of 5+ cups/day was higher in men than women (32.3 and 7.9%, respectively; p < 0.001). A higher proportion of current smokers was found in males than females (43.6 and 26.9%, respectively; p = 0.008), and among current smokers 16.7% of males and 3.1% of females smoked >20 cigarettes per day (p = 0.02). The 229 subjects who were re-interviewed were compared with the 72 subjects who were not: no significant difference was found between them and subjects re-interviewed as regards their alcohol consumption, coffee drinking and smoking habit at the 1st interview. However, a higher proportion of males was found among those lost (80.5%) compared to those with a complete follow-up (48%), and their mean age was 46.5 (S.D.: 14.5) compared to 51.8 (S.D.: 11.9) (data not shown in tables). Reliability of alcohol intake (Table 2) was fair to moderate as regards both the habit (current/former/never drinker), with a kappa of 0.58, and the total alcohol intake, categorized at three levels (ICC = 0.42) or considered as a continuous variable (grams of ethanol per day) (Spearman’s rho = 0.62).
Table 2 Test–retest reliability study investigating alcohol drinking habits and alcohol intake at the time of patients’ first contact with the Hepatitis Service, comparing the answers given at 1st and 2nd interview (6–12 months later). 1st interview
2nd interview
Alcohol drinking habits Current drinker Former drinker Never drinker
Current drinker 79 12 11
Alcohol intake (g/day) 0 Low High
0 103 20 4
Alcohol drinking habits Former drinker 9 23 4 Low 15 44 4
Never drinker 15 10 66
Alcohol intake (g/day) High 8 8 23
Agreement
K /ICC/Ra (95% CI)
73.4%
K = 0.58 (0.47–0.66)
ICC = 0.42 (0.23–0.46) 74.2% R = 0.62 (0.54–0.70)
The ICC was computed using the three categories ordinal variable whereas Spearman’s rho was computed using the continuous variable (g ethanol intake per day). a K = Cohen’s kappa; ICC = intraclass correlation coefficient; R = Spearman’s rho.
602
C. Zani et al. / Digestive and Liver Disease 41 (2009) 599–604
Table 3 Test–retest reliability study investigating coffee drinking habits and number of cups of coffee drunk per day at the time of patients’ first contact with the Hepatitis Service, comparing the answers given at 1st and 2nd interview (6–12 months after the former). 1st interview
2nd interview
Coffee drinking habits Current drinker Former and never drinker
Current drinker 186 3
Coffee intake (cups/day) 0 1–2 3–4 5+
0 49 25 1 2
Coffee drinking habits Former and never 14 26
1–2 9 55 15 11
Coffee intake (cups/day) 3–4 0 14 13 13
5+ 0 4 4 14
Agreement
K/ICC/Ra (95% CI)
92.5%
K = 0.71 (0.59–0.82)
ICC = 0.64 (0.53–0.72) R = 0.72 (0.65–0.78)
57.2%
The ICC was computed using the four categories ordinal variable whereas Spearman’s rho was computed using the continuous variable (number of cups of coffee per day). a K = Cohen’s kappa; ICC = intraclass correlation coefficient; R = Spearman’s rho. Table 4 Test–retest reliability study investigating smoking habits and number of cigarettes smoked per day at the time of patients’ first contact with the Hepatitis Service, comparing the answers given at 1st and 2nd interview (6–12 months after the former). 1st interview
2nd interview
Smoking habits Current smoker Former smoker Never smoker
Smoking habits Current smoker 68 5 2
No. cigarettes/day 0 1–10 11–20 >20
0 146 13 22 2
Former smoker 4 30 3
1–10 3 11 1 0
Never smoker 8 2 107 No. cigarettes/day 11–20 0 3 21 2
>20 0 0 0 5
Agreement
K /ICC/Ra (95% CI)
89.5%
K = 0.82 (0.74–0.89)
ICC = 0.64 (0.54–0.70) R = 0.63 (0.55–0.70)
84.6%
The ICC was computed using the four categories ordinal variable whereas Spearman’s rho was computed using the continuous variable (number of cigarettes smoked per day). a K = Cohen’s kappa; ICC = intraclass correlation coefficient; R = Spearman’s rho.
As regards coffee, former and never drinkers were collapsed into one category since there were very few former drinkers. The percentage of agreement regarding coffee drinking habit (Table 3) was 92.5% and the K value was 0.71. A good reliability was also found for coffee intake, with an ICC of 0.64 for the categorical variable, and Spearman’s rho of 0.72 for the continuous variable. Test–retest agreement for smoking habit was 89.5 and the K value was 0.82 (Table 4). A good reliability was found as regards the number of cigarettes smoked per day (agreement: 84.6%; ICC = 0.64; rho = 0.63). We found no association between unreliability of frequency and amount of alcohol and coffee drinking and cigarette smoking (data not shown in tables). Further analyses of agreement between the two interviews were performed for males and females, and subjects aged less than 50 years and those aged equal to or over 50 years, separately (data not shown in table). Compared to females, males showed a lower reliability regarding alcohol consumption (rho = 0.56 and rho = 0.67,
respectively) and smoking habits (rho = 0.58 and rho = 0.70) and no difference regarding coffee intake (rho = 0.70 and rho = 0.71). ICC values were very similar to rho values. Compared to older subjects, those aged <50 years showed a lower reliability for alcohol consumption (rho = 0.53 and rho = 0.68, respectively), smoking habits (rho = 0.53 and rho = 0.77) and coffee drinking (rho = 0.62 and rho = 0.74, respectively). Again, ICCs and rhos provided similar figures. Finally, in order to assess whether the subjects had changed their current habits from before to 6–12 months after the first contact with the Viral Hepatitis Service, we compared subjects’ answers regarding their current habits at each interview (Table 5). No change was found in the proportions of current users of alcohol and coffee and smokers. From before to after contact with the Viral Hepatitis Service, patients reported, among current users, a statistically significant decrease in the geometric means of alcohol intake (from 16.2 to 11.7 g/day, −27.7%), coffee intake (from 3.2 to 2.5 cups/day, −21.8%) and number of cigarettes smoked (from 13.6 to 13.2 per day, −2.9%).
Table 5 Alcohol and coffee drinking and smoking habit among HCV patients before and 6–12 months after the first contact with the Viral Hepatitis Service.
Current drinkers Geometric mean of alcohol intake (S.D.) (g/day)a Current coffee drinkers Geometric mean of number of cups of coffee/day (S.D.)a Current smokers Geometric mean of number cigarettes smoked/day (S.D.)a
Before 1st contact
After 6–12 months
114 16.2 (33.6) 200 3.2 (1.7) 80 13.6 (6.2)
112 11.7 (11.8) 189 2.5 (1.4) 75 13.2 (5.8)
Difference (%) −2 (−1.7) −4.5 (−27.7) −11 (−5.5) −0.7 (−21.8) −5 (−6.5) −0.4 (−2.9)
Pb NS p < 0.0001 NS p < 0.0001 NS p = 0.006
a The geometric means of alcohol and coffee intakes and number of cigarettes smoked were calculated among current users at the 1st and 2nd interview for each habit. The geometric instead of arithmetic mean was used for each variables due to its skewed distribution. b Wilcoxon test for paired data.
C. Zani et al. / Digestive and Liver Disease 41 (2009) 599–604
We investigated the association between ALT serum levels and changes in individuals’ habits from the 1st to the 2nd interview, dichotomising ALT values into normal and high according to the distribution of values in the general population living in the area (cut-off of 50 and 35 IU in men and women, respectively). We found no association between ALT serum values and each habit change. Particularly, 12.7 and 11.1% of patients reduced their alcohol intake among those with high and normal ALT serum levels, respectively (p > 0.1) (data not shown in table). 4. Discussion We aimed to estimate, among subjects with HCV infection: (a) alcohol and coffee intake and smoking habit, (b) the reliability of patients’ self-reported data on these habits and (c) the possible change in patients’ habits after their first contact with a Viral Hepatitis Service. About half of the patients reported to drink alcoholic beverages daily (males: 55.5%; females: 35.3%); 29.5% of males consumed >40 g/day, and 31% of females drank >30 g/day. Recent Italian studies of people with HCV infection found similar values for alcohol intake: 17% of anti-HCV-positive subjects (most of them males) referred to 79 Italian hospitals reported >40 g/day of alcohol intake [25], and 30–40% of patients continued to drink >40 g/day of ethanol after diagnosis of HCV-related chronic hepatitis or cirrhosis in another research [16]. As regards coffee drinking and cigarette smoking, we found that most patients were regular coffee drinkers and about one-third of them smoked cigarettes daily. These data are in agreement with current habits of the Italian general population. In fact, a national survey in 2001 [26] showed that, among people aged 45–54 years, 59% of males and 25% of females were daily alcohol drinkers and 37% of males and 24% of females were current smokers. Compared with other European populations, we found that Italians adults, aged 15 years and over, had a lower mean alcohol intake than those living in other Mediterranean areas, such as France, Greece, and Spain. The percentage of daily smokers in Italy was similar to that in other Mediterranean European countries [27]. No national or international data are available on coffee consumption. In a previous research, some of the Authors carried out in the same area [6], found that 84% of males and 92% of females in the general adult population were current coffee drinkers. The second aim of our study was to evaluate the reliability of information on patients’ alcohol and coffee drinking and smoking habits collected during a clinical assessment. The test–retest analysis showed a fair to very good reliability of the data collected, with a higher reliability for smoking habits (K = 0.82) compared to coffee (K = 0.71) and alcohol (K = 0.58) drinking. Women were more reliable than men and older people more than younger ones regarding alcohol drinking and smoking habits. These findings are in broad agreement with other reliability studies carried out in Italy among the general population [28,29] or patients admitted to hospital for diseases unrelated to alcohol intake or tobacco smoking [18,30]. The higher reliability we found for cigarette smoking compared to alcohol and coffee drinking is in line with the above-mentioned reliability studies and probably depends on the more stable characteristics of smoking habit, which is actually an addiction and is usually classified in broad categories of use. The value of the kappa statistic does not necessarily indicate a moderate reliability, although the overall agreement was very high (92.5%), due to the unbalanced table. Another Italian study confirmed a not particularly high reliability for coffee drinking, with an agreement of 61% [29]. The different methods used in the two interviews (first face-toface and second by telephone) may have caused a bias, especially
603
as regards alcohol drinking. A comprehensive review of reliability of measures of quantity, frequency and volume of alcohol drinking showed that telephone survey estimates were usually slightly lower than estimates from personal interview surveys [31]. However, significant differences were found for frequency but not for average alcohol consumption in the most recent and large studies [31]. One Italian study comparing in-hospital face-to-face interviews with subsequent in-home interviews, using the same questionnaire, showed that subjects tended to underreport alcohol, smoking and coffee consumption in the hospital setting compared to at home [30], which is contrary to the overall reduction in alcohol, coffee and cigarette consumption we observed from before to after the first contact with the Viral Hepatitis Service. A substantial bias due to the different modes of interview therefore seems unlikely. A possible selection bias for patients included in the study as regards their habits, particularly alcohol drinking, is to be considered. In fact, GPs may investigate and refer heavy alcohol drinkers with HCV infection more than those with HCV infection alone, and therefore heavy drinkers may be overrepresented in the sample. It does not seem likely, however, as patients claimed a lower alcohol intake than people unaffected by liver disease living in the same area and of the same age and gender [11]. The lower alcohol consumption of patients with HCV infection with respect to people unaffected by liver disease may be true and due to their GP’s advice or other sources of information on the increased risk of liver disease progression due to alcohol intake. Also a response bias should be considered because these patients were probably aware that heavy alcohol intake might not have been approved by the physician who interviewed them (socially undesirable response bias). Should this response bias be substantial, a poor reliability of alcohol intake is expected, as a low validity causes low reliability [19]. On the contrary, though a similar response bias in both interviews cannot be excluded definitely, the reliability of alcohol drinking was fair to moderate, according to the various measures used, and not much lower than that of coffee drinking. These findings seem satisfactory compared with other studies among people with liver diseases. Corrao et al. [32] found a low reproducibility for measuring alcohol intake by interviewadministered questionnaire in the hospital and in home interviews, the latter being performed one year after hospital discharge, with an ICC for category of intake of 0.32 among cirrhotic cases and 0.47 among patients unaffected by liver disease [32]. The ICC of 0.47 for alcohol intake found by Corrao et al. [32] among subjects unaffected by clinically evident liver disease is very close to the value we found. Like Corrao et al. [32], we found that patients reported a higher alcohol intake in the 2nd interview at home, compared to the 1st interview at hospital. When comparing the agreement between patients’ habits at the time of the 1st interview, as collected by the two interviews (reliability study), we found that the number of current alcohol drinkers, coffee drinkers and cigarette smokers did not vary substantially. On the contrary, we found, at the 1st compared to the 2nd interview, a slightly higher number of heavy alcohol drinkers (39 and 31, respectively) (Table 2), but a lower number of drinkers of 5 or more cups of coffee/day (22 and 40, respectively) (Table 3) and a lower number of smokers of >20 cigarettes/day (5 and 9, respectively) (Table 4). We see no clear reason for the opposite direction of the differences found between the two interviews for coffee intake and cigarette smoking compared to alcohol drinking, so we cannot rule out that these differences are due to chance. Lastly, we evaluated changes in the habits of patients with HCV infection before and 6–12 months after their first contact with the Viral Hepatitis Service. The number of subjects who regularly drank alcohol 6–12 months after the first contact with the Viral Hepatitis Service was the same as it was at the first contact (n = 112 and n = 114, respectively) but the mean intake decreased 27.7% from before to
604
C. Zani et al. / Digestive and Liver Disease 41 (2009) 599–604
after the contact (Table 5). Recall bias cannot be excluded, however, as alcohol intake may have been underestimated at the 1st interview in a clinical setting, with respect to the 2nd interview at home. However, a decrease in alcohol intake after the diagnosis of hepatitis was observed also in other studies. Pessione et al. [14] reported that in HCV patients who drank alcohol daily, one-third stopped drinking after the diagnosis of liver disease and others reduced their mean alcohol consumption from 29 to 11 g/day. An Italian study [16] showed that, among patients with HCV infection, alcohol was used daily by 68% of patients before and 36% after the diagnosis. The proportion of coffee drinkers did not change between the first contact and 2nd interview. Mean coffee intake, however, decreased by 21.8% after the first contact. A decrease in coffee intake cannot be regarded a useful to these patients, however. In fact, various cohort and case–control studies found that coffee has a protective effect on the liver [6,33,34] and a recent case–control study showed that coffee intake decreased the risk of developing HCC among patients with chronic type-C liver disease [35]. Coffee contains a large amount of antioxidants and several human studies have shown that coffee compounds have a direct inhibitory effect on carcinogenic potential in the liver [36]. Several studies found an inverse relation between coffee drinking and serum levels of gamma-glutamyltransferase and aminotransferase [37,38]. Lastly, we found no significant decrease in the number of cigarettes smoked daily, from 13.6 to 13.2 cigarettes/day. Some studies found that tobacco smoking is harmful to the liver, particularly in subjects with HCV infection (14–15). However, the mechanism whereby smoking may increase histological activity is not clear. The effect of nicotine on the liver was studied on rats, which developed significant liver lesions characterised by steatosis and focal or confluent necrosis [39]. Therefore, a precautionary advise of reducing or stopping tobacco smoking could be valuable to patients with HCV infection, also because of the well known burden of disease caused by tobacco [4]. In conclusion, we found that a fairly high proportion of patients with chronic HCV disease regularly drink alcohol and coffee beverages and smoke cigarettes, especially among males. The reliability of self-reported data regarding all these habits seems satisfactory. 6–12 months after the first contact with the Viral Hepatitis Service, patients reduced alcohol intake, and also coffee drinking and cigarette smoking, possibly due to the advice received from the doctors at the Viral Hepatitis Service. Further research is needed to disentangle the effects of physicians’ advice, and possible response bias due to different modes of interview or to the phenomenon of regression. More decisive action to modify alcohol intake, and possibly tobacco smoking as well, seems mandatory in these patients, as almost all alcohol drinkers and cigarette smokers were still drinking alcohol beverages or smoking cigarettes 6–12 months after their first contact with the Viral Hepatitis Service. Conflict of interest None declared. References [1] IARC. Alcohol Drinking. Monographs on the Evaluation of Carcinogenic Risks Humans. 44;1988:1–378. [2] Mandayam S, Jamal MM, Morgan TR. Epidemiology of alcoholic liver disease. Semin Liver Dis 2004;24:217–32. [3] Morgan TR, Mandayam S, Jamal MM. Alcohol and hepatocellular carcinoma. Gastroenterology 2004;127:S87–96. [4] IARC. Tobacco smoke and involuntary smoking. Monographs on the Evaluation of Carcinogenic Risks to Humans 2004;83:679–710. [5] Corrao G, et al. Coffee, caffeine, and the risk of liver cirrhosis. Ann Epidemiol 2001;11:458–65. [6] Gelatti U, Covolo L, Franceschini M, Pirali F, Tagger A, Ribero ML, et al. Coffee consumption reduces the risk of hepatocellular carcinoma independently of its aetiology: a case–control study. J Hepatol 2005;42:528–34.
[7] Ruhl CE, Everhart JE. Joint effects of body weight and alcohol on elevated serum alanine aminotransferase in the United States population. Clin Gastroenterol Hepatol 2005;3:1260–8. [8] Seeff LB. Natural history of chronic hepatitis C. Hepatology 2002;36:S35–46. [9] Peters MG, Terrault NA. Alcohol use and hepatitis C. Hepatology 2002;36: S220–225. [10] Poynard T, Bedossa P, Opolon P. Natural history of liver fibrosis progression in patients with chronic hepatitis C. The OBSVIRC, METAVIR, CLINIVIR, and DOSVIRC groups. Lancet 1997;349:825–32. [11] Donato F, Tagger A, Gelatti U, Parrinello G, Boffetta P, Albertini A, et al. Alcohol and hepatocellular carcinoma: the effect of lifetime intake and hepatitis virus infections in men and women. Am J Epidemiol 2002;155:323–31. [12] Corrao G, Arico S. Independent and combined action of hepatitis C virus infection and alcohol consumption on the risk of symptomatic liver cirrhosis. Hepatology 1998;27:914–9. [13] Anand BS, Currie S, Dieperink E, Bini EJ, Shen H, Ho SB, et al. Alcohol use and treatment of hepatitis C virus: results of a national multicenter study. Gastroenterology 2006;130:1912–4. [14] Pessione F, Ramond MJ, Njapoum C, Duchatelle V, Degott C, Erlinger S, et al. Cigarette smoking and hepatic lesions in patients with chronic hepatitis C. Hepatology 2001;34:121–5. [15] Hezode C, Lonjon I, Roudot-Thoraval F, Mavier JP, Pawlotsky JM, Zafrani ES, et al. Impact of smoking on histological liver lesions in chronic hepatitis C. Gut 2003;52:126–9. [16] Loguercio C, Di Pierro M, Di Marino MP, Federico A, Disalvo D, Crafa E, et al. Drinking habits of subjects with hepatitis C virus-related chronic liver disease: prevalence and effect on clinical, virological and pathological aspects. Alcohol Alcoholism 2000;35:296–301. [17] Strader DB, et al. Diagnosis, management, and treatment of hepatitis C. Hepatology 2004;39:1147–71. [18] Donato F, Boffetta P, Fazioli R, Gelatti U, Porru S. Reliability of data on smoking habit and coffee drinking collected by personal interview in a hospital-based case–control study. Eur J Epidemiol 1998;14:259–67. [19] Amstrong BK, White E, Saracci R. Principles of exposure measurement in epidemiology. New York: Oxford Medical Publications; 1992. [20] Dunn G. Design and analysis of reliability studies. New York: Oxford University Press; 1989. [21] Altman DG. Practical statistics for medical research. London: Chapman & Hall; 1991. [22] Thompson WG, Walter DW. A reappraisal of the kappa coefficient. J Clin Epidemiol 1988;41:949–58. [23] Feinstein AR, Cicchetti DV. High agreement but low kappa. I. The problems of two paradoxes. J Clin Epidemiol 1990;43:543–9. [24] Walter SD, Eliasziw M, Donner A. Sample size and optimal designs for reliability studies. Stat Med 1998;17:101–10. [25] Stroffolini T, et al. Characteristics of HCV positive subjects referring to hospitals in Italy: a multicentre prevalence study on 6,999 cases. J Viral Hepat 2006;13:351–4. [26] Istituto nazionale di statistica, Indicatori socio-sanitari [in Italian] http://www. istat.it/sanita/sociosan/. [27] European Commission, European statistic data on population and social conditions http://epp.eurostat.ec.europa.eu/. [28] Ferraroni M, Decarli A, Franceschi S, La Vecchia C, Enard L, Negri E, et al. Validity and reproducibility of alcohol consumption in Italy. Int J Epidemiol 1996;25:775–82. [29] Ferraroni M, Tavani A, Decarli A, Franceschi S, Parpinel M, Negri E, et al. Reproducibility and validity of coffee and tea consumption in Italy. Eur J Clin Nutr 2004;58:674–80. [30] D’Avanzo B, La Vecchia C, Katsouyanni K, Negri E, Trichopoulos D. Reliability of information on cigarette smoking and beverage consumption provided by hospital controls. Epidemiology 1996;7:312–5. [31] Rehm J. Measuring quantity, frequency, and volume of drinking. Alcohol Clin Exp Res 1998;22:4S–14S. [32] Corrao G, Busellu M, Valenti M, Torchio P, Galatola G, Di Orio F. Reproducibility of measuring lifetime alcohol consumption. Alcologia 1994;6:127–35. [33] Inoue M, et al. Influence of coffee drinking on subsequent risk of hepatocellular carcinoma: a prospective study in Japan. J Natl Cancer I 2005;97:293–300. [34] Klatsky AL, Armstrong MA, Friedman GD. Coffee, tea, and mortality. Ann Epidemiol 1993;3:375–81. [35] Ohfuji S, Fukushima W, Tanaka T, Habu D, Tamori A, Sakaguchi H, et al. Coffee consumption and reduced risk of hepatocellular carcinoma among patients with chronic type C liver disease: a case–control study. Hepatol Res 2006;36:201–8. [36] Cadden IS, Partovi N, Yoshida EM. Review article: possible beneficial effects of coffee on liver disease and function. Aliment Pharmacol Ther 2007;26:1–8. [37] Tanaka K, Tokunaga S, Kono S, Tokudome S, Akamatsu T, Moriyama T, et al. Coffee consumption and decreased serum gamma-glutamyltransferase and aminotransferase activities among male alcohol drinkers. Int J Epidemiol 1998;27:438–43. [38] Hu G, Tuomilehto J, Pukkala E, Hakulinen T, Antikainen R, Vartiainen E, et al. Joint effects of coffee consumption and serum gamma-glutamyltransferase on the risk of liver cancer. Hepatology 2008;48:129–36. [39] Yuen ST, Gogo Jr AR, Luk IS, Cho CH, Ho JC, Loh TT. The effect of nicotine and its interaction with carbon tetrachloride in the rat liver. Pharmacol Toxicol 1995;77:225–30.
Contents lists available at ScienceDirect
Digestive and Liver Disease journal homepage: www.elsevier.com/locate/dld
Liver, Pancreas and Biliary Tract
Alcohol and coffee drinking and smoking habit among subjects with HCV infection C. Zani a , F. Donato a,∗ , M. Chiesa a , C. Baiguera b , U. Gelatti a , L. Covolo a , M.G. Antonini b , P. Nasta b , F. Gatti b , G. Orizio a , M. Puoti b a b
Department of Experimental and Applied Medicine, Institute of Hygiene, Epidemiology and Public Health, University of Brescia, Brescia, Italy Liver Unit of the Infectious Diseases Department, Spedali Civili, Brescia, Italy
a r t i c l e
i n f o
Article history: Received 28 July 2008 Accepted 20 November 2008 Available online 29 January 2009 Keywords: Alcohol drinking Coffee drinking HCV infection Tobacco smoking
a b s t r a c t Background/aims: The aims were to estimate among patients with hepatitis C virus (HCV) infection the prevalence of alcohol and coffee intake and smoking habit, the reliability of these self-reported data and the possible change of patients’ habit after their first contact with a Viral Hepatitis Service. Methods: 229 patients were initially interviewed personally at the Viral Hepatitis Service and after 6 months they were re-interviewed by phone in regard to their alcohol, coffee drinking and smoking habits. Results: Alcohol drinkers were 55.5% of males and 35.3% of females. Most subjects drank coffee daily, both men (90.0%) and women (84.9%). The proportion of current smokers was higher in males (43.6%) than females (26.9%). We found a fair to good reliability of self-reported data regarding patients’ habits, alcohol and coffee intake, and number of cigarettes smoked daily. We observed a statistically significant decrease in alcohol and coffee intake and cigarettes smoked between baseline and follow-up interviews. Conclusion: We found a fairly high proportion of HCV-infected patients who regularly drink alcohol and coffee beverages and smoke cigarettes, especially among males. The reliability of self-reported data on these habits seems satisfactory. More decisive action to modify these habits, especially alcohol intake, is required in these patients. © 2009 Editrice Gastroenterologica Italiana S.r.l. Published by Elsevier Ltd. All rights reserved.
1. Introduction Some common lifestyle habits can influence the development of liver disease. Alcohol drinking is a well-known cause of chronic liver disease [1–3] and a dose–effect relationship between alcohol intake and the risk of developing a clinically evident liver disease is reported. Tobacco smoking is a biologically plausible risk factor for liver cancer, especially among people with other risk factors for chronic liver disease, and some findings support this hypothesis [4]. On the other hand, coffee drinking may play a protective role on the liver, since an inverse relation between coffee drinking and cirrhosis or hepatocellular carcinoma (HCC) has been found in some studies [5–7]. Many subjects with hepatitis C virus (HCV) infection have a progression of chronic hepatitis C towards cirrhosis and HCC. Various viral and host factors can influence the progression of the hepatic disease [8], among them, the effect of alcohol drinking on
∗ Corresponding author at: Sezione di Igiene, Epidemiologia e Sanità Pubblica, Università di Brescia, Facoltà di Medicina e Chirurgia, Viale Europa 11, 25123 Brescia, Italy. Tel.: +39 030 3838602; fax: +39 030 370 1404. E-mail address: [email protected] (F. Donato).
progression of HCV-related disease has been investigated widely [9,10]. An interaction between alcohol intake and HCV infection in causing cirrhosis and HCC has been found [3,11,12]. Furthermore, alcohol drinking can influence the response to HCV treatment [13]. Also a possible effect of tobacco smoking in increasing, and of coffee drinking in decreasing, the risk of the progression of liver damage towards cirrhosis have been shown in some studies [5,14,15]. In spite of the role that alcohol and coffee drinking and smoking habits may play in the development of chronic liver disease due to HCV infection, few studies are available on the epidemiologic pattern of these habits in HCV-infected individuals [16]. Furthermore, as data on lifestyle habits are usually collected through interviews, their reliability is questionable, but as far as we know these behaviours have not been investigated among patients with HCV infection outside the hospital setting. The aims of this study were to investigate among patients with HCV infection: (a) the prevalence of alcohol and coffee intake and smoking habit, (b) the reliability of self-reported data on these habits and (c) the possible change of patients’ habits between before and after their first contact with a Viral Hepatitis Service.
1590-8658/$36.00 © 2009 Editrice Gastroenterologica Italiana S.r.l. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.dld.2008.11.018
600
C. Zani et al. / Digestive and Liver Disease 41 (2009) 599–604
2. Methods 2.1. Study design All subjects were outpatients admitted for the first time due to anti-HCV positivity to the Viral Hepatitis Service of the Infectious Diseases Department of the “Spedali Civili” Hospital in Brescia, North Italy, between March 2006 and December 2007. They underwent two interviews using the same questionnaire: the first was performed at the first contact with the Viral Hepatitis Service and the second on average 6 months later (range 5–12 months). First, all participants were interviewed face-to-face during their visit by a trained interviewer, and they were told they would be recalled after 6 months. In accordance with international guidelines [17], physicians advised patients to abstain totally from drinking alcohol and to stop smoking. Patients who also had a serious disease, particularly a cardiovascular disorder, were advised to moderate their coffee consumption as well. Each subject was required to sign an informed consent form prior to participation. The study was approved by the Ethics Committee of the “Spedali Civili” Hospital. The data were analysed anonymously, in accordance with Italian privacy laws concerning the use of personal health data for research purposes. The 2nd interview was conducted, using the same questionnaire, over the telephone by an interviewer who was unaware of the answers given by each individual at the 1st interview. The subjects were asked about: (a) their habits, at the time of the 1st interview, so as to evaluate the reliability of self-report data, and (b) their habits, at the time of the 2nd interview, so as to detect any changes that had occurred in the interval between the two interviews. Each interview lasted an average of 10 min. 2.2. Population All subjects were outpatients who went to the Viral Hepatitis Service of the Infectious Disease Department for the first time, following their GP’s suggestions, because of anti-HCV positivity. The patients had not received any antiviral treatment before the visit. Among them, 47.2% started antiviral treatment after the visit. They had been referred by their general practitioners because they were either discovered to have an HCV infection or found to have an increase in transaminases or other signs or symptoms suggesting progression of liver disease. Subjects were enrolled according to the following inclusion criteria: born in Italy, aged between 18 and 65 years; not having HCC, other neoplasms, severe systemic diseases such as chronic heart failure, stroke, respiratory insufficiency and chronic renal failure, or severe physical or mental handicaps. Subjects who claimed intravenous drug use were also excluded. 2.3. Questionnaire and interview The questionnaire used to collect information on alcohol and coffee consumption and smoking habit was based on one that some authors used in previous case–control studies on cancer aetiology in this area, after having tested its reliability [11,18]. Alcohol consumption was assessed by asking about the frequency and amount of alcohol consumed in a typical day or week. We considered current drinkers patients who claimed to drink alcoholic beverages at least weekly. We evaluated the total amount of alcohol intake by summing up the number of glasses of wine, cans of beer, shots of aperitif, spirits and coffee with spirits taking account of their mean alcohol content. In accordance with international guidelines, we used different categories of alcohol intake among males and females, with thresholds of 40 and 30 g/day in males and females, respectively, to separate low-moderate from medium-
high intake. Accordingly, we dichotomised current drinkers into subjects with either “low” (1–40 and 1–30 g/day in men and women, respectively) or “high” (>40 and >30 g/day) alcohol consumption. Coffee intake was assessed by summing up the cups of coffee, cappuccino and “coffee and milk” drunk in a typical day. We considered coffee drinkers patients who drank at least one cup of coffee per day. Information on tobacco smoking was also collected. Subjects were classified as current smokers if they smoked at least one cigarette per day. Information provided by patients on their current alcohol and coffee consumption and tobacco smoking at the first contact with the Viral Hepatitis Service was referred to as “before” the first contact; those concerning patients’ current habits at the time of the 2nd interview were referred to as “6–12 months after” the first contact with the Viral Hepatitis Service. To evaluate if they decreased their consumption we asked them about their habits 10 years before, and if they had changed them. If they answered positively we asked the reason for this change. In this way we could evaluate if their GP had given them advice concerning their lifestyle. 2.4. Statistical methods In order to assess reliability, we compared the answers provided by each patient at the 1st interview and those provided at the 2nd interview regarding his or her habits 6–12 months earlier, at the time of the first contact with the Viral Hepatitis Service. Agreement between the two interviews was assessed by computing Cohen’s kappa statistic (K) for dichotomous variables, and the intra-class correlation coefficient (ICC) and Spearman’s rho for ordinal or continuous variables, respectively [19,20]. The ICC was computed using a two-way analysis of variance (ANOVA) with fixed effects [19]. The strength of agreement was evaluated for each of these statistics according to Altman’s suggestions for interpreting K values [21]: poor <0.20, fair 0.21–0.40, moderate 0.41–0.60, good 0.61–0.80 and very good 0.81–1.00. Low K values were not automatically regarded as indicative of non-agreement, however, since this statistic is strongly dependent on true prevalence [22] and may be low when agreement is high due to unbalanced, symmetrical marginals [23]. At the 2nd stage, we assessed the changes that had occurred in each patient’s current habits in the 6-month interval between the 1st and the 2nd interview (before and 6–12 months after the first contact with the Viral Hepatitis Service). To this end we compared each patient’s answers regarding his or her current habits at the time of each interview, using common statistical tests for the comparison of two proportions or means. The study was designed to enroll at least 100 men and 100 women in order to achieve a reliability of 0.5 using the ICC for ordinal variables (alcohol and coffee intake and number of cigarettes smoked per day) assuming a type I error of 5%, a power of 80% and a minimally acceptable level of reliability of 0.3, for each gender separately, using the formula suggested by Walter et al. [24]. The sample size was also large enough to detect a 10 mg/day reduction of alcohol intake in current alcohol drinkers among men, from before to after the first contact with the Viral Hepatitis Service (means of alcohol intake: from 30 to 20 mg/day), given a standard deviation of 30 mg/day, using a Student’s t-test for paired data (alpha = 0.05; power = 90%). Since we expected to lose up to 30% of the patients initially recruited, we enrolled 300 subjects in the study, 50% of whom males. The statistical analyses were performed using STATA software (Stata Statistical Software release 8.0, 2003; Stata Corporation, College Station, TX).
C. Zani et al. / Digestive and Liver Disease 41 (2009) 599–604
601
Table 1 Alcohol and coffee drinking and smoking habits of 229 patients with HCV infection at their first contact with the Viral Hepatitis Service (1st interview). All patients (%) (N = 229)
Males (%) (N = 110)
Females (%) (N = 119)
Alcohol drinking Never drinker Former drinker
81 (35.4) 45 (19.6)
20 (18.2) 29 (26.3)
61 (51.3) 16 (13.4)
Current drinker Low intake High intake
103 (45.0) 72 (69.9) 31 (30.1)
61 (55.5)a 43 (70.5) 18 (29.5)
42 (35.3)a 29 (69.0) 13 (31.0)
Coffee drinking Never drinker Former drinker
25 (10.9) 4 (1.8)
8 (7.3) 3 (2.7)
17 (14.3) 1 (0.8)
200 (87.3) 80 (40.0) 80 (40.0) 40 (20.0)
99 (90.0) 27 (27.3) 40 (40.4) 32 (32.3)b
101 (84.9) 53 (52.5) 40 (39.6) 8(7.9)b
112 (48.9) 37 (16.2)
35 (31.8) 27 (24.6)
77 (64.7) 10 (8.4)
80 (34.9) 27 (33.7) 44 (55.0) 9 (11.3)
48 (43.6)c 11 (22.9) 29 (60.4) 8 (16.7)
32 (26.9)c 16 (50.0) 15 (46.9) 1 (3.1)
Current drinker 1–2 cups of coffee/day 3–4 cups of coffee/day 5+ cups of coffee/day Smoking habit Never smoker Former smoker Current smoker 1–10 cigarettes/day 11–20 cigarettes/day > 20 cigarettes/day
Percentages are computed on the column total. However, percentages of alcohol drinkers according to ethanol intake categories, of coffee drinkers according to coffee intake categories and of smokers according to cigarettes smoked categories are computed on the total number of current users, respectively. Alcohol drinking: low intake ≤ 30 g alcohol/day for females and ≤40 g alcohol/day for males; high intake: >30 g alcohol/day for females and >40 g alcohol/day for males. a p = 0.002 by the Fisher’s exact test. b p < 0.001 by the Fisher’s exact test. c p = 0.008 by the Fisher’s exact test.
3. Results A total of 301 subjects were enrolled in the study and provided a valid 1st interview on their habits regarding alcohol and coffee drinking and tobacco smoking. Of these, 229 (76.0 %) were re-interviewed after 6–12 months and 72 subjects were phoned at least three times but could not be found or refused the interview. Among them 48.0% were males and the mean age was 51.8 ± 11.9 years (range 22–77 years) (data not shown in tables). All patients had a diagnosis of chronic hepatitis, 140 (61.2%) had elevated and 89 (38.8%) normal serum transaminases (data not shown in the table). Alcohol and coffee drinking and smoking habits of patients at their first contact with the Viral Hepatitis Service are reported in Table 1. Males drank more alcohol than females: current drinkers were 55.5 and 35.3%, respectively (p = 0.002). According to the different categorizations of alcohol intake for males and females, among current drinkers 29.5% males drank >40 g and 31.0% females drank >30 g/day of ethanol per day. Most subjects drank coffee daily, with similar percentages of current
drinkers in males and females (90.0 and 84.9%), although the proportion of drinkers of 5+ cups/day was higher in men than women (32.3 and 7.9%, respectively; p < 0.001). A higher proportion of current smokers was found in males than females (43.6 and 26.9%, respectively; p = 0.008), and among current smokers 16.7% of males and 3.1% of females smoked >20 cigarettes per day (p = 0.02). The 229 subjects who were re-interviewed were compared with the 72 subjects who were not: no significant difference was found between them and subjects re-interviewed as regards their alcohol consumption, coffee drinking and smoking habit at the 1st interview. However, a higher proportion of males was found among those lost (80.5%) compared to those with a complete follow-up (48%), and their mean age was 46.5 (S.D.: 14.5) compared to 51.8 (S.D.: 11.9) (data not shown in tables). Reliability of alcohol intake (Table 2) was fair to moderate as regards both the habit (current/former/never drinker), with a kappa of 0.58, and the total alcohol intake, categorized at three levels (ICC = 0.42) or considered as a continuous variable (grams of ethanol per day) (Spearman’s rho = 0.62).
Table 2 Test–retest reliability study investigating alcohol drinking habits and alcohol intake at the time of patients’ first contact with the Hepatitis Service, comparing the answers given at 1st and 2nd interview (6–12 months later). 1st interview
2nd interview
Alcohol drinking habits Current drinker Former drinker Never drinker
Current drinker 79 12 11
Alcohol intake (g/day) 0 Low High
0 103 20 4
Alcohol drinking habits Former drinker 9 23 4 Low 15 44 4
Never drinker 15 10 66
Alcohol intake (g/day) High 8 8 23
Agreement
K /ICC/Ra (95% CI)
73.4%
K = 0.58 (0.47–0.66)
ICC = 0.42 (0.23–0.46) 74.2% R = 0.62 (0.54–0.70)
The ICC was computed using the three categories ordinal variable whereas Spearman’s rho was computed using the continuous variable (g ethanol intake per day). a K = Cohen’s kappa; ICC = intraclass correlation coefficient; R = Spearman’s rho.
602
C. Zani et al. / Digestive and Liver Disease 41 (2009) 599–604
Table 3 Test–retest reliability study investigating coffee drinking habits and number of cups of coffee drunk per day at the time of patients’ first contact with the Hepatitis Service, comparing the answers given at 1st and 2nd interview (6–12 months after the former). 1st interview
2nd interview
Coffee drinking habits Current drinker Former and never drinker
Current drinker 186 3
Coffee intake (cups/day) 0 1–2 3–4 5+
0 49 25 1 2
Coffee drinking habits Former and never 14 26
1–2 9 55 15 11
Coffee intake (cups/day) 3–4 0 14 13 13
5+ 0 4 4 14
Agreement
K/ICC/Ra (95% CI)
92.5%
K = 0.71 (0.59–0.82)
ICC = 0.64 (0.53–0.72) R = 0.72 (0.65–0.78)
57.2%
The ICC was computed using the four categories ordinal variable whereas Spearman’s rho was computed using the continuous variable (number of cups of coffee per day). a K = Cohen’s kappa; ICC = intraclass correlation coefficient; R = Spearman’s rho. Table 4 Test–retest reliability study investigating smoking habits and number of cigarettes smoked per day at the time of patients’ first contact with the Hepatitis Service, comparing the answers given at 1st and 2nd interview (6–12 months after the former). 1st interview
2nd interview
Smoking habits Current smoker Former smoker Never smoker
Smoking habits Current smoker 68 5 2
No. cigarettes/day 0 1–10 11–20 >20
0 146 13 22 2
Former smoker 4 30 3
1–10 3 11 1 0
Never smoker 8 2 107 No. cigarettes/day 11–20 0 3 21 2
>20 0 0 0 5
Agreement
K /ICC/Ra (95% CI)
89.5%
K = 0.82 (0.74–0.89)
ICC = 0.64 (0.54–0.70) R = 0.63 (0.55–0.70)
84.6%
The ICC was computed using the four categories ordinal variable whereas Spearman’s rho was computed using the continuous variable (number of cigarettes smoked per day). a K = Cohen’s kappa; ICC = intraclass correlation coefficient; R = Spearman’s rho.
As regards coffee, former and never drinkers were collapsed into one category since there were very few former drinkers. The percentage of agreement regarding coffee drinking habit (Table 3) was 92.5% and the K value was 0.71. A good reliability was also found for coffee intake, with an ICC of 0.64 for the categorical variable, and Spearman’s rho of 0.72 for the continuous variable. Test–retest agreement for smoking habit was 89.5 and the K value was 0.82 (Table 4). A good reliability was found as regards the number of cigarettes smoked per day (agreement: 84.6%; ICC = 0.64; rho = 0.63). We found no association between unreliability of frequency and amount of alcohol and coffee drinking and cigarette smoking (data not shown in tables). Further analyses of agreement between the two interviews were performed for males and females, and subjects aged less than 50 years and those aged equal to or over 50 years, separately (data not shown in table). Compared to females, males showed a lower reliability regarding alcohol consumption (rho = 0.56 and rho = 0.67,
respectively) and smoking habits (rho = 0.58 and rho = 0.70) and no difference regarding coffee intake (rho = 0.70 and rho = 0.71). ICC values were very similar to rho values. Compared to older subjects, those aged <50 years showed a lower reliability for alcohol consumption (rho = 0.53 and rho = 0.68, respectively), smoking habits (rho = 0.53 and rho = 0.77) and coffee drinking (rho = 0.62 and rho = 0.74, respectively). Again, ICCs and rhos provided similar figures. Finally, in order to assess whether the subjects had changed their current habits from before to 6–12 months after the first contact with the Viral Hepatitis Service, we compared subjects’ answers regarding their current habits at each interview (Table 5). No change was found in the proportions of current users of alcohol and coffee and smokers. From before to after contact with the Viral Hepatitis Service, patients reported, among current users, a statistically significant decrease in the geometric means of alcohol intake (from 16.2 to 11.7 g/day, −27.7%), coffee intake (from 3.2 to 2.5 cups/day, −21.8%) and number of cigarettes smoked (from 13.6 to 13.2 per day, −2.9%).
Table 5 Alcohol and coffee drinking and smoking habit among HCV patients before and 6–12 months after the first contact with the Viral Hepatitis Service.
Current drinkers Geometric mean of alcohol intake (S.D.) (g/day)a Current coffee drinkers Geometric mean of number of cups of coffee/day (S.D.)a Current smokers Geometric mean of number cigarettes smoked/day (S.D.)a
Before 1st contact
After 6–12 months
114 16.2 (33.6) 200 3.2 (1.7) 80 13.6 (6.2)
112 11.7 (11.8) 189 2.5 (1.4) 75 13.2 (5.8)
Difference (%) −2 (−1.7) −4.5 (−27.7) −11 (−5.5) −0.7 (−21.8) −5 (−6.5) −0.4 (−2.9)
Pb NS p < 0.0001 NS p < 0.0001 NS p = 0.006
a The geometric means of alcohol and coffee intakes and number of cigarettes smoked were calculated among current users at the 1st and 2nd interview for each habit. The geometric instead of arithmetic mean was used for each variables due to its skewed distribution. b Wilcoxon test for paired data.
C. Zani et al. / Digestive and Liver Disease 41 (2009) 599–604
We investigated the association between ALT serum levels and changes in individuals’ habits from the 1st to the 2nd interview, dichotomising ALT values into normal and high according to the distribution of values in the general population living in the area (cut-off of 50 and 35 IU in men and women, respectively). We found no association between ALT serum values and each habit change. Particularly, 12.7 and 11.1% of patients reduced their alcohol intake among those with high and normal ALT serum levels, respectively (p > 0.1) (data not shown in table). 4. Discussion We aimed to estimate, among subjects with HCV infection: (a) alcohol and coffee intake and smoking habit, (b) the reliability of patients’ self-reported data on these habits and (c) the possible change in patients’ habits after their first contact with a Viral Hepatitis Service. About half of the patients reported to drink alcoholic beverages daily (males: 55.5%; females: 35.3%); 29.5% of males consumed >40 g/day, and 31% of females drank >30 g/day. Recent Italian studies of people with HCV infection found similar values for alcohol intake: 17% of anti-HCV-positive subjects (most of them males) referred to 79 Italian hospitals reported >40 g/day of alcohol intake [25], and 30–40% of patients continued to drink >40 g/day of ethanol after diagnosis of HCV-related chronic hepatitis or cirrhosis in another research [16]. As regards coffee drinking and cigarette smoking, we found that most patients were regular coffee drinkers and about one-third of them smoked cigarettes daily. These data are in agreement with current habits of the Italian general population. In fact, a national survey in 2001 [26] showed that, among people aged 45–54 years, 59% of males and 25% of females were daily alcohol drinkers and 37% of males and 24% of females were current smokers. Compared with other European populations, we found that Italians adults, aged 15 years and over, had a lower mean alcohol intake than those living in other Mediterranean areas, such as France, Greece, and Spain. The percentage of daily smokers in Italy was similar to that in other Mediterranean European countries [27]. No national or international data are available on coffee consumption. In a previous research, some of the Authors carried out in the same area [6], found that 84% of males and 92% of females in the general adult population were current coffee drinkers. The second aim of our study was to evaluate the reliability of information on patients’ alcohol and coffee drinking and smoking habits collected during a clinical assessment. The test–retest analysis showed a fair to very good reliability of the data collected, with a higher reliability for smoking habits (K = 0.82) compared to coffee (K = 0.71) and alcohol (K = 0.58) drinking. Women were more reliable than men and older people more than younger ones regarding alcohol drinking and smoking habits. These findings are in broad agreement with other reliability studies carried out in Italy among the general population [28,29] or patients admitted to hospital for diseases unrelated to alcohol intake or tobacco smoking [18,30]. The higher reliability we found for cigarette smoking compared to alcohol and coffee drinking is in line with the above-mentioned reliability studies and probably depends on the more stable characteristics of smoking habit, which is actually an addiction and is usually classified in broad categories of use. The value of the kappa statistic does not necessarily indicate a moderate reliability, although the overall agreement was very high (92.5%), due to the unbalanced table. Another Italian study confirmed a not particularly high reliability for coffee drinking, with an agreement of 61% [29]. The different methods used in the two interviews (first face-toface and second by telephone) may have caused a bias, especially
603
as regards alcohol drinking. A comprehensive review of reliability of measures of quantity, frequency and volume of alcohol drinking showed that telephone survey estimates were usually slightly lower than estimates from personal interview surveys [31]. However, significant differences were found for frequency but not for average alcohol consumption in the most recent and large studies [31]. One Italian study comparing in-hospital face-to-face interviews with subsequent in-home interviews, using the same questionnaire, showed that subjects tended to underreport alcohol, smoking and coffee consumption in the hospital setting compared to at home [30], which is contrary to the overall reduction in alcohol, coffee and cigarette consumption we observed from before to after the first contact with the Viral Hepatitis Service. A substantial bias due to the different modes of interview therefore seems unlikely. A possible selection bias for patients included in the study as regards their habits, particularly alcohol drinking, is to be considered. In fact, GPs may investigate and refer heavy alcohol drinkers with HCV infection more than those with HCV infection alone, and therefore heavy drinkers may be overrepresented in the sample. It does not seem likely, however, as patients claimed a lower alcohol intake than people unaffected by liver disease living in the same area and of the same age and gender [11]. The lower alcohol consumption of patients with HCV infection with respect to people unaffected by liver disease may be true and due to their GP’s advice or other sources of information on the increased risk of liver disease progression due to alcohol intake. Also a response bias should be considered because these patients were probably aware that heavy alcohol intake might not have been approved by the physician who interviewed them (socially undesirable response bias). Should this response bias be substantial, a poor reliability of alcohol intake is expected, as a low validity causes low reliability [19]. On the contrary, though a similar response bias in both interviews cannot be excluded definitely, the reliability of alcohol drinking was fair to moderate, according to the various measures used, and not much lower than that of coffee drinking. These findings seem satisfactory compared with other studies among people with liver diseases. Corrao et al. [32] found a low reproducibility for measuring alcohol intake by interviewadministered questionnaire in the hospital and in home interviews, the latter being performed one year after hospital discharge, with an ICC for category of intake of 0.32 among cirrhotic cases and 0.47 among patients unaffected by liver disease [32]. The ICC of 0.47 for alcohol intake found by Corrao et al. [32] among subjects unaffected by clinically evident liver disease is very close to the value we found. Like Corrao et al. [32], we found that patients reported a higher alcohol intake in the 2nd interview at home, compared to the 1st interview at hospital. When comparing the agreement between patients’ habits at the time of the 1st interview, as collected by the two interviews (reliability study), we found that the number of current alcohol drinkers, coffee drinkers and cigarette smokers did not vary substantially. On the contrary, we found, at the 1st compared to the 2nd interview, a slightly higher number of heavy alcohol drinkers (39 and 31, respectively) (Table 2), but a lower number of drinkers of 5 or more cups of coffee/day (22 and 40, respectively) (Table 3) and a lower number of smokers of >20 cigarettes/day (5 and 9, respectively) (Table 4). We see no clear reason for the opposite direction of the differences found between the two interviews for coffee intake and cigarette smoking compared to alcohol drinking, so we cannot rule out that these differences are due to chance. Lastly, we evaluated changes in the habits of patients with HCV infection before and 6–12 months after their first contact with the Viral Hepatitis Service. The number of subjects who regularly drank alcohol 6–12 months after the first contact with the Viral Hepatitis Service was the same as it was at the first contact (n = 112 and n = 114, respectively) but the mean intake decreased 27.7% from before to
604
C. Zani et al. / Digestive and Liver Disease 41 (2009) 599–604
after the contact (Table 5). Recall bias cannot be excluded, however, as alcohol intake may have been underestimated at the 1st interview in a clinical setting, with respect to the 2nd interview at home. However, a decrease in alcohol intake after the diagnosis of hepatitis was observed also in other studies. Pessione et al. [14] reported that in HCV patients who drank alcohol daily, one-third stopped drinking after the diagnosis of liver disease and others reduced their mean alcohol consumption from 29 to 11 g/day. An Italian study [16] showed that, among patients with HCV infection, alcohol was used daily by 68% of patients before and 36% after the diagnosis. The proportion of coffee drinkers did not change between the first contact and 2nd interview. Mean coffee intake, however, decreased by 21.8% after the first contact. A decrease in coffee intake cannot be regarded a useful to these patients, however. In fact, various cohort and case–control studies found that coffee has a protective effect on the liver [6,33,34] and a recent case–control study showed that coffee intake decreased the risk of developing HCC among patients with chronic type-C liver disease [35]. Coffee contains a large amount of antioxidants and several human studies have shown that coffee compounds have a direct inhibitory effect on carcinogenic potential in the liver [36]. Several studies found an inverse relation between coffee drinking and serum levels of gamma-glutamyltransferase and aminotransferase [37,38]. Lastly, we found no significant decrease in the number of cigarettes smoked daily, from 13.6 to 13.2 cigarettes/day. Some studies found that tobacco smoking is harmful to the liver, particularly in subjects with HCV infection (14–15). However, the mechanism whereby smoking may increase histological activity is not clear. The effect of nicotine on the liver was studied on rats, which developed significant liver lesions characterised by steatosis and focal or confluent necrosis [39]. Therefore, a precautionary advise of reducing or stopping tobacco smoking could be valuable to patients with HCV infection, also because of the well known burden of disease caused by tobacco [4]. In conclusion, we found that a fairly high proportion of patients with chronic HCV disease regularly drink alcohol and coffee beverages and smoke cigarettes, especially among males. The reliability of self-reported data regarding all these habits seems satisfactory. 6–12 months after the first contact with the Viral Hepatitis Service, patients reduced alcohol intake, and also coffee drinking and cigarette smoking, possibly due to the advice received from the doctors at the Viral Hepatitis Service. Further research is needed to disentangle the effects of physicians’ advice, and possible response bias due to different modes of interview or to the phenomenon of regression. More decisive action to modify alcohol intake, and possibly tobacco smoking as well, seems mandatory in these patients, as almost all alcohol drinkers and cigarette smokers were still drinking alcohol beverages or smoking cigarettes 6–12 months after their first contact with the Viral Hepatitis Service. Conflict of interest None declared. References [1] IARC. Alcohol Drinking. Monographs on the Evaluation of Carcinogenic Risks Humans. 44;1988:1–378. [2] Mandayam S, Jamal MM, Morgan TR. Epidemiology of alcoholic liver disease. Semin Liver Dis 2004;24:217–32. [3] Morgan TR, Mandayam S, Jamal MM. Alcohol and hepatocellular carcinoma. Gastroenterology 2004;127:S87–96. [4] IARC. Tobacco smoke and involuntary smoking. Monographs on the Evaluation of Carcinogenic Risks to Humans 2004;83:679–710. [5] Corrao G, et al. Coffee, caffeine, and the risk of liver cirrhosis. Ann Epidemiol 2001;11:458–65. [6] Gelatti U, Covolo L, Franceschini M, Pirali F, Tagger A, Ribero ML, et al. Coffee consumption reduces the risk of hepatocellular carcinoma independently of its aetiology: a case–control study. J Hepatol 2005;42:528–34.
[7] Ruhl CE, Everhart JE. Joint effects of body weight and alcohol on elevated serum alanine aminotransferase in the United States population. Clin Gastroenterol Hepatol 2005;3:1260–8. [8] Seeff LB. Natural history of chronic hepatitis C. Hepatology 2002;36:S35–46. [9] Peters MG, Terrault NA. Alcohol use and hepatitis C. Hepatology 2002;36: S220–225. [10] Poynard T, Bedossa P, Opolon P. Natural history of liver fibrosis progression in patients with chronic hepatitis C. The OBSVIRC, METAVIR, CLINIVIR, and DOSVIRC groups. Lancet 1997;349:825–32. [11] Donato F, Tagger A, Gelatti U, Parrinello G, Boffetta P, Albertini A, et al. Alcohol and hepatocellular carcinoma: the effect of lifetime intake and hepatitis virus infections in men and women. Am J Epidemiol 2002;155:323–31. [12] Corrao G, Arico S. Independent and combined action of hepatitis C virus infection and alcohol consumption on the risk of symptomatic liver cirrhosis. Hepatology 1998;27:914–9. [13] Anand BS, Currie S, Dieperink E, Bini EJ, Shen H, Ho SB, et al. Alcohol use and treatment of hepatitis C virus: results of a national multicenter study. Gastroenterology 2006;130:1912–4. [14] Pessione F, Ramond MJ, Njapoum C, Duchatelle V, Degott C, Erlinger S, et al. Cigarette smoking and hepatic lesions in patients with chronic hepatitis C. Hepatology 2001;34:121–5. [15] Hezode C, Lonjon I, Roudot-Thoraval F, Mavier JP, Pawlotsky JM, Zafrani ES, et al. Impact of smoking on histological liver lesions in chronic hepatitis C. Gut 2003;52:126–9. [16] Loguercio C, Di Pierro M, Di Marino MP, Federico A, Disalvo D, Crafa E, et al. Drinking habits of subjects with hepatitis C virus-related chronic liver disease: prevalence and effect on clinical, virological and pathological aspects. Alcohol Alcoholism 2000;35:296–301. [17] Strader DB, et al. Diagnosis, management, and treatment of hepatitis C. Hepatology 2004;39:1147–71. [18] Donato F, Boffetta P, Fazioli R, Gelatti U, Porru S. Reliability of data on smoking habit and coffee drinking collected by personal interview in a hospital-based case–control study. Eur J Epidemiol 1998;14:259–67. [19] Amstrong BK, White E, Saracci R. Principles of exposure measurement in epidemiology. New York: Oxford Medical Publications; 1992. [20] Dunn G. Design and analysis of reliability studies. New York: Oxford University Press; 1989. [21] Altman DG. Practical statistics for medical research. London: Chapman & Hall; 1991. [22] Thompson WG, Walter DW. A reappraisal of the kappa coefficient. J Clin Epidemiol 1988;41:949–58. [23] Feinstein AR, Cicchetti DV. High agreement but low kappa. I. The problems of two paradoxes. J Clin Epidemiol 1990;43:543–9. [24] Walter SD, Eliasziw M, Donner A. Sample size and optimal designs for reliability studies. Stat Med 1998;17:101–10. [25] Stroffolini T, et al. Characteristics of HCV positive subjects referring to hospitals in Italy: a multicentre prevalence study on 6,999 cases. J Viral Hepat 2006;13:351–4. [26] Istituto nazionale di statistica, Indicatori socio-sanitari [in Italian] http://www. istat.it/sanita/sociosan/. [27] European Commission, European statistic data on population and social conditions http://epp.eurostat.ec.europa.eu/. [28] Ferraroni M, Decarli A, Franceschi S, La Vecchia C, Enard L, Negri E, et al. Validity and reproducibility of alcohol consumption in Italy. Int J Epidemiol 1996;25:775–82. [29] Ferraroni M, Tavani A, Decarli A, Franceschi S, Parpinel M, Negri E, et al. Reproducibility and validity of coffee and tea consumption in Italy. Eur J Clin Nutr 2004;58:674–80. [30] D’Avanzo B, La Vecchia C, Katsouyanni K, Negri E, Trichopoulos D. Reliability of information on cigarette smoking and beverage consumption provided by hospital controls. Epidemiology 1996;7:312–5. [31] Rehm J. Measuring quantity, frequency, and volume of drinking. Alcohol Clin Exp Res 1998;22:4S–14S. [32] Corrao G, Busellu M, Valenti M, Torchio P, Galatola G, Di Orio F. Reproducibility of measuring lifetime alcohol consumption. Alcologia 1994;6:127–35. [33] Inoue M, et al. Influence of coffee drinking on subsequent risk of hepatocellular carcinoma: a prospective study in Japan. J Natl Cancer I 2005;97:293–300. [34] Klatsky AL, Armstrong MA, Friedman GD. Coffee, tea, and mortality. Ann Epidemiol 1993;3:375–81. [35] Ohfuji S, Fukushima W, Tanaka T, Habu D, Tamori A, Sakaguchi H, et al. Coffee consumption and reduced risk of hepatocellular carcinoma among patients with chronic type C liver disease: a case–control study. Hepatol Res 2006;36:201–8. [36] Cadden IS, Partovi N, Yoshida EM. Review article: possible beneficial effects of coffee on liver disease and function. Aliment Pharmacol Ther 2007;26:1–8. [37] Tanaka K, Tokunaga S, Kono S, Tokudome S, Akamatsu T, Moriyama T, et al. Coffee consumption and decreased serum gamma-glutamyltransferase and aminotransferase activities among male alcohol drinkers. Int J Epidemiol 1998;27:438–43. [38] Hu G, Tuomilehto J, Pukkala E, Hakulinen T, Antikainen R, Vartiainen E, et al. Joint effects of coffee consumption and serum gamma-glutamyltransferase on the risk of liver cancer. Hepatology 2008;48:129–36. [39] Yuen ST, Gogo Jr AR, Luk IS, Cho CH, Ho JC, Loh TT. The effect of nicotine and its interaction with carbon tetrachloride in the rat liver. Pharmacol Toxicol 1995;77:225–30.