Results
Characteristics of the included studies and patients
In total, 45 studies comprising 33,272 patients with HBV infection met the inclusion criteria, and their basic characteristics are listed in Table 115,17,26–68 and Supplementary Table 2. The NOS assessment showed that 23 (51.1%) and 22 (48.9%) studies had a low and moderate risk of bias, respectively (Supplementary Tables 3 and 4). The studies included 36 Chinese studies, four Italian studies, four Japanese studies, and one Korean study, with the majority of studies (n = 41, 91.1%) originating from Asian countries. The sample sizes ranged from 57 to 19,673, and the publication years spanned from 1982 to 2022. Of the studies, 41 (91.1%) had a case-control design and four (8.9%) had a cohort design. Among the patients, 26,687 (80.2%) were male and 6,585 (19.8%) were female. A total of 28,895 (86.5%) patients were alcohol drinkers, while 4,477 (13.5%) were non-drinkers. The cut-off points for categorizing alcohol consumption levels were 20, 40, 50, 60, 80, and 100 g per day in 10, six, five, four, four, and three studies, respectively. The assessment of alcohol consumption varied among the included studies, with most classifying high-level drinkers as those consuming ≥ 40 g of alcohol per day and low-level drinkers as those consuming no alcohol or <40 g per day. A total of 17,189 patients had cirrhosis or HCC, 13,407 had cirrhosis only, 14,621 had HCC only, and 10,839 had both cirrhosis and HCC (Fig. 1B).
Table 1Characteristics of studies included in the meta-analysis
NO. | Authors (Year) | Country | Study design | T | Mean age | Measure of liver progression |
---|
1a | Ikeda et al. (1998)15 | Japan | Cohort study | 610 | 34 | Cirrhosis | 208 (32.20%) |
1b | Ikeda et al. (1998)15 | Japan | Cohort study | 610 | 34 | Hepatocellular carcinoma | 121 (18.80%) |
2 | Lin et al. (2013)17 | China | Cohort study | 764 | 47 | Hepatocellular carcinoma | 138 (18%) |
3 | Hou et al. (2016)26 | China | Case-control | 298 | 43 | Hepatocellular carcinoma | 141 (47%) |
4 | Chen et al. (2013)27 | China | Case-control | 242 | 51 | Hepatocellular carcinoma | 121 (50%) |
5 | Sun et al. (2017)28 | China | Case-control | 283 | 47 | Cirrhosis | 89 (31%) |
6 | Wu et al. (2016)29 | China | Case-control | 700 | 45 | Hepatocellular carcinoma | 350 (50%) |
7 | Zhan et al. (2016)30 | China | Case-control | 266 | 52 | Hepatocellular carcinoma | 133 (50%) |
8 | Zhang et al. (2019)31 | China | Case-control | 238 | 50 | Hepatocellular carcinoma | 120 (50%) |
9 | Yu et al. (2015)32 | China | Case-control | 215 | 42 | Cirrhosis | 84 (39%) |
10a | Ohnishi et al. (1982)33 | Japan | Case-control | 57 | 44 | Cirrhosis | 35 (61%) |
10b | Ohnishi et al. (1982)33 | Japan | Case-control | 57 | 44 | Hepatocellular carcinoma | 52 (60%) |
11a | Xiaerfuhazi et al. (2014)34 | China | Case-control | 2,054 | 50 | Cirrhosis | 1,334 (65%) |
11b | Xiaerfuhazi et al. (2014)34 | China | Case-control | 2,054 | 50 | Hepatocellular carcinoma | 578 (28%) |
12 | Song et al. (2017)35 | China | Case-control | 151 | 55 | Hepatocellular carcinoma | 75 (50%) |
13 | Sun et al. (2002)36 | China | Case-control | 470 | 51 | Hepatocellular carcinoma | 89 (19%) |
14 | Lin et al. (2013)37 | China | Case-control | 94 | 55 | Hepatocellular carcinoma | 47 (50%) |
15 | Xin et al. (2020)38 | China | Cohort study | 343 | 55 | Hepatocellular carcinoma | 83 (24%) |
16 | Nie et al. (2017)39 | China | Case-control | 303 | 51 | Hepatocellular carcinoma | 137 (45%) |
17 | Lin et al. (2020)40 | China | Case-control | 308 | 50 | Hepatocellular carcinoma | 116 (38%) |
18 | Chen et al. (2018)41 | China | Case-control | 331 | 45 | Hepatocellular carcinoma | 46 (14%) |
19 | Wang et al. (2019)42 | China | Case-control | 152 | 51 | Hepatocellular carcinoma | 72 (47%) |
20 | Li et al. (2015)43 | China | Case-control | 158 | 52 | Hepatocellular carcinoma | 138 (87%) |
21 | Cao et al. (2003)44 | China | Case-control | 193 | 52 | Hepatocellular carcinoma | 148 (77%) |
22 | Li et al. (2002)45 | China | Case-control | 1,091 | 55 | Hepatocellular carcinoma | 626 (57%) |
23 | Genya et al. (2005)46 | China | Case-control | 205 | 53 | Hepatocellular carcinoma | 174 (85%) |
24 | Liu et al. (2011)47 | China | Case-control | 80 | 49 | Hepatocellular carcinoma | 42 (53%) |
25a | Villa et al. (1988)48 | Italy | Cohort study | 165 | 52 | Cirrhosis | 99 (60%) |
25b | Villa et al. (1988)48 | Italy | Cohort study | 165 | 52 | Hepatocellular carcinoma | 42 (25%) |
26 | Nonomu et al. (1986)49 | Japan | Case-control | 77 | 55 | Hepatocellular carcinoma | 51 (66%) |
27 | Sun et al. (2021)50 | China | Case-control | 80 | 60 | Hepatocellular carcinoma | 38 (48%) |
28 | Xie et al. (2021)51 | China | Case-control | 417 | 53 | Hepatocellular carcinoma | 57 (14%) |
29a | Zhang et al. (2015)52 | China | Case-control | 715 | 52 | Cirrhosis | 281 (39%) |
29b | Zhang et al. (2015)52 | China | Case-control | 715 | 52 | Hepatocellular carcinoma | 434 (61%) |
30a | Kwon et al. (2010)53 | Korea | Case-control | 292 | 53 | Cirrhosis | 146 (50%) |
30b | Kwon et al. (2010)53 | Korea | Case-control | 292 | 53 | Hepatocellular carcinoma | 146 (50%) |
31 | Bai et al. (2021)54 | China | Case-control | 80 | 43 | Cirrhosis | 40 (50%) |
32 | Wu et al. (2015)55 | China | Case-control | 240 | 49 | Hepatocellular carcinoma | 120 (50%) |
33 | Abassa et al. (2022)56 | China | Case-control | 19,673 | 52 | Hepatocellular carcinoma | 8,454 (43%) |
34 | Cao et al. (2018)57 | China | Case-control | 160 | 44 | Hepatocellular carcinoma | 20 (12.50%) |
35 | Stroffolini et al. (2010)58 | Italy | Case-control | 62 | 50 | Cirrhosis | 12 (19.30%) |
36a | Zhong et al. (2013)59 | China | Case-control | 106 | 64 | Cirrhosis | 73 (68.90%) |
36b | Zhong et al. (2013)59 | China | Case-control | 106 | 64 | Hepatocellular carcinoma | 47 (44.30%) |
37 | Donato et al. (2002)60 | Italy | Case-control | 136 | 60 | Hepatocellular carcinoma | 92 (67.60%) |
38 | Donao et al. (1997)61 | Italy | Case-control | 59 | 60 | Hepatocellular carcinoma | 41 (69%) |
39 | Tsutsum et al. (1996)62 | Japan | Case-control | 215 | 50 | Hepatocellular carcinoma | 92 (43%) |
40 | Yan et al. (2015)63 | China | Case-control | 274 | 51 | Hepatocellular carcinoma | NR |
41 | Wang et al. (2015)64 | China | Case-control | 120 | 45 | Hepatocellular carcinoma | 60 (50%) |
42 | Hao et al. (2015)65 | China | Case-control | 170 | 50 | Hepatocellular carcinoma | 85 (50%) |
43 | Shi et al. (2013)66 | China | Case-control | 160 | 49 | Hepatocellular carcinoma | 80 (50%) |
44 | Sun et al. (2015)67 | China | Case-control | 147 | 52 | Hepatocellular carcinoma | 110 (75%) |
45 | Liao et al. (2014)68 | China | Case-control | 318 | 55 | Hepatocellular carcinoma | 241 (76%) |
Alcohol consumption increases the risk of cirrhosis in a dose-dependent manner
Patients with cirrhosis were included in the analysis of the association between alcohol consumption and the risk of cirrhosis. Compared with non-drinkers, the pooled OR for the risk of cirrhosis in drinkers was 2.61 (95% CI, 1.46–4.66; I2 = 94%, p < 0.001) (Fig. 2A). Subgroup analyses showed that drinkers had an increased risk of cirrhosis, with ORs of 2.70 (95% CI: 1.43–5.11; I2 = 93%, p < 0.001) and 2.15 (95% CI: 0.30–15.64; I2 = 97%, p < 0.001) in studies with low and moderate risk of bias, respectively (Supplementary Fig. 1). Additionally, the ORs were 3.11 (95% CI: 1.61–6.10; I2 = 94%, p < 0.001) and 1.76 (95% CI: 0.50–6.19; I2 = 94%, p < 0.001) for patients aged ≥50 years and <50 years, respectively (Supplementary Fig. 2), and 3.00 (95% CI: 1.73–5.20; I2 = 94%, p < 0.001) and 0.40 (95% CI: 0.14–1.17) for studies published after and before 2000, respectively (Supplementary Fig. 3).
Compared with low-level drinkers, the estimated overall OR for the risk of cirrhosis in high-level drinkers was 2.34 (95% CI: 1.59–3.44; I2 = 87%, p < 0.001) (Supplementary Fig. 4). Subgroup analyses showed that high-level drinkers had an increased risk of cirrhosis, with ORs of 2.36 (95% CI: 1.34–4.15; I2 = 90%, p < 0.001) and 2.31 (95% CI: 1.53–3.49; I2 = 60%, p = 0.02) in studies with low and moderate risk of bias, respectively (Supplementary Fig. 5). Additionally, the ORs were 2.39 (95% CI: 1.53–3.71; I2 = 88%, p < 0.001) and 2.32 (95% CI: 0.95–5.70; I2 = 82%, p < 0.001) for patients aged ≥50 years and <50 years, respectively (Supplementary Fig. 6), and 2.29 (95% CI: 1.54–3.38; I2 = 87%, p < 0.001) and 5.28 (95% CI: 0.23–122.54; I2 = 87%, p = 0.005) for patients from Asia and Europe, respectively (Supplementary Fig. 7). Furthermore, the ORs were 2.04 (95% CI: 1.12–3.72; I2 = 83%, p < 0.001) and 6.88 (95% CI: 4.65–10.16; I2 = 90%, p < 0.001) for patients with an alcohol consumption history of ≥5 years and <5 years, respectively (Supplementary Fig. 8), and 2.75 (95% CI: 1.86–4.07; I2 = 88%, p < 0.001) and 0.87 (95% CI: 0.40–1.87; I2 = 59%, p = 0.09) for studies published after and before 2000, respectively (Supplementary Fig. 9).
Alcohol consumption increases the risk of HCC in a dose-dependent manner
Patients with HCC were included in the analysis of the association between alcohol consumption and the risk of HCC. Compared with non-drinkers, the estimated overall OR for the risk of HCC was 2.27 (95% CI: 1.50–3.43; I2 = 90%, p < 0.001) for drinkers (Fig. 2B). Subgroup analyses showed that drinkers had an increased risk of HCC, with ORs of 2.15 (95% CI: 1.37–3.36; I2 = 89%, p < 0.001) and 3.44 (95% CI: 1.17–10.09; I2 = 92%, p < 0.001) in studies with low and moderate risk of bias, respectively (Supplementary Fig. 10). In addition, the ORs were 2.41 (95% CI: 1.63–3.55; I2 = 87%, p < 0.001) and 1.64 (95% CI: 0.28–9.68; I2 = 97%, p < 0.001) for patients aged ≥50 years and <50 years, respectively (Supplementary Fig. 11), and 2.26 (95% CI: 1.47–3.48; I2 = 91%, p < 0.001) and 2.27 (95% CI: 1.50–3.43) for studies published after and before 2000, respectively (Supplementary Fig. 12).
Compared with low-level drinkers, the estimated overall OR for the risk of HCC was 2.42 (95% CI: 1.90–3.09; I2 = 80%, p < 0.001) for high-level drinkers (Supplementary Fig. 13). Subgroup analyses showed that high-level drinkers had an increased risk of HCC, with ORs of 2.40 (95% CI: 1.78–3.23; I2 = 82%, p < 0.001) and 2.51 (95% CI: 1.74–2.04; I2 = 48%, p = 0.08) in studies with low and moderate risk of bias, respectively (Supplementary Fig. 14); 2.33 (95% CI: 1.72–3.15; I2 = 81%, p < 0.001) and 2.66 (95% CI: 1.77–3.99; I2 = 67%, p = 0.002) for patients aged ≥50 years and <50 years, respectively (Supplementary Fig. 15); and 2.46 (95% CI: 1.90–3.18; I2 = 81%, p < 0.001) and 2.01 (95% CI: 1.15–3.50; I2 = 0.0%, p = 0.950) for patients from Asia and Europe, respectively (Supplementary Fig. 16). The ORs were 2.45 (95% CI: 1.89–3.18; I2 = 81%, p < 0.001) and 2.13 (95% CI: 1.45–3.13; I2 = 0.0%, p = 0.92) for case–control and cohort studies, respectively (Supplementary Fig. 17); 2.59 (95% CI: 2.01–3.34; I2 = 82%, p < 0.001) and 1.44 (95% CI: 0.40–1.87; I2 = 45%, p = 0.14) for studies published after and before 2000, respectively (Supplementary Fig. 18); and 1.75 (95% CI: 0.94–3.28) and 3.74 (95% CI: 1.89–7.42; I2 = 83%, p < 0.001) for patients with an alcohol consumption history of ≥5 years and <5 years, respectively (Supplementary Fig. 19).
Dose-dependent model of the effects of alcohol consumption on the risks of cirrhosis and HCC
The dose-dependent model of the effect of pure alcohol consumption (grams per day) on the incidence of cirrhosis was calculated using the formula log (OR) = 0.004954 × (alcohol intake per day) with p = 0.0318, as determined by the least squares method. According to the formula, each daily consumption of 12 g of alcohol would increase the risk of cirrhosis attributable to HBV infection by 6.2% (Fig. 3A). Meanwhile, the dose-dependent model of the effect of pure alcohol consumption on the incidence of HCC was calculated using the formula log (OR) = 0.009135 × (alcohol intake per day) with p = 0.0005. According to this formula, each daily consumption of 12 g of alcohol would increase the risk of HCC attributable to HBV infection by 11.5% (Fig. 3B).
Publication bias and sensitivity analysis
Funnel plots and Egger’s test showed evidence of some publication biases (Supplementary Figs. 20–23 and Supplementary Table 5), and the sensitivity analysis indicated that the results were reliable (Supplementary Figs. 24, 25). We performed sensitivity analyses excluding studies with fewer than 200 patients, considering the total number of patients in the included studies, and those published before 2000, considering the years when nucleoside or nucleotide drugs were developed, and observed that the results had robust reliability (Supplementary Figs. 26–29). In addition, we performed a trim-and-fill analysis to further test the stability and reliability of the results and found that the existence of publication bias did not affect the overall results (Supplementary Fig. 30).
Meta-regression analysis
The age, race, alcohol consumption duration of patients, study design, publication year of the study, and risk of bias were not significantly associated with the heterogeneity of the results in our study, except for publication year, which was a significant contributor to heterogeneity in the results related to patients with liver cirrhosis (p < 0.05) when compared with high-level alcohol consumption, as determined by meta-regression analysis (Supplementary Tables 6–9).
Discussion
Alcohol consumption promotes the progression of chronic liver diseases, but the exact quantitative dose-dependent effects of alcohol on the development of cirrhosis and HCC have not been well demonstrated. Our meta-analysis is the first report to show that alcohol consumption increases the risk of cirrhosis and HCC in patients with HBV infection in a dose-dependent manner. Importantly, we established a dose-dependent model of the effects of alcohol consumption on the risks of cirrhosis and HCC, which revealed that each daily consumption of 12 g of alcohol increases the risk of cirrhosis by 6.2% and the risk of HCC by 11.5%.
Alcohol consumption is a leading risk factor for various diseases with serious consequences, accounting for nearly 10% of deaths in populations aged 15–49 years worldwide.69 Over the past decades, the percentage of drinkers has increased from 45% in 1990 to 47% in 2017 globally.70 A modeling study from the United States predicted that the age-standardized death rate related to alcohol-associated liver disease would increase from 8.2 deaths per 100,000 patient-years in 2019 to 15.2 per 100,000 patient-years by 2040 if the current trends in alcohol consumption continued.71 After the COVID-19 pandemic in 2019, the frequency of alcohol consumption further increased by 14% due to various factors such as psychological, economic, and social environments.72 A recent study suggested that the increase in alcohol consumption during the COVID-19 pandemic could increase the long-term morbidity and mortality of alcohol-associated liver diseases.73 In the present meta-analysis, we demonstrated that alcohol consumption was closely associated with the occurrence of cirrhosis or HCC in patients with chronic HBV infection. The increased risk attributable to alcohol consumption was independent of age and the quality rating of the included studies for cirrhosis and independent of age, country of the patient population, study design, and quality rating of the included studies for HCC. These findings suggest that alcohol consumption promotes the development of cirrhosis and HCC in patients with chronic HBV infection regardless of their age or the quality rating of the included studies. Mechanistically, alcohol-mediated oxidative stress might contribute to the progression of liver cirrhosis and HCC74 via the nuclear factor kappa B pathway, which plays an important role in liver injury and regeneration.75 Furthermore, a systematic review and meta-analysis showed that active alcohol abstinence might decrease the risk of HCC in patients with alcohol-related diseases.76
Next, we determined the potential effects of alcohol doses on the incidence of cirrhosis and HCC in patients with HBV infection. We first demonstrated that alcohol consumption increased the risks of cirrhosis and HCC in a dose-dependent manner. In particular, the risks of cirrhosis and HCC were higher in high-level drinkers than in low-level drinkers. Importantly, we established a dose-dependent model of the effects of alcohol consumption on the incidence of cirrhosis and HCC. The model revealed that each daily consumption of 12 g of alcohol increased the risk of cirrhosis by 6.2% and the risk of HCC by 11.5%. The results are highly consistent with a systematic review of 19 cohort studies published in 2014, which showed that the consumption of more than three alcoholic drinks would increase the risk of HCC by 16% among alcohol drinkers.77 Therefore, attention should be paid to the effect of alcohol consumption on the development of cirrhosis and HCC in patients with HBV infection, in addition to genetic vulnerability and metabolic risk factors, all of which are known to contribute to the progression of chronic HBV infection to liver cirrhosis or HCC.78 Patients who consume more than 12 g of alcohol per day should be strictly monitored in the management of HBV infection.
These results should be carefully considered in light of several limitations. First, most of the 45 studies were case-control studies (n = 41, 91.1%), and some studies were published before 2000 (n = 6, 14.6%). The effects of potential biases, including memory bias, changes in diagnostic guidelines, and the discovery of nucleoside or nucleotide drugs, cannot be ruled out. Second, most of the 45 studies were from Asian countries (n = 41, 91.1%), with only four studies from Europe included in the present study, which may not be representative. Third, significant heterogeneity was present among the included studies in our meta-analysis, and the heterogeneity remained even when various subgroup analyses were conducted. The primary studies incorporated into our analysis lacked individual-level data, which compromised the precision of alcohol consumption quantification. This limitation was particularly evident considering the absence of detailed data on adolescent alcohol consumption—an important factor for liver disease—and information regarding periods of abstinence and the chronicity of alcohol intake. These gaps contributed substantially to the high heterogeneity observed in our study, a challenge that persisted even after conducting meta-regression analyses adjusted for age, race, and the methodological rigor of the literature. Despite these efforts, the sources of heterogeneity remained elusive and could not be effectively pinpointed. Therefore, we performed trim-and-fill analysis and sensitivity analysis, finding that the results of the meta-analysis were reliable. In addition, a dose-response meta-analysis could not be performed because only two groups of alcohol doses were present in 76% of the included studies. To establish an appropriate dose-dependent model, we chose a linear dose-dependent model rather than a quadratic or mixed model based on the Akaike information criterion and Bayesian information criterion statistics. In our analysis, we noted that the included studies did not offer detailed quantification of alcohol consumption, particularly for alcohol consumption in adolescents—an important factor for liver disease. Furthermore, the studies lacked comprehensive data on abstinence periods and the chronicity of alcohol consumption. Given the absence of individual-level data, our meta-analysis had limited ability to draw conclusions at the individual level. Instead, we extrapolated our findings to address population-level implications, focusing on broader public health and epidemiological perspectives rather than individual outcomes.
Supporting information
Supplementary Table 1
Electronic search strategy.
(DOCX)
Supplementary Table 2
Characteristics of studies included in the meta-analysis.
(DOCX)
Supplementary Table 3
Results of quality assessment using the Newcastle-Ottawa Scale for case-control studies.
(DOCX)
Supplementary Table 4
Results of quality assessment using the Newcastle-Ottawa Scale for cohort study studies.
(DOCX)
Supplementary Table 5
Egger test.
(DOCX)
Supplementary Table 6
Meta-regression of HCC for high level of drinkers versus low level of drinkers.
CI, confidence interval; HCC: Hepatocellular carcinoma.
(DOCX)
Supplementary Table 7
Meta-regression of cirrhosis for high level of drinkers versus low level of drinkers.
CI, confidence interval; HCC: Hepatocellular carcinoma.
(DOCX)
Supplementary Table 8
Meta-regression of HCC for drinkers versus non-drinkers.
CI, confidence interval; HCC: Hepatocellular carcinoma.
(DOCX)
Supplementary Table 9
Meta-regression of cirrhosis for drinkers versus non-drinkers.
CI, confidence interval.
(DOCX)
Supplementary Fig. 1
Forest plot of subgroup analysis for cirrhosis for drinkers versus non-drinkers.
Forest plot of subgroup accounting for study quality of studies included. Studies are named by author and year of publication.3–6,7–14
(TIF)
Supplementary Fig. 2
Forest plot of subgroup analysis for cirrhosis for drinkers versus non-drinkers.
Forest plot of subgroup analysis accounting for age of patients. OR; odds ratio; CI, confidence interval. Studies are named by author and year of publication.3–6,7–14
(TIF)
Supplementary Fig. 3
Forest plot of subgroup analysis for cirrhosis for drinkers versus non-drinkers.
Forest plot of subgroup analysis for publication year of the included studies. OR, odds ratio; CI, confidence interval. Studies are named by author and year of publication.3–6,7–14
(TIF)
Supplementary Fig. 4
Forest plot of subgroup analysis for cirrhosis for high level of drinkers versus low level of drinkers.
Forest plot for pooled OR of cirrhosis for high level drinkers versus low level drinkers. OR, odds ratio; CI, confidence interval. Studies are named by author and year of publication.4,6,7,10,12,25,26,29,31–34,70,72,76–78
(TIF)
Supplementary Fig. 5
Forest plot of subgroup analysis for cirrhosis for high level of drinkers versus low level of drinkers.
Forest plot of subgroup accounting for study quality of studies included. OR, odds ratio; CI, confidence interval. Studies are named by author and year of publication.4,6,7,10,12,25,26,29,31–34,70,72,76–78
(TIF)
Supplementary Fig. 6
Forest plot of subgroup analysis for cirrhosis for high level of drinkers versus low level of drinkers.
Forest plot of subgroup analysis accounting for age of patients. OR, odds ratio; CI, confidence interval. Studies are named by author and year of publication. 4,6,7,10,12,25,26,29,31–34,70,72,76–78
(TIF)
Supplementary Fig. 7
Forest plot of subgroup analysis for cirrhosis for high level of drinkers versus low level of drinkers.
Forest plot of subgroup accounting for race of studies included. OR, odds ratio; CI, confidence interval. Studies are named by author and year of publication. 4,6,7,10,12,25,26,29,31–34,70,72,76–78
(TIF)
Supplementary Fig. 8
Forest plot of subgroup analysis for cirrhosis accounting for alcohol intake times for high level drinkers versus low level.
Forest plot of subgroup accounting for alcohol times, OR, odds ratio; CI, confidence interval HCC: Hepatocellular carcinoma. Studies are named by author and year of publication.2,4,7,29,32,33,69,70,76
(TIF)
Supplementary Fig. 9
Forest plot of subgroup analysis for cirrhosis high level drinkers versus low level.
Forest plot of subgroup analysis for publication year of the included studies, OR, odds ratio; CI, confidence interval HCC: Hepatocellular carcinoma. Studies are named by author and year of publication. 4,6,7,10,12,25,26,29,31–34,70,72,76–78
(TIF)
Supplementary Fig. 10
Forest plot of subgroup analysis for HCC for drinkers versus non-drinkers.
Forest plot of subgroup accounting for study quality of studies included. OR, odds ratio; CI, confidence interval. HCC: Hepatocellular carcinoma. Studies are named by author and year of publication.4,6–8,10–18,21–23
(TIF)
Supplementary Fig. 11
Forest plot of subgroup analysis for HCC for drinkers versus non-drinkers.
Forest plot of subgroup analysis accounting for age of patients. OR, odds ratio; CI, confidence interval. HCC: Hepatocellular carcinoma. Studies are named by author and year of publication.4,6–8,10–18,21–23
(TIF)
Supplementary Fig. 12
Forest plot of subgroup analysis for HCC for drinkers versus non-drinkers.
Forest plot of subgroup analysis accounting for publication year of the included studies. OR, odds ratio; CI, confidence interval. HCC: Hepatocellular carcinoma. Studies are named by author and year of publication.4,6–8,10–18,21–23
(TIF)
Supplementary Fig. 13
Forest plot of subgroup analysis for HCC for high level of drinkers versus low level of drinkers.
Forest plot for pooled OR of HCC for high level drinkers versus low level drinkers. OR, odds ratio; CI, confidence interval. HCC: Hepatocellular carcinoma. Studies are named by author and year of publication.2,4,6–8,10,23–27,29,31–35,69–73,75–77
(TIF)
Supplementary Fig. 14
Forest plot of subgroup analysis for HCC for high level of drinkers versus low level of drinkers.
Forest plot of subgroup accounting for study quality of studies included. OR, odds ratio; CI, confidence interval. HCC: Hepatocellular carcinoma. Studies are named by author and year of publication. 2,4,6–8,10,23–26,29,31–35,69–73,75–77,27
(TIF)
Supplementary Fig. 15
Forest plot of subgroup analysis for HCC for high level of drinkers versus low level of drinkers.
Forest plot of subgroup analysis accounting for age of patients. OR, odds ratio; CI, confidence interval. HCC: Hepatocellular carcinoma. Studies are named by author and year of publication. 2,4,6–8,10,23–27,29,31–35,69–73,75–77
(TIF)
Supplementary Fig. 16
Forest plot of subgroup analysis for HCC for high level of drinkers versus low level of drinkers.
Forest plot of subgroup accounting for race of studies included. OR, odds ratio; CI, confidence interval. HCC: Hepatocellular carcinoma. Studies are named by author and year of publication. 2,4,6–8,10,23–27,29,31–35,69–73,75–77
(TIF)
Supplementary Fig. 17
Forest plot of HCC for high level drinkers versus low level drinkers.
Forest plot of subgroup analysis accounting for study design of the included studies. OR, odds ratio; CI, confidence interval. HCC: Hepatocellular carcinoma. Studies are named by author and year of publication. 2,4,6–8,10,23–27,29,31–35,69–73,75–77
(TIF)
Supplementary Fig. 18
Forest plot of subgroup analysis for HCC high level drinkers versus low level.
Forest plot of subgroup analysis for publication year of the included studies, OR, odds ratio; CI, confidence interval HCC: Hepatocellular carcinoma. Studies are named by author and year of publication. 2,4,6–8,10,23–27,29,31–35,69–73,75–77
(TIF)
Supplementary Fig. 19
Forest plot of subgroup analysis accounting for alcohol intake times for high level drinkers versus low level.
Forest plot for study-specific and pooled OR of HCC OR, odds ratio; CI, confidence interval HCC: Hepatocellular carcinoma. Studies are named by author and year of publication.2,4,7,22,29,32,33,69,70,76,77
(TIF)
Supplementary Fig. 20
Funnel plot The dashed lines represent the pseudo 95% CI.
The sold line represents the effect estimate, the included studies are shown by the dots in dark blue. Log (OR) is plotted in the horizontal axis and the standard error of log (OR) is plotted in the vertical axis. (A) Funnel plot of cirrhosis for drinkers versus non-drinkers. (B) Funnel plot of cirrhosis for high level drinkers versus low level drinkers. OR, odds ratio; CI, confidence interval; HCC: Hepatocellular carcinoma.
(TIF)
Supplementary Fig. 21
Funnel plot (A) Funnel plot of HCC for drinkers versus non-drinkers. (B) Funnel plot of HCC for high level drinkers versus low level drinkers.
OR, odds ratio; CI, confidence interval; HCC: Hepatocellular carcinoma.
(TIF)
Supplementary Fig. 22
Egger test (A) Egger test of cirrhosis for drinkers versus non-drinkers. (B) Egger test of cirrhosis for high level drinkers versus low level drinkers.
OR, odds ratio; CI, confidence interval; HCC: Hepatocellular carcinoma.
(TIF)
Supplementary Fig. 23
Egger test (A) Egger test of HCC for drinkers versus non-drinkers. (B) Egger test of HCC for high level drinkers versus low level drinkers.
OR, odds ratio; CI, confidence interval; HCC: Hepatocellular carcinoma.
(TIF)
Supplementary Fig. 24
Sensitivity analysis.
The vertical line in the middle represents the total pooled effect size, and the two vertical lines on either side represent the 95% CI range for the total effect size. The points corresponding to each study represent the pooled effect size of the remaining studies after the study was deleted. The pooled OR is plotted in the horizontal axis and the names and publication years of the included studies are plotted in the vertical axis. (A) Sensitivity analysis of cirrhosis for drinkers versus non-drinkers. (B) Sensitivity analysis of cirrhosis for high level drinkers versus low level drinkers. OR, odds ratio; CI, confidence interval; HCC: Hepatocellular carcinoma. Studies are named by author and year of publication.3–13,25,26,29,31–34,70,72,76–78
(TIF)
Supplementary Fig. 25
Sensitivity analysis (A) Sensitivity analysis of HCC for drinkers versus non-drinkers. (B) Sensitivity analysis of HCC for high level of drinkers versus low level of drinkers.
OR, odds ratio; CI, confidence interval; HCC: Hepatocellular carcinoma. Studies are named by author and year of publication.2,4,6–8,10–18,21–27,29,31–35,69–73,75–77
(TIF)
Supplementary Fig. 26
Sensitivity analysis (A) Sensitivity analysis of cirrhosis (exclude less than 200 cases studies) for drinkers versus non-drinkers. (B) Sensitivity analysis of cirrhosis (exclude less than 200 cases studies) for high level of drinkers versus low level of drinkers.
OR, odds ratio; CI, confidence interval; HCC: Hepatocellular carcinoma. Studies are named by author and year of publication.4,6–8,10–20,22–27,29,31–35,69,71,73
(TIF)
Supplementary Fig. 27
Sensitivity analysis Sensitivity analysis of cirrhosis (exclude studies published before the year 2000) for high level of drinkers versus low level of drinkers.
OR, odds ratio; CI, confidence interval; HCC: Hepatocellular carcinoma. Studies are named by author and year of publication.4,6–8,29,71–73,76
(TIF)
Supplementary Fig. 28
Sensitivity analysis (A) Sensitivity analysis of HCC (exclude less than 200 cases studies) for high level of drinkers versus low level of drinkers. (B) Sensitivity analysis of HCC (exclude less than 200 cases studies) for drinkers versus non-drinkers.
OR, odds ratio; CI, confidence interval; HCC: Hepatocellular carcinoma. Studies are named by author and year of publication.2,4,6–8,11,13,15–18,22,23,29,35,71–73,75,76
(TIF)
Supplementary Fig. 29
Sensitivity analysis (A) Sensitivity analysis of HCC (exclude published before 2000 studies) for high level of drinkers versus low level of drinkers. (B) Sensitivity analysis of HCC (exclude published before 2000 studies) for drinkers versus non- drinkers.
OR, odds ratio; CI, confidence interval; HCC: Hepatocellular carcinoma. Studies are named by author and year of publication.2,4,6–8,11,13–23,26,31–35,70,72,73,75–77
(TIF)
Supplementary Fig. 30
Trim and Fill analysis.
(A) Trim and fill analysis of HCC for high level drinkers versus low level drinkers. (B) Trim and fill analysis of HCC with cirrhosis for high -level of drinkers versus low-level of drinkers. OR, odds ratio; CI, confidence interval; HCC: Hepatocellular carcinoma.
(TIF)