Alimentary Pharmacology and Therapeutics
Systematic review with meta-analysis: meat consumption and the risk of hepatocellular carcinoma J. Luo, Y. Yang, J. Liu, K. Lu, Z. Tang, P. Liu, L. Liu & Y. Zhu
Department of Radiation Therapy, Zhejiang Cancer Hospital, Hangzhou, China.
Correspondence to: Prof. Y. Zhu, Department of Radiation Therapy, Zhejiang Cancer Hospital, No. 38, Guangji Road, Hangzhou, 310022, P.R. China. E-mail: [email protected]
Publication data Submitted 27 October 2013 First decision 18 November 2013 Resubmitted 6 February 2014 Accepted 9 February 2014 EV Pub Online 3 March 2014 As part of AP&T’s peer-review process, a technical check of this meta-analysis was performed by Dr P. Collins. This article was accepted for publication after full peer-review.
SUMMARY Background The association between meat consumption and the risk of hepatocellular carcinoma (HCC) is still inconclusive. Aim To conduct a systematic review and meta-analysis to quantitatively assess the relationship between meat intake and the risk of HCC. Methods We searched the PubMed, Web of Science and EMBASE databases for relevant studies published before July 2013. The summary relative risks were pooled using the fixed-effects model when no substantial heterogeneity was detected, otherwise, the random-effects model was used. Heterogeneity and publication bias were also analysed. Results Finally, seven cohort studies and 10 case–control studies were included. The pooled relative risks (RRs) of HCC for the highest vs. lowest consumption levels were 1.10 (95% confidence interval, CI: 0.85–1.42) for red meat, 1.01 (95% CI: 0.79–1.28) for processed meat and 0.97 (95% CI: 0.85–1.11) for total meat. Moreover, white meat and fish consumption were found to be inversely associated with HCC risk, the summary RRs were 0.69 (95% CI: 0.58–0.81) and 0.78 (95% CI: 0.67–0.90) respectively, and the results remained quite stable after stratification by the confounding factors. Conclusions The present meta-analysis indicates that a high level of white meat or fish consumption can reduce the risk of HCC significantly, while intake of red meat, processed meat or total meat is not associated with HCC risk. Our findings suggest that dietary intervention may be a promising approach for prevention of HCC, which still need to be confirmed by further welldesigned prospective studies and experimental research. Aliment Pharmacol Ther 2014; 39: 913–922
ª 2014 John Wiley & Sons Ltd doi:10.1111/apt.12678
913
J. Luo et al. INTRODUCTION Hepatocellular Carcinoma (HCC) is one of the most common cancers worldwide, especially in Eastern Asia and sub-Saharan Africa, most HCC cases are diagnosed at advanced stages and the prognosis is poor.1 The major risk factors for HCC are hepatitis B or hepatitis C infection, cirrhosis and aflatoxin contamination, other factors such as alcohol drinking, obesity, smoking and diabetes were also found to be associated with increased risk of HCC.2 Considering the rising incidence of HCC, especially in Europe and North America, identifying of new risk factors and protective factors is important for HCC prevention, which will provide new strategies to lower its incidence.3, 4 For the past few decades, a lot of the epidemiological studies have shown that dietary factors are closely correlated with cancer, a healthy diet is known to reduce the rates of some types of cancer, while a poor diet will increase cancer risk.5 As a main component of diet, meats, especially red meat and processed meat are found to be associated with the increased risk of certain types of cancer, including most gastrointestinal cancers, and even lung cancer and breast cancer.6–10 As to HCC, also one of gastrointestinal cancers, Nanji AA et al. firstly reported that pork consumption was related with a high mortality of HCC,11 then some other studies found similar results,12, 13 while some others found no such association.14–19 On the contrary, many investigators found that white meat or fish might decrease the risk of HCC.14, 16, 20, 21 However, up to now, there is still no quantitative assessment to examine the relationship between meat consumption and the risk of HCC. To better understand the issue, we therefore searched relevant publications from the biomedical databases, and performed a systematic review and meta-analysis to summarise the evidence of meat consumption (i.e. red meat, processed meat, fish, poultry and total meat) on the HCC risk, MATERIALS AND METHODS Publication selection We identified related studies by an electronic search in PubMed (Medline), Web of science and EMBASE databases from 1956 to July 2013, the key words ‘hepatocellular carcinoma’ or ‘hepatoma’ or ‘liver cancer’ or ‘liver tumor’ in combination with ‘meat’ or ‘red meat’ or ‘processed meat’ or ‘white meat’ or ‘poultry’ or ‘fish’ or ‘diet’ or ‘foods’ or ‘beef’ or ‘pork’ or ‘lamb’ or ‘goat’ were used, no language restriction was applied. Further, we reviewed 914
the references in the identified articles to find more relevant studies. Studies were included in the meta-analysis if they met the following criteria: (i) A cohort or case–control study with HCC incidence or mortality as the outcome; (ii) Relative risk estimates (odds ratio or hazard ratio estimates) and their corresponding 95% CIs of HCC were reported relating to each category of meat intake.
Data extraction and quality assessment For each study, the following information was extracted: first author’s name, publication year, country or region in which the study was conducted, duration of data collection, number of cases (or death cases), sample size, meat items, comparison of consumption levels and adjusted confounders. In our study, red meat was defined as ‘pork’ or ‘beef’ or ‘lamb’ or meat defined as ‘red meat’ in the individual studies. Processed meat was defined as ‘processed’, ‘preserved’ or ‘cured meat’, including any of the following types of meat: sausage, ham, bacon, hot dogs and salami. White meat definition included ‘poultry’, or ‘chicken’ or ‘poultry’ or meat defined as ‘white meat’ in the individual studies. While fish also included ‘fermented fish’ and ‘shrimp’ and the total meat in our analysis was referred to as ‘all meat’ without specifying the type, or ‘total meat’ defined in the individual studies. Concerning to the specified type of meat, if several articles were derived from the same study population, we extracted the data from the most recent publication. Data were extracted from each study by LJL and Yang Y independently, and further confirmed for accuracy by Zhu Y. Three independent investigators (Liu JJ, Lu K and Tang ZZ) performed quality assessment according to the Newcastle-Ottawa quality assessment scale, which is widely used in meta-analyses for nonrandomised studies.22 Briefly, each study was assigned a maximum of 9 points: 4 for selection, 2 for comparability, and 3 for outcomes (for cohort study) or exposures (for case–control study), scores of 0–3, 4–6, and 7–9 were regarded as low, moderate and high quality, respectively in our study. The scores provided by each investigator were compared and a consensus value was achieved by discussion if there was any difference. Statistical methods We extracted the study-specific relative risks and their corresponding 95% CIs of specific meat consumption (comparing the highest with the lowest category) for data synthesis, and transformed them to their natural logarithms to stabilise the variance and normalise the distributions.23 We estimated heterogeneity among Aliment Pharmacol Ther 2014; 39: 913-922 ª 2014 John Wiley & Sons Ltd
Systematic review with meta-analysis: meat consumption and risk of hepatocellular carcinoma studies by employing the Q and I2 statistics, for the Q statistic, statistical significance was set at P < 0.1, for the I2 value, 25%, 50% and 75% were considered to indicate low, medium and high levels of heterogeneity, respectively.24 We used the fixed-effects model (DerSimonian and Laird method) to pool the relative risks when no substantial heterogeneity was detected (I2 < 50%, P > 0.1), otherwise, the random-effects model (the inverse variance method) was used.25, 26 To assess the influence of specific studies on the pooled results, we performed a sensitivity analysis in which one study was removed and the combined estimates of the rest were recalculated. We evaluated publication bias using the Egger test27 and Begg’s rank correlation method,28 P < 0.1 indicated a statistically significant publication bias. If publication bias existed, ‘trim and fill’ method was used to correct such bias.29 All statistical analyses were performed by STATA12.0 (StataCorp, College Station, TX, USA).
RESULTS Literature search and study characteristics We identified 17 studies that met the inclusion criteria, including 7 cohort studies and 10 case–control studies, the
search process was summarised in Figure 1. The main characteristics of each study were presented in Table 1, a total of 1 670 093 participants and 4826 HCC cases were involved after combining all the studies. Among these studies, three were carried out in US, six in Europe, and eight in Asia. It is should be noted that relative risks for different types of meat were reported in different publications for the cohort of the NIH-AARP Diet and Health Study in US and the National Cancer Institute study in Italy. Most studies adjusted for potential confounding factors including age, sex, HBV infection, alcohol drinking, etc. Overall, the quality score ranged from 4 to 9, with an average score 6.76, and no study of low quality was found.
Red meat Among all the studies, 9 studies reported the association between red meat and HCC risk, including three cohort studies12, 14, 30 and six case–control studies.15–19, 21 After combining all the RRs, a pooled RR of 1.10 (95% CI: 0.85–1.42) was calculated with the random-effects model (Figure 2 and Table 2), indicating high consumption of red meat was not associated with increased HCC risk. The RRs were 0.97 (95% CI: 0.71–1.32) for case–control studies and 1.43 (95% CI: 1.08–1.90) for cohort studies, respectively.
Citations in PubMed (n = 2065) Web of Science (n = 3658) EMBASE (n = 1800)
After removing duplicates (n = 1351)
Records identified for further review (n = 4372) Excluded by title review (n = 4138) Abstracts (n = 234) Excluded by abstract review (n = 179) Articles for detailed evaluation (n = 55)
Excluded by full-text article review (n = 38) 3 were abstracts, 5 were reviews and 3 were editorials 11 did not report meat consumption as an exposure 16 without enough information on outcomes
Figure 1 | Flow chart of study selection. Aliment Pharmacol Ther 2014; 39: 913-922 ª 2014 John Wiley & Sons Ltd
Studies included in the meta-analysis (n = 17)
915
J. Luo et al. Table 1 | Main characteristics of studies on meat consumption and HCC risk Author, year [Ref]
Duration
Design
Location
Cases
Size/ Controls
Srivatanakul, 199119
Unreported
PCC
Thailand
65
Fernandez, 199932
1983–1996
PCC
Italy
Kuper, 200034
1995–1998
PCC
Tavani, 200018
1983–1991
Yu, 200217
Meat items
Comparison
Adjusted variables
QS
65
Fish and red meat
HBsAg status, alcohol, education and food items
4
428
7990
Fish
97
128
Total meat
HCC*
Italy
428
7990
Red meat
1995–1997
PCC
China
248
248
Fish, red meat and chicken
Age, sex, area of residence, education, smoking, alcohol consumption, and body mass index Age, gender, schooling, HBV/ HCV infection, alcohol consumption, smoking, total energy intake Age, year of recruitment, sex, education, smoking habits and alcohol, fat, fruit and vegetable intakes HBsAg, history of intravenous injection, average income
7
Greece
3/day > vs. vs.
Systematic review with meta-analysis: meat consumption and the risk of hepatocellular carcinoma J. Luo, Y. Yang, J. Liu, K. Lu, Z. Tang, P. Liu, L. Liu & Y. Zhu
Department of Radiation Therapy, Zhejiang Cancer Hospital, Hangzhou, China.
Correspondence to: Prof. Y. Zhu, Department of Radiation Therapy, Zhejiang Cancer Hospital, No. 38, Guangji Road, Hangzhou, 310022, P.R. China. E-mail: [email protected]
Publication data Submitted 27 October 2013 First decision 18 November 2013 Resubmitted 6 February 2014 Accepted 9 February 2014 EV Pub Online 3 March 2014 As part of AP&T’s peer-review process, a technical check of this meta-analysis was performed by Dr P. Collins. This article was accepted for publication after full peer-review.
SUMMARY Background The association between meat consumption and the risk of hepatocellular carcinoma (HCC) is still inconclusive. Aim To conduct a systematic review and meta-analysis to quantitatively assess the relationship between meat intake and the risk of HCC. Methods We searched the PubMed, Web of Science and EMBASE databases for relevant studies published before July 2013. The summary relative risks were pooled using the fixed-effects model when no substantial heterogeneity was detected, otherwise, the random-effects model was used. Heterogeneity and publication bias were also analysed. Results Finally, seven cohort studies and 10 case–control studies were included. The pooled relative risks (RRs) of HCC for the highest vs. lowest consumption levels were 1.10 (95% confidence interval, CI: 0.85–1.42) for red meat, 1.01 (95% CI: 0.79–1.28) for processed meat and 0.97 (95% CI: 0.85–1.11) for total meat. Moreover, white meat and fish consumption were found to be inversely associated with HCC risk, the summary RRs were 0.69 (95% CI: 0.58–0.81) and 0.78 (95% CI: 0.67–0.90) respectively, and the results remained quite stable after stratification by the confounding factors. Conclusions The present meta-analysis indicates that a high level of white meat or fish consumption can reduce the risk of HCC significantly, while intake of red meat, processed meat or total meat is not associated with HCC risk. Our findings suggest that dietary intervention may be a promising approach for prevention of HCC, which still need to be confirmed by further welldesigned prospective studies and experimental research. Aliment Pharmacol Ther 2014; 39: 913–922
ª 2014 John Wiley & Sons Ltd doi:10.1111/apt.12678
913
J. Luo et al. INTRODUCTION Hepatocellular Carcinoma (HCC) is one of the most common cancers worldwide, especially in Eastern Asia and sub-Saharan Africa, most HCC cases are diagnosed at advanced stages and the prognosis is poor.1 The major risk factors for HCC are hepatitis B or hepatitis C infection, cirrhosis and aflatoxin contamination, other factors such as alcohol drinking, obesity, smoking and diabetes were also found to be associated with increased risk of HCC.2 Considering the rising incidence of HCC, especially in Europe and North America, identifying of new risk factors and protective factors is important for HCC prevention, which will provide new strategies to lower its incidence.3, 4 For the past few decades, a lot of the epidemiological studies have shown that dietary factors are closely correlated with cancer, a healthy diet is known to reduce the rates of some types of cancer, while a poor diet will increase cancer risk.5 As a main component of diet, meats, especially red meat and processed meat are found to be associated with the increased risk of certain types of cancer, including most gastrointestinal cancers, and even lung cancer and breast cancer.6–10 As to HCC, also one of gastrointestinal cancers, Nanji AA et al. firstly reported that pork consumption was related with a high mortality of HCC,11 then some other studies found similar results,12, 13 while some others found no such association.14–19 On the contrary, many investigators found that white meat or fish might decrease the risk of HCC.14, 16, 20, 21 However, up to now, there is still no quantitative assessment to examine the relationship between meat consumption and the risk of HCC. To better understand the issue, we therefore searched relevant publications from the biomedical databases, and performed a systematic review and meta-analysis to summarise the evidence of meat consumption (i.e. red meat, processed meat, fish, poultry and total meat) on the HCC risk, MATERIALS AND METHODS Publication selection We identified related studies by an electronic search in PubMed (Medline), Web of science and EMBASE databases from 1956 to July 2013, the key words ‘hepatocellular carcinoma’ or ‘hepatoma’ or ‘liver cancer’ or ‘liver tumor’ in combination with ‘meat’ or ‘red meat’ or ‘processed meat’ or ‘white meat’ or ‘poultry’ or ‘fish’ or ‘diet’ or ‘foods’ or ‘beef’ or ‘pork’ or ‘lamb’ or ‘goat’ were used, no language restriction was applied. Further, we reviewed 914
the references in the identified articles to find more relevant studies. Studies were included in the meta-analysis if they met the following criteria: (i) A cohort or case–control study with HCC incidence or mortality as the outcome; (ii) Relative risk estimates (odds ratio or hazard ratio estimates) and their corresponding 95% CIs of HCC were reported relating to each category of meat intake.
Data extraction and quality assessment For each study, the following information was extracted: first author’s name, publication year, country or region in which the study was conducted, duration of data collection, number of cases (or death cases), sample size, meat items, comparison of consumption levels and adjusted confounders. In our study, red meat was defined as ‘pork’ or ‘beef’ or ‘lamb’ or meat defined as ‘red meat’ in the individual studies. Processed meat was defined as ‘processed’, ‘preserved’ or ‘cured meat’, including any of the following types of meat: sausage, ham, bacon, hot dogs and salami. White meat definition included ‘poultry’, or ‘chicken’ or ‘poultry’ or meat defined as ‘white meat’ in the individual studies. While fish also included ‘fermented fish’ and ‘shrimp’ and the total meat in our analysis was referred to as ‘all meat’ without specifying the type, or ‘total meat’ defined in the individual studies. Concerning to the specified type of meat, if several articles were derived from the same study population, we extracted the data from the most recent publication. Data were extracted from each study by LJL and Yang Y independently, and further confirmed for accuracy by Zhu Y. Three independent investigators (Liu JJ, Lu K and Tang ZZ) performed quality assessment according to the Newcastle-Ottawa quality assessment scale, which is widely used in meta-analyses for nonrandomised studies.22 Briefly, each study was assigned a maximum of 9 points: 4 for selection, 2 for comparability, and 3 for outcomes (for cohort study) or exposures (for case–control study), scores of 0–3, 4–6, and 7–9 were regarded as low, moderate and high quality, respectively in our study. The scores provided by each investigator were compared and a consensus value was achieved by discussion if there was any difference. Statistical methods We extracted the study-specific relative risks and their corresponding 95% CIs of specific meat consumption (comparing the highest with the lowest category) for data synthesis, and transformed them to their natural logarithms to stabilise the variance and normalise the distributions.23 We estimated heterogeneity among Aliment Pharmacol Ther 2014; 39: 913-922 ª 2014 John Wiley & Sons Ltd
Systematic review with meta-analysis: meat consumption and risk of hepatocellular carcinoma studies by employing the Q and I2 statistics, for the Q statistic, statistical significance was set at P < 0.1, for the I2 value, 25%, 50% and 75% were considered to indicate low, medium and high levels of heterogeneity, respectively.24 We used the fixed-effects model (DerSimonian and Laird method) to pool the relative risks when no substantial heterogeneity was detected (I2 < 50%, P > 0.1), otherwise, the random-effects model (the inverse variance method) was used.25, 26 To assess the influence of specific studies on the pooled results, we performed a sensitivity analysis in which one study was removed and the combined estimates of the rest were recalculated. We evaluated publication bias using the Egger test27 and Begg’s rank correlation method,28 P < 0.1 indicated a statistically significant publication bias. If publication bias existed, ‘trim and fill’ method was used to correct such bias.29 All statistical analyses were performed by STATA12.0 (StataCorp, College Station, TX, USA).
RESULTS Literature search and study characteristics We identified 17 studies that met the inclusion criteria, including 7 cohort studies and 10 case–control studies, the
search process was summarised in Figure 1. The main characteristics of each study were presented in Table 1, a total of 1 670 093 participants and 4826 HCC cases were involved after combining all the studies. Among these studies, three were carried out in US, six in Europe, and eight in Asia. It is should be noted that relative risks for different types of meat were reported in different publications for the cohort of the NIH-AARP Diet and Health Study in US and the National Cancer Institute study in Italy. Most studies adjusted for potential confounding factors including age, sex, HBV infection, alcohol drinking, etc. Overall, the quality score ranged from 4 to 9, with an average score 6.76, and no study of low quality was found.
Red meat Among all the studies, 9 studies reported the association between red meat and HCC risk, including three cohort studies12, 14, 30 and six case–control studies.15–19, 21 After combining all the RRs, a pooled RR of 1.10 (95% CI: 0.85–1.42) was calculated with the random-effects model (Figure 2 and Table 2), indicating high consumption of red meat was not associated with increased HCC risk. The RRs were 0.97 (95% CI: 0.71–1.32) for case–control studies and 1.43 (95% CI: 1.08–1.90) for cohort studies, respectively.
Citations in PubMed (n = 2065) Web of Science (n = 3658) EMBASE (n = 1800)
After removing duplicates (n = 1351)
Records identified for further review (n = 4372) Excluded by title review (n = 4138) Abstracts (n = 234) Excluded by abstract review (n = 179) Articles for detailed evaluation (n = 55)
Excluded by full-text article review (n = 38) 3 were abstracts, 5 were reviews and 3 were editorials 11 did not report meat consumption as an exposure 16 without enough information on outcomes
Figure 1 | Flow chart of study selection. Aliment Pharmacol Ther 2014; 39: 913-922 ª 2014 John Wiley & Sons Ltd
Studies included in the meta-analysis (n = 17)
915
J. Luo et al. Table 1 | Main characteristics of studies on meat consumption and HCC risk Author, year [Ref]
Duration
Design
Location
Cases
Size/ Controls
Srivatanakul, 199119
Unreported
PCC
Thailand
65
Fernandez, 199932
1983–1996
PCC
Italy
Kuper, 200034
1995–1998
PCC
Tavani, 200018
1983–1991
Yu, 200217
Meat items
Comparison
Adjusted variables
QS
65
Fish and red meat
HBsAg status, alcohol, education and food items
4
428
7990
Fish
97
128
Total meat
HCC*
Italy
428
7990
Red meat
1995–1997
PCC
China
248
248
Fish, red meat and chicken
Age, sex, area of residence, education, smoking, alcohol consumption, and body mass index Age, gender, schooling, HBV/ HCV infection, alcohol consumption, smoking, total energy intake Age, year of recruitment, sex, education, smoking habits and alcohol, fat, fruit and vegetable intakes HBsAg, history of intravenous injection, average income
7
Greece
3/day > vs. vs.
