Meta-Analysis
Copyright ©2013 Baishideng Publishing Group Co.
World J Gastroenterol. Sep 28, 2013; 19(36): 6098-6107
Published online Sep 28, 2013. doi: 10.3748/wjg.v19.i36.6098
Table 1 Characteristics of literatures included in the meta-analysis
Ref.Study areaYearStudy typeHP DmHp+definition1ESCC
EAC
Matched/adjusted3Control
Case
Control
Case
Control
Hp+CagA+Hp+CagA+Hp+CagA+Hp+CagA+
Murphy et al[7]Finnish2012PopSHpSe+64/8235/8263/8236/82----Yes/yesMatched with age, smoking and Alcohol consumption, and date of blood draw to controls, all of cases and controls were male smokers
Khoshbaten et al[9]Iran2011PopSHpSe+58/10028/10083/10036/100----Yes/noSex-matched and age-matched health individuals, any clinical evidence of gastrointestinal symptoms were excluded
Venerito et al[8]Germany2011ClinS, H, UHis+, HpSe+, CagA+ or U+53/7542/7553/7540/75----Yes/noSex and age-matched individuals with upper GI symptoms but no malignancy of the upper GI tract
Whiteman et al[11]Australia2010PopSHpSe+54/208-302/1316-35/260-302/1316-Yes/yesRandomly selected from the same areas, matched to each stratum of age and state
Cook et al[10]Finnish2010PopSHpSe+64/7835/7871/9142/91----Yes/yesMatched to case for age, years of smoking, cigarettes per day, or body mass index
Wu et al[5]Taiwan2009PopSHpSe+112/31791/317563/1103268/700----No/yesOne part of control is matched by gender and age, but another part wasn't matched
Hu et al[6]Taiwan2009PopSHpSe+66/180-102/194-----Yes/yesHealthy and cancer-free individuals, matched to age, sex and ethnicity
Früh et al[12]Canada2008ClinSCagA + or VacA+----36/10029/10043/10130/101Yes/yesHealthy GRERD-free, non-blood-related family member and friends of other cancer/surgical patients
Derakhshan et al[24]Iran2008ClinSHpSe+----9/19-28/38-Yes/yesDyspeptic patients with no peptic ulcer or tumor in their endoscopy
Anderson et al[25]Ireland2008PopSHpSe+----55/12357/123157/253150/253Yes/yesRandomly selected population-based controls, frequency matched to EAC cases for age and sex
Löfdahl et al[23]Sweden2008PopSHpSe+ or CagA+----130/230-304/499-Yes/noRandom selected from the population register, frequency matched for age and sex
Anandasabapathy et al[26]United States2007ClinHHis+----4/25-10/30-No/noBarrett’s patients with no dysplasia
Iijima et al[27]Japan2007ClinS, H, UHpSe+, His+ or U+60/73-56/73-----Yes/yesEndoscoped patients with no localized lesion, matched to cases for age and sex
Kamangar et al[14]China2007PopSHpSe+231/335178/335662/992552/992----No/yesRandomly selected from the entire baseline participants in the stdy cohort
Simán et al[13]Sweden2007PopSHpSe+15/3724/3768/12982/1294/126/1224/4732/47Yes/yesRandomly selected from the study cohorted, matched with age, sex, and date enrollment
Wang et al[28]China2006PopSHpSe+?/107-?/107-----Yes/yesNeighborhood controls, randomly selected, and matched to cases for age and gender
Wu et al[15]Taiwan2005PopSHpSe+28/127-74/171-----No/yesRandomly selected from the same community
de Martel et al[29]United States2005PopSHpSe+----19/519/1874/15044/71Yes/yesRandomly selected from the study cohorted, matched with age, sex, and date enrollment race, and study site
Ye et al[16]Sweden2004PopSHpSe+32/8563/85198/499293/49918/9742/97198/499293/499Yes/yesRandomly selected population-based controls, frequency matched to EAC cases for age and sex
2Wang et al[30]China2003PopSHpSe+33/63-145/310-----Yes/noHealthy subjects with no difference in age and gender
Wu et al[22]United States2003PopSHpSe+----49/8018/80230/356106/356Yes/yesMatched to cases for age, sex, neighborhood of residence, and race
El Omar et al[21]United States2003PopSHpSe+31/537/2684/21046/22435/1085/6884/21046/224Yes/noMatched to cases for age, sex, and study center
Weston et al[20]United States2000ClinHHis+----3/20-96/217-No/noPatients with GERD symptoms but no Barrett’s esophagus
Vieth et al[19]Germany2000ClinHHis+----66/138-468/712-No/noPatients with non-ulcer dyspepsia and no endoscopic signs of GERD
Peek et al[18]United States1999ClinS, HHpSe+ or His+----11/303/3020/4825/48No/noPatients endoscoped for reasons other Than GERD or Barrett’s
Oberg et al[31]United States1999ClinHHis+----5/37-32/229-No/noPatients with foregut symptoms and benign diseases
Talley et al[17]United States1991ClinSHpSe+20/41-96/252-----No/noAsymptomatic volunteers and patients with benign esophageal, lung, or culoskeletal disorders
Table 2 Meta-analysis of the Helicobacter pylori infection on the risk of esophageal squamous cell carcinoma and esophageal adenocarcinoma
StudiesP1I2,2Overall OR (95%CI)
Esophageal squamous cell carcinoma
Case/control (1961/5704)
All studies16< 0.0174.50%0.83 (0.63, 1.03)
Population-based studies14< 0.0176.00%0.79 (0.59, 1.00)
Clinical-based studies20.86< 0.011.49 (0.66, 2.31)
Eastern studies8< 0.0179.50%0.66 (0.43, 0.89)
Western studies80.421.20%1.02 (0.80, 1.25)3
Studies with matched controls11< 0.0171.80%0.90 (0.61, 1.20)
Studies without matched controls5< 0.0182.90%0.79 (0.46, 1.12)
Hp+ only definition as HpSe+14< 0.0176.80%0.81 (0.60, 1.02)
Adjusted results11< 0.0180.50%0.84 (0.56, 1.12)
CagA+vs Hp-90.0352.00%0.97 (0.76, 1.24)
Eastern study30.2235.00%0.77 (0.65, 0.92)3
Western studies60.393.60%1.26 (0.97, 1.63)3
Esophageal adenocarcinoma
Case/control (1330/4705)
All studies150.13129.90.59 (0.51, 0.68)
Population-based studies80.10640.90.62 (0.52, 0.73)
Clinical-based studies70.31914.50.53 (0.40, 0.68)
Eastern study1---
Western studies140.128310.60 (0.52, 0.68)
Studies with matched controls100.13933.60.62 (0.53, 0.72)
Studies without matched controls50.33312.70.49 (0.36, 0.66)
Hp+ definition as HpSe+80.29916.60.55 (0.45, 0.66)
Hp+ definition as His+40.33411.70.46 (0.33, 0.64)
Adjusted results80.20028.60.51 (0.40, 0.61)
CagA+vs Hp-80.1139.90.56 (0.45, 0.70)
Eastern study0---
Western studies80.1139.90.56 (0.45, 0.70)