Meta-Analysis
Copyright ©2014 Baishideng Publishing Group Co.
World J Gastroenterol. Feb 21, 2014; 20(7): 1858-1870
Published online Feb 21, 2014. doi: 10.3748/wjg.v20.i7.1858
Table 1 Randomized controlled trials included in the meta-analysis
StudyAgentAll subjectsDuration (yr)Incident colorectal cancer cases
RR95%CI
Statin groupControl group
JUPITER[43]Rosuvastatin16304Median: 2.016 of 815428 of 81500.570.31-1.05
WOSCOPS[44]Pravastatin6577Mean: 4.912 of 329120 of 32860.600.29-1.23
4S[47]Simvastatin4444Median: 10.425 of 222132 of 22230.780.46-1.32
ALLHAT-LLT[46]Pravastatin10355Mean: 4.8461 of 5170381 of 51851.210.79-1.86
HPS[45]Simvastatin20536Mean: 5.0114 of 10269131 of 102670.870.68-1.12
LIPID[48]Pravastatin9014Mean: 8.075 of 451271 of 45021.050.76-1.45
AFCAPS[5]Lovastatin6605Mean: 5.2251 of 3304201 of 33011.250.70-2.24
CARE[4]Pravastatin4159Mean: 4.812 of 208121 of 20780.570.28-1.16
1Figures for colon cancer rather than colorectal cancer.
Table 2 Cohort studies included in the meta-analysis
StudyStudy locationAll subjectsCRC casesRR95%CIControl for potential confounding factors2
Clancy et al[50], 2013Italy26610924200.840.76-0.931-9
Leung et al[49], 2013Taiwan342056540.570.45-0.721-
Simon et al[51], 2012United States15921920000.990.83-1.201, 4, 10-27
Jacobs et al[53], 2011United States13325517391.030.94-1.141-
Lee et al[52], 2011United States13192216800.970.84-1.121, 6, 10, 12, 13, 28-34
Haukka et al[55], 2010Finland94496250161.040.98-1.101-
Matsushita et al[54], 2010Japan13724761.220.77-1.94-
Flick et al[57], 2009United States691151710.890.61-1.301, 4, 6, 10-13, 15, 18, 19, 28, 35-38
Singh et al[56], 2009Canada41327166371.131.02-1.251-4, 6, 7, 16, 39-42
Friedman et al[58], 2008United States424306756840.880.82-0.961-
Farwell et al[59], 2008United States628426870.650.55-0.781, 6, 11, 13, 19, 29, 43-53
Setoguchi et al[60], 2007United States317232490.960.70-1.311-9, 39, 43, 54-64
Friis et al[61], 2005Denmark33475430060.850.65-1.111, 2, 4, 16, 66
1Calculated crude relative risk (RR);
21: Age; 2: Sex; 3: Inflammatory bowel disease; 4: Use of nonsteroidal anti-inflammatory drugs; 5: Obesity; 6: Colonoscopy; 7: Comorbidity score; 8: Distinct generic medicines taken; 9: Prior hospitalizations; 10: Body mass index; 11: Smoking status; 12: Family History of colorectal cancer; 13: Alcohol use; 14: Education; 15: Physical activity level; 16: Hormone replacement therapy; 17: Ethnic group; 18: Colorectal polyps; 19: Cardiovascular disease; 20: Calcium intake; 21: Percent energy from fat; 22: Fruit and vegetable intake; 23: Calcium supplement use; 24: Selenium supplement use; 25: Current healthcare provider; 26: Last medical visit within one year; 27: Colon screening; 28: Red meat consumption; 29: Use of aspirin; 30: Calendar year; 31: Study; 32: Pack-years of smoking before age 30; 33: Height; 34: Total energy intake; 35: Hypercholesterolemia; 36: Multivitamin use; 37: Energy-adjusted fibre intake; 38: Folate intake; 39: Benign mammary dysplasia; 40: Coronary heart disease; 41: Resective colorectal surgery; 42: Socioeconomic status; 43: Diabetes; 44: Weight; 45: Thyroid disease; 46: Hypertension; 47: Renal failure; 48: Chest pain; 49: Mental illness; 50: Lung disease; 51: Gastrointestinal Disease; 52: Prostate disease; 53: Total cholesterol; 54: Race; 55: Arthritis; 56: Use of gastroprotective drugs; 57: Estrogen use; 58: Tobacco abuse; 59: Mammography; 60: Gynecologic examination; 61: Pap smear; 62: Fecal occult blood; 63: Number of physician visits; 64: Prior nursing home stay; 65: History of heart attack; 66: Use of cardiovascular drugs. CRC: Colorectal cancer.
Table 3 Case-control studies included in the meta-analysis
StudyStudy locationAll subjectsCRC casesOR95%CIControl for potential confounding factors2
Deshpande[62], 2013United States73472367360.960.89-1.051-4
Broughton et al[64], 2012United Kingdom2331010.430.25-0.801, 2, 5-8
Lakha et al[63], 2012United Kingdom6033090.330.15-0.691-3, 9-15
Cheng et al[65], 2011Taiwan578011561.090.91-1.301-3, 5, 8, 9, 16-21
Vinogradova et al[26], 2011United Kingdom60373117491.071.00-1.151-3, 7, 9, 11, 12, 22-27
Robertson et al[66], 2010Denmark10976999790.870.80-0.961, 2, 5-7, 9, 10, 19
Hachem et al[68], 2009United States3040060800.910.86-0.961-4, 9, 16, 19, 20, 28, 29
Shadman et al[67], 2009United States20446691.170.74-1.851, 4, 11, 12, 30-33
Boudreau et al[70], 2008United States13306651.020.65-1.591, 2, 5, 9, 11, 12, 32
Yang et al[69], 2008United Kingdom4872444321.100.50-2.201, 2, 4, 6, 7, 9, 11, 12, 32, 34
Hoffmeister et al[71], 2007Germany11545400.690.45-1.061, 2, 4-7, 9, 11, 12, 30-32, 35-37
Coogan et al[72], 2007United States361818090.920.78-1.091, 2, 4, 9, 10
Li et al[73], 2006United States74133910.80.34-1.871, 2, 5, 6, 9, 11, 12, 30, 38
Poynter et al[25], 2005Israel396819530.530.38-0.741, 2, 9, 30, 36, 39-41
Rubin et al[74], 2005United States387240184400.920.89-0.961, 2, 4, 9, 17, 42-45
Kaye et al[76], 2004United Kingdom1808832911.00.6-1.71, 2, 11, 12, 42
Graaf et al[77], 2004Netherlands2010529210.870.48-1.571, 2, 5, 9, 17, 18, 32, 34, 46-49
Khurana et al[75], 2004United States534273533910.940.89-1.001, 2, 6, 9, 12, 38, 50
Blais et al[78], 2000Canada59625610.830.37-1.891, 2, 14, 18, 34, 46
1Figures for colon cancer rather than colorectal cancer;
21: Age; 2: Sex; 3: Inflammatory bowel disease; 4: Colonoscopy; 5: Diabetes; 6: Alcohol use; 7: Use of aspirin; 8: Use of metformin; 9: Use of NSAIDs; 10: Precinct of residence; 11: Body mass index; 12: Smoking status; 13: Physical activity level; 14: History of neoplasia; 15: Family history of cancer; 16: Fecal occult blood; 17: Prior hospitalizations; 18: Other lipid-lowering therapy; 19: Cholecystectomy; 20: Liver disease; 21: Colorectal polyps; 22: History of heart attack; 23: Hypertension; 24: Coronary heart disease; 25: Socioeconomic status; 26: Rheumatoid arthritis; 27: Use of COX-2 inhibitors; 28: Diabetic nephropathy; 29: Use of sulfonylurea; 30: Family history of colorectal cancer; 31: Education; 32: Hormone replacement therapy; 33: Calendar year; 34: Duration of follow-up; 35: Red meat consumption; 36: Hypercholesterolemia; 37: History of rheumatic disease; 38: Race; 39: Ethnic group; 40: Sports participation; 41: Level of vegetable consumption; 42: Number of physician visits; 43: Use of glucocorticosteroids; 44: Use of immunomodulators; 45: Use of 5-aminosalicylic acids; 46: Comorbidity score; 47: Use of diuretics; 48: Use of ACE inhibitors; 49: Use of CCBs; 50: Obesity. CRC: Colorectal cancer.
Table 4 Meta-analysis results
No. of studiesFixed-effects model
Random-effects model
Tests of homogeneity
Tests of publication bias
RR95%CIRR95%CIQ value (df)P valueI2Begg’s P valueEgger’s P value
Randomized Controlled Trials80.900.78-1.040.890.74-1.079.31 (7)0.2325%0.220.31
Placebo-controlled RCTs70.860.74-1.010.850.71-1.037.21 (6)0.3017%0.180.24
RCTs of lipophilic statins30.900.73-1.100.900.73-1.101.55 (2)0.460%0.600.70
RCTs of lipophobic statins50.900.73-1.120.830.60-1.157.75 (4)0.1048%0.620.05
Cohort Studies130.960.93-0.990.910.83-1.0070.85 (12)< 0.00183%0.540.22
Case-control studies190.930.91-0.960.920.87-0.9850.31 (18)< 0.00164%0.460.27
Published in full text form160.940.91-0.980.900.83-0.9949.12 (15)< 0.00169%0.470.19
Observational studies320.940.92-0.960.920.87-0.96122.68 (31)< 0.00175%0.360.16
All studies400.940.92-0.960.910.87-0.96132.3 (39)< 0.00171%0.330.11
RCT: Randomized controlled trials.