Table 8Colorectal cancer preventive interventions

StudySpecific intervention(s) evaluatedType of SREligible populationsEligible study designsNumber and type of trialsFollowp duration (years)Sample sizesAgesSummary effect sizesMain conclusion
Type of intervention: Chemoprevention (antioxidants, NSAIDS, and statins)
Bardia18 2008Antioxidant supplementation:
Vitamin C
Vitamin E
or any others
QuantPeople at average risk of cancerRCTs;
Supplements with fully disclosed components and not dietary increases in nutrients; placebo-controlled;
≥1 y of followup; reported overall cancer incidence
≤12NR178,086 (total across all studies)NRCRC incidence (any antioxidant)
RR=1.00 (0.90–1.10)
No evidence of protective effect of antioxidants against CRC
Alkhenizan19 2007Vitamin E alone
Vitamin E as part of other supplements
Quant≥18 y; range of pre-existing health states, but not high risk for colorectal cancer.RCTs;
Supplementation in capsule or tablet form;
control = placebo or no intervention;
reported total mortality, cancer mortality, cancer-specific mortality
2–4 depending on outcome≥510 days – ≤10 y (not reported separately for individual cancers)24,114 – 91,099 (across all trials, depending on analysis)NR

CRC incidence (Vit E any combination)

RR=0.95 (0.81–1.12)


CRC incidence (Vit E alone)

RR=1.05 (0.79–1.39)

No evidence of protective effect of Vitamin E against CRC
Dubé23 2007ASAQuant and Qual“Average’ riskRCTs, controlled clinical trials2
PHS 325mg every other day


CRC mortality (WHS) “No statistically significant benefit of ASA on CRC mortality”


CRC incidence

RR=1.02 (0.84–1.25)


Adenoma incidence (PHS only)

RR=0.86 (0.68–1.10)

Observational data appear to indicate that ASA use reduces the risk of colorectal neoplasia, but this effect is not seen in large trials of low dose ASA use
Browning20 2006StatinsQuantAll populations except highly specific statin-using patients (e.g. familial hypercholesterolemia, renal transplant)RCTs;
Control = placebo;
Reported all cancer, site-specific cancer incidence or mortality
1–9 (depending on analysis)Median followup:
Trials: 3.6 y
n = 103,573
RCTs: 18–82 yCRC incidence from use of statins:

All studies

RR=1.02 (0.89–1.16)


Mortality (1 trial)

RR=0.33 (0.07–1.63)


LI (4 trials)

RR=1.0 (0.85–1.18)


LO (5 trials)

RR=1.04 (0.85–1.30)


LP (2 trials)

RR=1.07 (0.61–1.85)


MP (5 trials)

RR=1.04 (0.84–1.30)


HP (2 trials)

RR=0.99 (0.83–1.18)

No evidence of protective effect
Type of intervention: Population-based screening
Hewitson24 2006FOBT (Hemoccult)
investigation following positive screen result - colonoscopy or sigmoidoscopy and double contrast barium enema, with removal of colorectal cancers or adenomas found at diagnostic investigation
Quant, QualAdults;
volunteers or individuals/households identified from primary care records or population registries
RCTs (individual or groups); control = no screening; report results based on participation in >1 screening round; report colorectal cancer mortality410–1746,551–150,251
45–80 y

CRC mortality (all groups)

RR=0.84 (0.78–0.90)


CRC mortality (biennial screening groups)

RR=0.85 (0.78–0.92)


All cause mortality (all groups)

RR=1.00 (0.99–1.01)


All cause mortality without CRC mortality (all groups)

RR=1.01 (1.00–1.03).


CRC mortality (attended ≥1 screening)

RR=0.75 (0.66–0.84)

15–16% reduction in relative risk of CRC mortality for individuals allocated to receive screening, rising to 25% risk reduction for those who actually participated at least once. No evidence of reduction in overall mortality, and borderline evidence of increased non-CRC mortality
Kerr25 2007Once-only flexible sigmoidoscopy in addition to FOBT test, followed by colonoscopy for investigation of positive screen result*QuantNot specified, but population-based impliedRCTs; control = no screening control for this study was FOBT alone12–510,97850–75

CRC mortality

RR=0.78 (0.36–1.73)


CRC incidence

RR=1.37 (0.88–2.15)

The trial was limited by short followup period and no repeat screening. Results do not support benefit of combined screening strategy of FS with FOBT over FOBT alone in asymptomatic populations Poor compliance with combined screening group

Abbreviations: ASA=acetylsalicylic acid; BC=breast cancer; BSE=breast self-examination; DRE=digital rectal exam; e.g.=example; FOBT=fecal occult blood test; HP=High potency; LI=Lipophilic; LO=Lipophobic; LP=Low potency; mg=milligrams; MP=Medium potency; NR=not reported; PHS=Physician's Health Study; PSA=prostate specific antigen; Qual=Qualitative; Quant=Quantitative; RCT=Randomized controlled trial; RR=Relative risk; SR=systematic review; TRUS=transrectal ultrasound; y=year; WHS=Women's Health Study


Kerr review also examines FOBT screening alone, but results overlap, and are consistent, with those of Hewitson & Heresbach, therefore not reported here.

From: 3, Results

Cover of Clinical Utility of Cancer Family History Collection in Primary Care
Clinical Utility of Cancer Family History Collection in Primary Care.
Evidence Reports/Technology Assessments, No. 179.
Wilson B, Qureshi N, Little J, et al.

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.