Table 1.
Features and summary of the findings of the studies included in the review evaluating the impact of broccoli consumption on various types of cancer.
| Author; Year | Design | Sample (n) | Population/Country/Age | Exposure | Comparison | Outcomes | Effect Size | Follow-Up | Adjustments |
|---|---|---|---|---|---|---|---|---|---|
| Cancer mortality or general cancer incidence | |||||||||
| Colditz et al., 1985 [43] | Cohort study | 1271 | Men and women/USA 66 years | Broccoli | High intake vs. low intake | Cancer mortality (any type); n = 42 cases |
RR: 0.8; 95% CI from 0.4 to 1.6 | 5 years | Age |
| Wang L et al., 2009 [55] | Cohort study | 38,408 | Women/USA ≥ 45 years |
Broccoli | High intake vs. low intake | Cancer incidence (any type); n = 3234 cases |
RR: 1.05; 95% CI from 0.88 to 1.25 | 11.5 years | Multivariate |
| Breast cancer | |||||||||
| Adebamowo CA et al., 2005 [56] | Cohort study | 90,630 | Women/USA 25–46 years |
Broccoli | High intake vs. low intake | Breast cancer | RR (adjusted by age): 1.11; 95% CI: 0.67 to 1.85; RR (multivariable adjustment): 0.99; 95% CI from 0.59 to 1.65 |
5 years | Multivariate |
| Lin T et al., 2017 [68] | Cases and controls | 1491 cases and 1482 controls | Women/USA 21–97 years |
Broccoli | High intake vs. low intake | Breast cancer | OR: 0.68; 95% CI from 0.56 to 0.82 Raw → OR: 0.78; 95% CI from 0.66 to 0.91 Cooked → OR: 0.83; 95% CI from 0.70 to 0.99 |
1982–1998 | Multivariate |
| Ambrosone CB et al., 2004 [69] | Cases and controls | 740 cases and 810 controls | Caucasian women/USA < 50 years, >50 years |
Broccoli | High intake vs. low intake | Breast cancer | Premenopausal → OR: 0.6; 95% CI from 0.4 to 1.0 Postmenopausal → OR: 1.0; 95% CI from 0.7 to 1.4 |
1986–1991 | Multivariate |
| Lung and respiratory tract cancer | |||||||||
| Fontham ET et al., 1988 [39] | Cases and controls | 1253 cases; 1274 controls | Men and women/USA | Broccoli | High intake vs. low intake | Lung cancer | OR: 0.64; 95% CI from 0.54 to 0.78 | 1979–1982 | Multivariate |
| Steinmetz KA et al., 1993 [44] | Cases and controls | 138 cases and 2814 controls (random); base cohort (n = 41.837 women) | Women/USA 55–69 years |
Broccoli | High intake vs. low intake | Lung cancer; n = 179 cases | OR: 0.72; 95% CI from 0.40 to 1.29 | 4 years | Multivariate |
| Tarrazo-Antelo AM et al., 2014 [70] | Cases and controls | 371 cases and 496 controls | Men and women/Spain Median > 63 years |
Broccoli | High intake vs. low intake | Lung cancer | OR: 0.54; 95% CI from 0.35 to 0.84 | 2004–2008 | Multivariate |
| García-Lavandeira JA et al. (2022) [72] | Cases and controls | 438 cases and 781 controls | Men and women. Never smokers patients/Spain > 66 years |
Broccoli | High intake vs. low intake | Lung cancer; adenocarcinoma, n = 289. | OR: 0.55 (0.35–0.83) | 2002–2019 | Multivariate |
| Mettlin C. et al., 1989 [85] | Cases and controls | 569 cases (355 men/214 women) and 569 controls | Men and women/USA | Broccoli | High intake vs. low intake | Lung cancer | RR: 0.31; 95% CI 0.16 to 0.57 | 1989 | Multivariate/multiple regression |
| Goodman MT et al., 1992 [86] | Cases and controls | 675 cases (463 men and 212 women) and 675 controls | Men and women/USA | Broccoli | High intake vs. low intake | Lung cancer | Women → RR: 2.2; p < 0.01) Men → RR: 1.0; p = 0.37). Survival of small cell lung cancer in men RR: 2.6; p = 0.02 |
1979–1985 | Multivariate |
| Graham S et al., 1981 [88] | Cases and controls | 374 with laryngeal cancer and 381 controls | Men and women/USA | Broccoli | High intake vs. low intake | Lung cancer | NS | 1981 | - |
| Digestive tract cancer | |||||||||
| Gastric | |||||||||
| Morrison MEW et al., 2020 [40] | Cases and controls | 292 cases and 1168 controls | Men and women/USA | Broccoli | High intake vs. low intake | Gastric cancer | OR: 0.61; 95% CI: 0.43 to 0.86 | 1992–1998 | Multivariate |
| Correa P et al., 1985 [45] | Cases and controls | 391 cases and 391 controls | Men and women/USA | Broccoli | High intake vs. low intake | Gastric cancer | OR: 1.0; 95% CI from 0.7 to 1.7 | 1985 | Multivariate |
| Hansson LE et al., 1993 [73] |
Cases and controls | 338 cases and 669 controls | Men and women/Sweden | Broccoli | High intake vs. low intake | Gastric cancer | OR: 0.63; 95% CI from 0.41 to 0.96 | Adolescence and 20 years prior to the study | Multivariate |
| Hara M et al., 2003 [74] | Cases and controls | 149 cases and 287 controls | Men and women/Japan 20–70 years |
Broccoli | High intake vs. low intake | Gastric cancer | OR: 0.60; 95% CI: 0.34 to 1.08 | 1998–2002 | Multivariate |
| Graham S et al., 1972 [89] | Cases and controls | 228 cases and 228 controls | Men and women/USA | Broccoli | High intake vs. low intake | Gastric cancer | Inverse association, NS | 2004–2008 | |
| Colorectal | |||||||||
| Steinmetz KA et al., 1994 [63] | Cohort study | 41,837 | Women/USA 55–69 years |
Broccoli | High intake vs. low intake | Colon cancer; n = 212 cases | RR: 1.0; 95% CI from 0.7 to 1.7 | 5 years | - |
| Flood A et al., 2002 [64] | Cohort study | 45,490 | Women/USA Mean > 60 years |
Broccoli | High intake vs. low intake | Colon cancer; n = 485 cases | RR: 0.78; 95% CI from 0.58 to 1.06 | 7 years | Multivariate/multiple regression |
| Nomura AM et al., 2008 [65] | Cohort study | 85,903 men and 105,108 women | Men and women/USA 45–75 years |
Broccoli | High intake vs. low intake | Colorectal cancer | Women → RR: 0.92; 95% CI from 0.75 to 1.15, p = 652 Men → RR: 0.94; 95% CI from 0.76 to 1.15, p = 652 |
Average follow-up of 7.3 years | Multivariate/multiple regression |
| Steinmetz and Potter JD et al., 1993 [71] | Cases and controls | 220 cases and 438 controls | Men and women/Australia | Broccoli | High intake vs. low intake | Colon cancer | OR: 0.91; 95% CI from 0.48 to 1.72 | 1979–1980 | Multivariate |
| Hara M et al., 2003 [74] | Cases and controls | 115 cases and 230 controls | Men and women/Japan 20–70 years |
Broccoli | High intake vs. low intake | Colorectal cancer | OR: 0.18; 95% CI from 0.06 to 0.58 | 1998–2002 | Multivariate |
| Witte JS et al., 1996 [75] | Cases and controls | 488 cases and 488 controls | Men and women/USA 50–74 years | Broccoli | High intake vs. low intake | Adenomatous polyps | OR: 0.64; 95% CI from 0.44 to 0.92 | 1991–1993 | Multivariate |
| Lin HJ et al., 1998 [76] | Cases and controls | 459 cases and 507 controls | Men and women/USA 50–74 years |
Broccoli | High intake vs. low intake | Colorectal adenomas | OR: 0.47; 95% CI of 0.30–0.73; | 1991–1993 | Multivariate |
| Evans RC et al., 2002 [77] | Cases and controls | 512 cases and 512 controls | Men and women/UK | Broccoli | High intake vs. low intake | Colorectal cancer | Left side colon and rectal cancer (OR: 0.61; 95% CI 0.39 to 0.96); colorectal cancer in general (OR: 0.67; 95% CI 0.45 to 1.00); right colon cancer (OR: 1.00; 95% CI 0.39 to 2.57) | 6 years | Univariate |
| Mahfouz EM et al., 2014 [78] | Cases and controls | 150 cases and 300 controls | Men and women/Egypt | Broccoli | High intake vs. low intake | Colorectal cancer | OR: 0.11; 95% CI from 0.01 to 0.48: p = 0.03 | 2010–2011 | |
| Le Marchand et al., 1997 [79] | Cases and controls | Men (698 case–control pairs) Women (494 case–control pairs) |
Men and women (different ethnic groups)/USA < 84 years |
Broccoli | High intake vs. low intake | Colorectal cancer | Men → OR: 0.7; 95% CI from 0.4 to 1.0; p = 0.05 Women → OR: 0.7; 95% CI from 0.4 to 1.1; p = 0.18 |
1987–1991 | |
| Graham S et al., 1978 [87] | Cases and controls | 256 colon cancer cases and 783 controls; 330 rectal cancer cases and 628 controls | Men/USA | Broccoli | High intake vs. low intake | Colon and rectal cancer | Inverse association between the consumption of broccoli and the risk of colon cancer, but not rectal cancer, NS | 1978 | - |
| Miller et al., 1983 [90] | Cases and controls | 194 rectal cancer cases and 542 controls (2nd control series, 535) | Men and women 1st control series without pathologies 2nd series of surgical patients/Canada |
Broccoli | High intake vs. low intake | Colon and rectal cancer | Colon cancer OR (men): 1.0; p-value: 0.48 OR (women): 1.0; p-value 0.43 n = 348 cases Rectal cancer OR (men): 1.0; p-value: 0.34 OR (women): 1.2; p-value: 0.29 n = 194 cases. |
1983 | - |
| Freudenheim JL et al., 1990 [91] | Cases and controls | 422 cases (277 men and 145 women) and 422 controls | Men and women/USA | Broccoli | High intake vs. low intake | Rectal cancer | Inversely associated with the risk of rectal cancer in men, but not in women, NS | 1978–1986. | - |
| Slattery ML et al., 2000 [92] | Cases and controls | 1579 cases and 1898 controls | Men and women/USA 30–79 years |
Broccoli | High intake vs. low intake | Colon cancer | GSTM-1 genotype. OR:1.23; 95% CI from 0.86 to 1.76 for the GSTM1-null genotype OR:0.92; 95% CI from 0.63 to 1.33 for the GSTM1-present genotype OR: 0.30; 95% CI from 0.13 to 0.70; only for the GSTM1-null genotype and age less than 55 years |
1991–1994 | Multivariate in GSTM1-null genotype |
| Lin HJ et al., 2002 [93] | Cases and controls | 459 cases and 507 controls | Men and women 50–74 years |
Broccoli | High intake vs. low intake | Colorectal adenomas | OR 0.41; 95% CI: 0.24 to 0.70 for the GSTM1-null and GSTT1-null genotypes | 1991–1993 | - |
| Pancreas | |||||||||
| Azeem K et al., 2016 [41] | Cases and controls | 310 cases and 220 controls | Men and women/Czech Republic | Broccoli | High intake vs. low intake | Pancreatic cancer | OR: 0.37; 95% CI from 0.25 to 0.53 | 2006–2009 | - |
| Liver | |||||||||
| Zhao L et al., 2023 [67] | Cohort study | 485,403 | Men and women/USA 50–71 years |
Broccoli | High intake vs. low intake | Liver cancer | HR: 0.66; 95% CI from 0.54 to 0.81; p trend < 0.001. | 1995–1996 | Multivariate |
| Urinary tract cancer | |||||||||
| Prostate | |||||||||
| Kirsh VA et al., 2007 [37] | Cohort study | 29,361 | Men/USA Mean > 62 years |
Broccoli | High intake vs. low intake | Prostate cancer | All prostate cancer → RR: 0.91; 95% CI from 0.77 to 1.06 Aggressive prostate cancer → RR: 0.76; 95% CI from 0.59 to 0.99 Extraprostatic cancer → RR: 0.55; 95% CI from 0.34 to 0.89 |
4.2 years | Multivariate |
| Ambrosini GL et al., 2008 [57] | Cohort study | 1985 | Men in a prevention program supplemented with beta-carotene and retinol/Australia. Median 62.6 years |
Broccoli | High intake vs. low intake | Prostate cancer. n = 97 |
RR: 0.56; 95% CI from 0.31 to 1.0 | 1990–2004 | - |
| Giovannucci E et al., 2003 [66] | Cohort study | 47,365 | Men/USA < 65 years and ≥65 years |
Broccoli | High intake vs. low intake | Total prostate cancer (excluding stage T1a tumors); n = 962 | RR: 0.87; 95% CI from 0.73 to 1.05 | 1986–2000 | Multivariate |
| Joseph MA et al., 2004 [80] | Cases and controls | 428 cases and 537 controls | Caucasian Men/USA 45–85 years |
Broccoli | High intake vs. low intake | Incident prostate cancer | OR: 0.72; 95% CI from 0.49 to 1.06 | 1986–1991 | Multivariate |
| Bladder | |||||||||
| Michaud DS et al., 1999 [38] | Cohort study | 47,909 | Men/USA 40–75 years |
Broccoli | High intake vs. low intake | Bladder cancer n = 252 cases. |
RR: 0.61; 95% CI from 0.42 to 0.87 | 10 years | Multivariate |
| Tang L et al., 2010 [58] | Cohort study | 239 | Men/USA < 60 years, 60–70 years, and >70 years |
Broccoli | High intake vs. low intake | Survival of patients with bladder cancer Cancer deaths n = 101 cases |
General death (HR: 0.57; 95% CI 0.39 to 0.83); Disease-specific death (HR: 0.43; 95% CI 0.25 to 0.74) | 8 years | Multivariate |
| Castelao JE et al., 2004 [81] | Cases and controls | 1592 cases and controls | Men and women (non-Asians)/USA 25–64 years |
Broccoli | High intake vs. low intake | Bladder cancer | OR: 0.81; 95% CI from 0.59 to 1.09 | 1987–1996 | Multivariate |
| Lin J et al., 2009 [82] | Cases and controls | 884 cases and 878 controls | Men and women/USA mean age 64 years cases, 65 years controls |
Broccoli | High intake vs. low intake | Bladder cancer, patients who had not received previous chemotherapy or radiotherapy | OR: 0.71; 95% CI from 0.53 to 0.96 | 1999-Currently ongoing | |
| Tang L et al., 2008 [83] | Cases and controls | 275 cases and 825 controls | Men and women (Predominantly Caucasian)/USA 25–86 years cases; 21–92 years controls |
Broccoli | High intake vs. low intake | Bladder cancer | Broccoli raw → OR: 0.57; 95% CI from 0.40 to 0.81 Broccoli cooked → OR: 0.88; 95% CI from 0.65 to 1.20 |
1982–1998 | Multivariate |
| Reproductive system cancer | |||||||||
| Shen Y et al., 2016 [42] | Cases and controls | 600 cases and 236 controls | Women/China 30–50 years |
Broccoli | High intake vs. low intake | Uterine fibroids | OR: 0.55; 95% CI from 0.32 to 0.96 | 2010–2014 | Multivariate |
| Gates MA et al., 2007 [59] | Cohort study | 66,940 | Women/USA Mean 50–51 years |
Broccoli | High intake vs. low intake | Ovarian cancer | RR: 0.67; 95% CI from 0.45 to 1.01 | 1984–2002 | Multivariate |
| Chang E et al., 2007 [62] | Cohort study | 97,275 | Women/USA Median age at baseline 50 years |
Broccoli | High intake vs. low intake | Ovarian cancer | RR: 0.91; 95% CI from 0.61 to 1.36 | 1995–2003 | Multivariate |
| Barbone F et al., 1993 [84] | Cases and controls | 103 cases and 236 controls | Women/USA | Broccoli | High intake vs. low intake | Endometrial cancer | OR: 0.5; 95% CI from 0.3 to 1.0 | 1985–1988 | Multivariate |
| Thyroid cancer | |||||||||
| Braganza MZ et al., 2015 [60] | Cohort study | 292,477 | Men and women/USA Mean: 63.4 years |
Broccoli | High intake vs. low intake | Thyroid cancer | HR: 2.13; 95% CI from 1.13 to 3.99; p trend < 0.01. | 1996–2006 | Multivariate |
| Ron E et al., 1987 [94] | Cases and controls | 159 cases and 285 controls | Men and women/USA | Broccoli | High intake vs. low intake | Thyroid cancer | OR: 0.8; p trend: 0.20 | 1987 | - |
| Lymphoid cancer | |||||||||
| Thompson CA et al., 2010 [61] | Cohort study | 35,159 | Women/USA 55–69 years |
Broccoli | High intake vs. low intake | Non-Hodgkin lymphoma (NHL), diffuse large B-cell lymphoma (DLBCL) and follicular lymphoma (FL) n = 415 NHL; 184 DLBCL and 90 FL cases |
NHL (RR: 0.72; p-value: 0.018). mainly for FL and weaker or not apparent for DLBCL. | 1986–2005 | Multivariate |
Abbreviations: AMI = acute myocardial infarction; CC = coronary heart disease; CI = confidence interval; CVD = cardiovascular diseases; DLBCL = diffuse large B-cell lymphoma; FL = follicular lymphoma; GSTM1 = glutathione S-transferase Mu 1 gene; HR = hazard ratio; n = number of cases; NHL = non-Hodgkin lymphoma; NS = quantitative data not supplied; OR = odds ratio; RCT: randomized controlled trial; RR = relative risk or risk ratio.