Broccoli Consumption and Risk of Cancer: An Updated Systematic Review and Meta-Analysis of Observational Studies

Background: The scientific literature has reported an inverse association between broccoli consumption and the risk of suffering from several types of cancer; however, the results were not entirely consistent across studies. A systematic review and meta-analysis of observational studies were conducted to determine the association between broccoli consumption and cancer risk with the aim of clarifying the beneficial biological effects of broccoli consumption on cancer. Methods: PubMed/MEDLINE, Web of Science, Scopus, Cochrane Library (CENTRAL), and Epistemonikos databases were searched to identify all published papers that evaluate the impact of broccoli consumption on the risk of cancer. Citation chasing of included studies was conducted as a complementary search strategy. The risk of bias in individual studies was assessed using the Newcastle-Ottawa Scale. A random-effects model meta-analysis was employed to quantitatively synthesize results, with the I2 index used to assess heterogeneity. Results: Twenty-three case–control studies (n = 12,929 cases and 18,363 controls; n = 31,292 individuals) and 12 cohort studies (n = 699,482 individuals) were included in the meta-analysis. The results suggest an inverse association between broccoli consumption and the risk of cancer both in case–control studies (OR: 0.64, 95% CI from 0.58 to 0.70, p < 0.001; Q = 35.97, p = 0.072, I2 = 30.49%—moderate heterogeneity; τ2 = 0.016) and cohort studies (RR: 0.89, 95% CI from 0.82 to 0.96, p = 0.003; Q = 13.51, p = 0.333, I2 = 11.21%—low heterogeneity; τ2 = 0.002). Subgroup analysis suggested a potential benefit of broccoli consumption in site-specific cancers only in case–control studies. Conclusions: In summary, the findings indicate that individuals suffering from some type of cancer consumed less broccoli, suggesting a protective biological effect of broccoli on cancer. More studies, especially cohort studies, are necessary to clarify the possible beneficial effect of broccoli on several types of cancer.


Introduction
Cancer has become the second leading cause of death worldwide, only surpassed by cardiovascular diseases.However, mortality from cancer is actually higher than from cardiovascular diseases in more-developed countries [1].In 2020, approximately 10 million patients died from cancer and 19.3 million people were diagnosed with cancer for the first time worldwide, as prevalence continues to rise [2,3].
Nutrients 2024, 16, 1583 2 of 25 Since the 1990s, cancer incidence and mortality have tended to decline, while the five-year relative survival rate has increased between 2014 and 2018.Advances in treatment have led to an accelerated decrease in mortality rates in lung, prostate and colorectal cancers in men and in breast and colorectal cancers in women [4].
Researchers have long focused their efforts on identifying multiple risk factors that contribute to the possibility of developing cancer.Tobacco smoking, air pollution, asbestos, alcohol consumption, ultraviolet radiation, Helicobacter pylori infection, lifestyle, excess body weight and poor diet are considered exogenous cancer risk factors associated with a higher incidence of certain types of cancer [5].In theory, most of these exogenous cancer risk factors are potentially modifiable, which can contribute to preventing and reducing the incidence and mortality of various types of cancer [6].
Broccoli (Brassica oleracea var.Italica) is an herbaceous plant of the family Brassicaceae, commonly called cruciferous vegetables (Cruciferae), characterized by low energy content and high nutritional value due to its fiber, potassium, folate and vitamins C and K contents [21].Sulforaphane is a compound present in broccoli derived from the hydrolysis of glucoraphanin by the action of myrosinase.This compound is a glucosinolate that is a potent inducer of nuclear factor erythroid 2-related factor 2 (Nrf2), a transcription factor that positively regulates genes associated with the production of antioxidant proteins key to neutralizing oxidative damage.Nrf2 has recently been associated with the modulation of central metabolic pathways [22,23].Sulforaphane has shown a variety of biological properties that contribute positively to human health.It has been revealed as a chemopreventive and protective agent in various types of cancer, such as colon, gastric, bladder, prostate, breast, skin and lung cancers [24].Sulforaphane, isolated from broccoli aqueous extract, has shown an inhibitory effect on the damage induced by ultraviolet radiation and the progression of skin cancer [25], as well as decreasing the harmful effects of prostate cancer [26].In brief, sprouts were boiled in deionized water for 30 min at over 95 • C to extract glucoraphanin.The resulting aqueous extract was cooled to 37 • C, and myrosinase was added, catalyzing the conversion of glucoraphanin to sulforaphane over a 4-h period [26].The consumption of fresh broccoli is associated with the presence of sulforaphane in plasma and urine [27,28], which is maintained beyond 24 h after consumption.The presence of sulforaphane is greater when it is consumed as a part of vegetables than when taken as an extract.This is possibly due to the presence of myrosinase and other modulating compounds present in fresh broccoli [29][30][31][32], a fact also demonstrated in new varieties of broccoli [33].The bioavailability of sulforaphane is higher in raw than in cooked broccoli [34].It also seems that its absorption and bioavailability improve when body mass index is higher than 26 kg/m 2 [35,36].
These data suggest biological plausibility that would explain and reinforce the possible benefits of broccoli consumption in cancer prevention.Some studies reported positive associations between broccoli consumption and risk in several types of cancer [37][38][39][40][41][42]; however, not all studies were consistent [43][44][45].To our knowledge, there is no updated systematic review on this topic that includes all types of relevant studies and evaluates broccoli consumption associated with cancer risk.Likewise, not all reviews showed subgroup analyses by study design, which seems relevant for a correct interpretation of the results.Therefore, we conducted a meta-analysis evaluating the association between broccoli consumption and cancer risk with the aim of clarifying the beneficial biological effects of broccoli consumption on cancer.

Materials and Methods
A systematic review and meta-analysis of observational studies was conducted following the guidelines of Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA 2020) statement [46].

Information Sources and Search Strategies
Systematic searches were performed in electronic databases: MEDLINE via PubMed, Web of Science, Scopus, Cochrane Library (CENTRAL) and Epistemonikos.The initial search strategy was designed for PubMed and adapted to the syntax of the other databases using Polyglot software (available at: https://sr-accelerator.com/#/polyglot; accessed on 14 November 2023) from Systematic Review Accelerator [47].
Search strategies adapted for each database's syntax are provided in the Supplementary File S1.These strategies were tailored to optimize search efficiency and ensure thorough coverage of relevant literature.
To identify unpublished and ongoing studies, study protocols and scientific conference proceedings and abstracts were also retrieved.Additionally, we carried out forward and backward citation chasing from each included article using the Citation Chaser software (available at: https://estech.shinyapps.io/citationchaser/;accessed on 20 November 2023) [48].The last search in databases was performed on 22 December 2023.

Eligibility Criteria
The research team used the population, exposure factor, comparator, outcomes, types of study designs (PECOT) approach to specify the eligibility criteria as follows: Population: healthy or sick individuals of any age, sex, ethnicity or country; Exposure: broccoli as a food in any culinary preparation, including raw, cooked and even as beverages; dehydrated broccoli was also included, ensuring that the active components remained intact; studies with broccoli extracts or any active ingredients extracted from broccoli were excluded; Comparison: high consumption of broccoli vs. low or no consumption of broccoli; Outcomes: cancer outcomes were included; and Types of Studies: randomized controlled trials, cohort studies, case-control studies.Systematic reviews, meta-analyses and umbrella reviews were also retrieved as a source of primary studies not retrieved from the database search.

Study Selection Process and Data Extraction
The studies retrieved were managed using ZOTERO software (version 6.0, Corporation for Digital Scholarship, Vienna, VA, USA), and duplicates were manually removed.The resulting library was exported in RIS format and imported to Abstrackr software (available at: http://abstrackr.cebm.brown.edu/account/login;accessed on 12 December 2023) [49].
Blind peer review title and abstract screening was performed by three researchers, EB, MMo and MVG-C, applying the eligibility criteria to identify articles as "included", "excluded" or "not sure".Articles tagged as "included" and "not sure" were independently selected for full-text screening by EB, MMo and MVG-C.Discrepancies during the screening were resolved by consensus between EB, MMo and MVG-C.The identification of full-text studies and data extraction were performed at the same time by EB, MMo and MVG-C.The following data were extracted from each study using a piloted form: first author and year, article's title, objective, study design, sample size, details about the exposure factor, results (association measure and 95% confidence interval) and conclusions.
The search and selection process results were reported using a flow diagram according to the PRISMA 2020 statement [46].

Risk of Bias Assessment
All articles selected for inclusion in this review were case-control and cohort studies.To assess the risk of bias in case-control and cohort studies, EB, MMo and MVG-C made a blinded assessment of each study applying the Newcastle-Ottawa Scale [50].The scale assesses the study's methodological quality and covers key aspects of the risk of bias in case-control and cohort studies, and is composed of 4 items that evaluate the selection process, 1 item to assess whether groups are homogeneous (control of confounding factors in the design and/or analysis), and 3 items to assess exposure factors (cases and controls) or outcomes (cohorts) (available at: https://www.ohri.ca/programs/clinical_epidemiology/oxford.asp;accessed on 10 January 2024).

Summary of Outcomes
The characteristics of included studies and main findings were presented in a table of findings, stratifying and organizing the studies based both on cancer outcome and study design, and in line with the methods proposed by the Centre for Reviews and Dissemination [51].

Statistical Analysis
A meta-analysis with Open Metaanalyst software (version 5.26.14; available at: http://www.cebm.brown.edu/openmeta/;accessed on 12 February 2024) was performed using a random effects model (DerSimonian-Laird method), which considers heterogeneity within and between studies, to calculate summary relative risks (RR, cohort studies) and odds ratio (OR, case-control studies) with a confidence interval (CI) of 95% and 3 digits of precision.Forest plots of all possible comparisons were performed.The Cochran's Q statistic, I 2 index and tau-squared (τ 2 ) were used to evaluate heterogeneity [52].For the Q statistic, a p-value < 0.1 was considered to be representative of statistically significant heterogeneity.For I 2 index, heterogeneity was classified as follows: 25%-low; 50%-moderate; 75%-high levels of variance.A τ 2 equal to zero indicates no heterogeneity between studies, and a τ 2 close to zero indicates lower levels of heterogeneity.
Publication bias was evaluated by Egger's and Begg's tests [53,54].A p-value < 0.05 for Egger's or Begg's tests was considered representative of significant statistical publication bias.

Article Selection Process
We identified 3026 articles from databases (772 articles from PubMed/Medline, 893 articles from Web of Science, 1082 articles from Scopus, and 279 from Epistemonikos).After removing duplicates and adding five additional articles identified in published systematic reviews, the complete search strategy resulted in 1369 unique titles and abstracts to be screened.
During the title and abstract screening, researchers reached agreement on including 34 articles, agreed to exclude 1186 articles, and had doubts on 149 papers, mainly due to a lack of data to assess if they met all eligibility criteria.Researchers retrieved 183 full-text to be screened.Finally, 49 articles were selected after complete full-text reading, and 134 were excluded due to not meeting the population, exposure factor or study design criteria.Of the 49 included studies, 14 studies could not be meta-analyzed due to lack of data and data heterogeneity.Finally, 35 studies were meta-analyzed.Figure 1 displays the flow diagram of the search and screening process.
The snowball search using the 49 included studies yielded 172 potential non-screened additional records.From this, 159 articles were excluded after the second-round title and abstract screening, and 13 were selected for complete full-text reading.Finally, all articles were discarded because they did not meet the eligibility criteria or were already included in the review.The snowball search using the 49 included studies yielded 172 potential nonscreened additional records.From this, 159 articles were excluded after the second-round title and abstract screening, and 13 were selected for complete full-text reading.Finally, all articles were discarded because they did not meet the eligibility criteria or were already included in the review.
High broccoli intake was compared with low consumption.High broccoli intake ranged from daily to weekly consumption, with a minimum frequency of once per week and a maximum of once per day.Low broccoli intake was established from not occurring weekly to three times per month or nonconsumption.
High broccoli intake was compared with low consumption.High broccoli intake ranged from daily to weekly consumption, with a minimum frequency of once per week and a maximum of once per day.Low broccoli intake was established from not occurring weekly to three times per month or nonconsumption.

Outcomes of Cohort Studies
Of the 16 included cohort studies that evaluated the association between broccoli intake and the risk of cancer, only 12 could be meta-analyzed [37,38,43,[55][56][57]59,[62][63][64][65][66]. Figure 3 presents the detailed data of the meta-analysis of cohort studies that evaluated the association between broccoli consumption and several types of cancer.
lower risk of several types of cancer (RR: 0.89, 95% CI from 0.82 to 0.96, p = 0.003; Q = 13.51,p = 0.333, I 2 = 11.21%-lowheterogeneity; τ 2 = 0.002).In subgroup analysis, prostate cancer showed a statistically significant inverse association between broccoli intake and cancer risk (p = 0.042).However, this association was not statistically significant in colorectal cancer (p = 0.136), in reproductive system cancer (p = 0.113), and in cancer in general (p = 0.713).All cancer subgroups showed low statistical heterogeneity (Figure 3).In the case-control studies, no evidence of significant publication bias was verified with the Begg's funnel plot (Figure 4A, p = 0.06) or with the Egger's test (p = 0.07).In the cohort studies, no evidence of significant publication bias was observed with the Begg's funnel plot (Figure 4B, p = 0.36) or with the Egger's test (p = 0.08).The analysis showed that a high consumption of broccoli could be associated with a lower risk of several types of cancer (RR: 0.89, 95% CI from 0.82 to 0.96, p = 0.003; Q = 13.51,p = 0.333, I 2 = 11.21%-lowheterogeneity; τ 2 = 0.002).In subgroup analysis, prostate cancer showed a statistically significant inverse association between broccoli intake and cancer risk (p = 0.042).However, this association was not statistically significant in colorectal cancer (p = 0.136), in reproductive system cancer (p = 0.113), and in cancer in general (p = 0.713).All cancer subgroups showed low statistical heterogeneity (Figure 3).
In the case-control studies, no evidence of significant publication bias was verified with the Begg's funnel plot (Figure 4A, p = 0.06) or with the Egger's test (p = 0.07).In the cohort studies, no evidence of significant publication bias was observed with the Begg's funnel plot (Figure 4B, p = 0.36) or with the Egger's test (p = 0.08).

Discussion
To our knowledge, this is the first meta-analysis that evaluates the associati tween broccoli intake and several types of cancer.According to the results of the analysis, findings from cohort and case-control studies suggested a greater reduc cancer risk in people who consumed more broccoli compared to those who consum or no broccoli; however, there is uncertainty about the robustness of the current av evidence.While more cohort studies are needed to draw more precise conclusio results of the case-control studies showed borderline statistical significance with m ate heterogeneity.
Various meta-analyses have verified that the general consumption of cruciferou etables is inversely associated with the risk of various types of cancer, such as col cancer, gastric cancer, bladder cancer, pancreatic cancer, breast cancer, and ovarian [7,11,12,14,17,19,20]; however, not all meta-analyses have confirmed this associatio Specific analysis for broccoli yielded similar findings, and these results are consisten the review led by Verhoeven et al., 1996 [96], which considered only case-control s and suggested that in most of them (56%), the high consumption of broccoli was ated with a lower probability of developing cancer.In our meta-analysis, 17 of 23 control studies (73.9%) showed a protective effect of higher broccoli consumption o ious types of cancer.
For specific types of cancer, case-control studies suggested that high brocco sumption was inversely associated with lung or respiratory tract cancer, reproductiv cer, pancreatic cancer, gastric cancer and bladder cancer.Although case-control s analyzed in colorectal and breast cancer also demonstrated this inverse association findings should be interpreted with caution due to the subtle and moderate heterog confirmed in the meta-analyzed studies.
The inverse association observed in case-control studies between greater con tion of broccoli and the risk of suffering from some type of cancer was corroborated cohort studies, also observing low heterogeneity in the studies.In this regard, the heterogeneity detected in the cohort studies can be attributed, at least in part, to the g number of case-control studies compared to the cohort studies.
For specific types of cancer, meta-analysis of cohort studies raised doubts abo likely beneficial effect of broccoli consumption in reducing the risk of colorectal and cancer of the reproductive system [59,[62][63][64][65].In alignment with these findin did not find an inverse association between broccoli consumption and the risk of su from cancer in general (two cohort studies) and breast cancer (one cohort study) (N Health Study II; 2005) [43,55,56].Conversely, a protective benefit of broccoli consum was observed for bladder cancer [38] and prostate cancer [37,57,66].These results s

Discussion
To our knowledge, this is the first meta-analysis that evaluates the association between broccoli intake and several types of cancer.According to the results of the meta-analysis, findings from cohort and case-control studies suggested a greater reduction in cancer risk in people who consumed more broccoli compared to those who consumed less or no broccoli; however, there is uncertainty about the robustness of the current available evidence.While more cohort studies are needed to draw more precise conclusions, the results of the casecontrol studies showed borderline statistical significance with moderate heterogeneity.
Various meta-analyses have verified that the general consumption of cruciferous vegetables is inversely associated with the risk of various types of cancer, such as colorectal cancer, gastric cancer, bladder cancer, pancreatic cancer, breast cancer, and ovarian cancer [7,11,12,14,17,19,20]; however, not all meta-analyses have confirmed this association [95].Specific analysis for broccoli yielded similar findings, and these results are consistent with the review led by Verhoeven et al., 1996 [96], which considered only casecontrol studies and suggested that in most of them (56%), the high consumption of broccoli was associated with a lower probability of developing cancer.In our meta-analysis, 17 of 23 case-control studies (73.9%) showed a protective effect of higher broccoli consumption on various types of cancer.
For specific types of cancer, case-control studies suggested that high broccoli consumption was inversely associated with lung or respiratory tract cancer, reproductive cancer, pancreatic cancer, gastric cancer and bladder cancer.Although case-control studies analyzed in colorectal and breast cancer also demonstrated this inverse association, these findings should be interpreted with caution due to the subtle and moderate heterogeneity confirmed in the meta-analyzed studies.
The inverse association observed in case-control studies between greater consumption of broccoli and the risk of suffering from some type of cancer was corroborated in the cohort studies, also observing low heterogeneity in the studies.In this regard, the lower heterogeneity detected in the cohort studies can be attributed, at least in part, to the greater number of case-control studies compared to the cohort studies.
For specific types of cancer, meta-analysis of cohort studies raised doubts about the likely beneficial effect of broccoli consumption in reducing the risk of colorectal cancer and cancer of the reproductive system [59,[62][63][64][65].In alignment with these findings, we did not find an inverse association between broccoli consumption and the risk of suffering from cancer in general (two cohort studies) and breast cancer (one cohort study) (Nurses' Health Study II; 2005) [43,55,56].Conversely, a protective benefit of broccoli consumption was observed for bladder cancer [38] and prostate cancer [37,57,66].These results suggest that broccoli intake could be associated with certain specific cancers.More cohort studies would be necessary to support such claims and further improve the perception of the possible healthy effect of broccoli consumption on specific type of cancer.
Other reviews have previously evaluated the association between broccoli consumption and colorectal cancer with similar findings.In the meta-analysis conducted by Wu et al., 2013, six studies on broccoli were included: three cohort and three case-control studies.The set of studies showed a lower risk of colorectal cancer associated with the consumption of broccoli, although the association was not statistically significant (RR: 0.82; 95% CI 0.65 to 1.02) [13].By type of study, a non-statistically significant association was observed in the case-control studies (RR: 0.60; 95% CI 0.32 to 1.13) as well as in the cohort studies (RR: 0.91; 95% CI 0.80 to 1.03) [13].Furthermore, the meta-analysis conducted by Tse et al., 2014 revealed that broccoli intake showed protective benefits against colorectal neoplasia (OR: 0.80, 95% CI: 0.65 to 0.99).It is worth noting that Tse et al. conducted a joint meta-analysis of both case-control and cohort studies, unlike our approach in this meta-analysis [12].
A similar trend was observed in cancer of the reproductive system.The findings from the cohort studies in this review on the impact of broccoli consumption on the risk of reproductive cancer [59,62] are consistent with the results reported by Hu et al., 2015, who observed a 22% reduction in the risk of ovarian cancer with marginal significance for broccoli (RR: 0.78; 95% CI 0.58 to 1.06) [18].
When examining thyroid cancer, a case-control study found no significant association between broccoli consumption and the risk of thyroid cancer [94].In contrast, findings from a prospective cohort study hinted at a possible positive association between broccoli consumption and thyroid cancer risk in men [60].However, it is important to note that the results in this study should be interpreted with caution due to the possibility of bias introduced by the presence of other natural goitrogens.Regarding lymphatic cancer, a cohort study suggested that broccoli intake could be associated with a lower risk of non-Hodgkin lymphoma and follicular lymphoma, with a less evident association with diffuse large B-cell lymphoma [61].
As a whole, the results of this review seem to be consistent with the results of previous systematic reviews, both with those that evaluated the impact of the consumption of cruciferous vegetables on cancer [7,11,12,14,17,19,20] and those that assessed the consumption of broccoli [12,13,96].In this regard, (a) the association between high broccoli consumption and the risk of suffering from various types of cancer showed an inverse trend, that is, the higher the consumption, the lower the risk; (b) the inverse association between higher broccoli intake and the risk of several specific cancers was identified in both case-control and cohort studies; (c) these findings should be interpretated with caution.Case-control studies showed marginal statistical significance with moderate heterogeneity, and further cohort studies are needed.
Several biological mechanisms have been proposed to determine the positive effect of broccoli uptake associated with cancer.The possible protective effect of broccoli could be explained, at least in part, by the chemopreventive and anticancer properties of the metabolites present in this cruciferous plant.Broccoli serves as a significant source of isothiocyanates, small biologically active molecules derived from glucosinolates.Sulforaphane, as an essential compound in broccoli, is an isothiocyanate with notable anticancer and chemopreventive properties [97].Sulforaphane plays a crucial role in diverse biological processes associated with cancer, including enzymatic detoxification of carcinogens, attenuation of oxidative stress, initiation of cell cycle arrest, promotion of apoptosis, and regulation of the epithelial-to-mesenchymal transition [97][98][99][100][101].Numerous studies have highlighted sulforaphane's effectiveness in targeting cancer stem cells across various cancer types, thereby enhancing its potential to prevent drug resistance, metastasis, and tumor recurrence.Sulforaphane has shown its effectiveness against various tumors, including lung cancer, prostate cancer, breast cancer, and colon cancer [97][98][99][100][101][102][103][104].
The observed variance in cancer risk associated with broccoli consumption across studies may predominantly stem from divergent characteristics within studied populations, encompassing variances not only across different cancer types but also across demographic, geographic, cultural, and genetic factors.Additionally, methodological disparities in the assessment of broccoli consumption, including variations in measurement tools and criteria for categorizing high versus low intake levels, could contribute to this heterogeneity.Discrepancies in broccoli preparation methods, notably differences in cooking techniques and the preservation of bioactive compounds, may further amplify this variability.Exposure to high temperatures during cooking can cause the degradation of myrosinase, decreasing its functionality and hindering the synthesis of sulforaphane.Therefore, it is preferable to opt for the consumption of raw broccoli to enhance the bioavailability and protective effects of sulforaphane [97].Moreover, inconsistencies in the adjustment for confounding factors, despite attempts at standardization across studies, introduce another layer of potential variation, given the divergent types and quantities of confounding factors considered.It is imperative not to discount other potential contributors to disparate findings, including dissimilarities in study methodologies or identified limitations encountered during critical analysis.Finally, it is important to mention that in the studies in which the impact of broccoli consumption was analyzed based on the presence of the GSTM1-null gene polymorphism that is associated with glutathione S-transferase inactivity [76,92], individuals with this polymorphism seemed to benefit more than other subgroups, a finding that could also explain the inconsistency and imprecision of the results.
The primary strength of this study lies in its substantial sample size, comprising 699,482 subjects in the cohort studies and 31,292 participants in the case-control studies.Such a large sample size bestows considerable statistical power, enabling the identification of a robust association between broccoli consumption and the risk of developing various types of cancer.
Several limitations must be considered.It should be considered that the definition of the exposure level varied depending on each study (maximum intake vs. minimum intake of broccoli).Furthermore, the methods for evaluating the level of broccoli intake were heterogeneous due to the types of surveys or tools to measure consumption in each study.The followed-up groups of people were also heterogeneous between the different studies, which could contribute to the inconsistency of the results.As the data from the conducted studies relied on observational methods, it is plausible that the observed inverse association between broccoli consumption and the risk of various cancer types could have been influenced by unmeasured variables or residual confounding factors.In addition, several biases were detected, mainly relating to the representativeness of the cases, selection of controls in hospital environments (Berkson bias), self-reported exposure, nonblinded interviews, and times for the collection of food consumption data.

Conclusions
This review and meta-analysis may be the most comprehensive to date due to the broad coverage of outcomes for various types of cancer related to broccoli consumption.
From a biological perspective, the consumption of broccoli, regardless of its varieties, shows a protective and chemoprotective effect on cancer and cancer biomarkers.From a methodological perspective, this beneficial effect of broccoli consumption on cancer should be interpreted with caution.Cohort studies should be increased in various specific cancer types, and case-control studies showed subtle moderate heterogeneity.
As a final remark, while broccoli is generally considered a healthy food choice and is associated with various health benefits, including potential cancer-preventive effects, it is important to note that there is generally no significant risk associated with high broccoli consumption for most individuals.However, certain groups may need to exercise caution, including individuals on warfarin medications and people with thyroid issues, allergies/hypersensitivities, or digestive sensitivities [105].
More in-depth studies are warranted to report more-detailed results and stratified results by different cancer types.

Figure 1 .
Figure 1.Flowchart of the article selection process.

Figure 1 .
Figure 1.Flowchart of the article selection process.

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.

Table 2 .
Summary of critical appraisal process based on the Newcastle-Ottawa Scale in case-control studies.

Table 3 .
Summary of critical appraisal process based on the Newcastle-Ottawa Scale in cohort studies.
Color legend: Green: response marked in the tool as the most appropriate option, with a low risk of bias; Red: answer marked in the tool as the least appropriate option; Orange: answer marked as doubtful.Cohort study dimensions: A.1.Representativeness of the exposed cohort (broccoli consumption); A.2. Selection of the unexposed cohort; A.3.Record of exposure; A.4. Verification of outcome not present at the beginning of the study; B1.Cohort comparability; C.1.Evaluation of the outcome; C.2. Sufficient exposure time; C.3.Adequate follow-up of the cohort.