Folate exposures and risk of colorectal cancer: an umbrella review of meta-analyses of observational studies and randomised controlled trials

Abstract

Objectives

To systematically summarise and evaluate the existing evidence of the associations between diverse folate exposures and the risk of colorectal cancer (CRC), while identifying evidence quality.

Design

Umbrella review of meta-analyses.

Data sources

PubMed, Web of Science, Cochrane and Embase were searched from the database inception to March 2024, with an update to 12 October 2025.

Eligibility criteria

We included meta-analyses of randomised controlled trials or observational studies that investigated the associations between folate exposures and CRC or precancerous lesions (ie, adenoma and polyps).

Data extraction and synthesis

For each association, we recalculated the summary effect size with 95% CI using the DerSimonian and Laird random-effects model, heterogeneity (I² statistic), 95% prediction interval, small-study effect (Egger’s test) and excess significance bias (χ² test).

Results

This umbrella review included five meta-analyses describing 10 associations between folate exposures and CRC risk. In the general population, moderate-quality evidence supported an inverse association between total folate intake (from foods and supplements) and CRC risk (RR 0.84; 95% CI 0.80 to 0.90), while low-quality evidence suggested inverse associations of dietary folate intake (from foods alone) (RR 0.88; 95% CI 0.81 to 0.96) and folic acid supplement intake (RR 0.83; 95% CI 0.77 to 0.90) with CRC risk. Among patients with inflammatory bowel disease, low-quality evidence suggested an inverse association between folic acid supplement intake and CRC incidence (HR 0.71; 95% CI 0.53 to 0.96). Additionally, elevated circulating folate levels were observed to have a provoking effect on advanced-stage tumours (OR 1.95; 95% CI 1.18 to 3.22; Grading of Recommendations Assessment, Development and Evaluation (GRADE): very low). Sensitivity analysis revealed a potential increased risk of adenoma recurrence associated with folic acid supplement use among patients with a history of adenoma (RR 1.05; 95% CI 0.86 to 1.29; GRADE: high).

Conclusions

These findings suggest that consuming dietary folate and total folate intake may be beneficial in CRC primary prevention. Specifically, folic acid supplements may inhibit colorectal carcinogenesis in normal tissues while promoting cancer in the established neoplastic foci.

PROSPERO registration number

CRD42024537550.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• This umbrella review was conducted according to a preregistered protocol.

• Standard tools were employed to assess methodological quality (A Measurement Tool to Assess Systematic Review−2) and quality of evidence (Grading of Recommendations Assessment, Development and Evaluation) of selected studies.

• The present umbrella review consists of only published meta-analyses; individual studies might be omitted if they were not evaluated in these meta-analyses.

• The umbrella review’s credibility is directly dependent on the included meta-analyses, as well as indirectly on the original studies; bias in the original studies could not be controlled.

Introduction

Colorectal cancer (CRC) ranked as the third most frequently diagnosed cancer and the second-leading cause of global cancer death in 2022.¹ Both genetic and environmental factors play a central role in the occurrence and progression of CRC. Established evidence suggests that healthy dietary and lifestyle habits can help protect against disease development.² Recent studies have increasingly investigated the impact of dietary factors on CRC, as these may represent modifiable risk factors.^{3 4} Previous literature supports inverse associations of higher dietary fibre and calcium intake with the incidence of CRC.⁵ However, the relationship between folate and CRC appears to be contradictory. For example, a cohort study revealed that folate intake had a protective effect, whereas a randomised controlled trial (RCT) concluded that folic acid (FA) supplementation increased the risk of colorectal adenomas (CRAs).^{6 7}

Folate (vitamin B₉), a water-soluble vitamin, cannot be synthesised in the body and must be obtained from the diet.⁸ It is essential for de novo synthesis of thymidine and purines. Since folate plays a significant role in DNA methylation, DNA repair and synthesis and maintenance of genetic stability, folate deficiency may increase the risk of tumour development.⁹ Numerous studies have explored the associations between folate and CRC, comprising numerous systematic reviews and meta-analyses with inconsistent results and diverse quality. Previous attempts to review this meta-analytic literature focused on specific folate exposure, such as dietary folate intake, total folate intake, FA supplement or blood levels.10,12 It is necessary to comprehensively evaluate all folate exposures examined in meta-analyses to provide guidance for both clinical practice and public health policy.

Several umbrella reviews have synthesised evidence on dietary factors and CRC risk,13,15 including one that specifically summarised the associations between folate and various health outcomes,¹⁵ with literature current through May 2018. However, these broader reviews either provided only a high-level summary of folate or did not conduct detailed, folate-specific subgroup analyses (eg, by age, sex, region, cancer site or fortification period). In addition, the expanding body of evidence on folate’s relationship with CRC has seen a proliferation of original research and meta-analyses since 2018, underscoring the imperative for an updated umbrella review to harmonise these advancements. Therefore, we conducted this umbrella review of meta-analyses to provide an updated and systematic summary of the associations between diverse folate exposures and the CRC risk and to assess the possible biases and quality of evidence of the identified associations.

Materials and methods

We performed an umbrella review to systematically collect and review published meta-analyses investigating the relationship between folate and CRC risk. The protocol has been registered in the International Prospective Register of Systematic Reviews (ID: CRD42024537550).

Search strategy

A systematic literature search was conducted in PubMed, Web of Science, Cochrane and Embase from database inception to March 2024, with an update to 12 October 2025, using a predefined search strategy. The full search strategy for all databases can be found in the online supplemental table S1. The titles, abstracts and full texts of potentially eligible articles were screened by two reviewers independently. A third reviewer resolved disagreements.

Eligibility criteria

Studies were included if they met the following criteria: (1) participants: all participants without restriction to age and specific health conditions; (2) exposure: folate intake (ie, dietary folate (folate from foods alone), total folate (folate from foods and supplements) and FA supplement) and blood folate; (3) outcome: CRC or precancerous lesions (ie, CRA and colorectal polyps); (4) study design: meta-analyses of either observational (case-control, nested case-control and cohort studies) or RCTs, published in peer-reviewed journals in English. When multiple meta-analyses examined the same association, we prioritised the most recent publication with the largest number of cases.

We did not consider meta-analyses that evaluated the effects of genetic polymorphisms related to folate metabolism on the CRC risk. Moreover, studies were excluded if they were primary studies, systematic reviews without meta-analyses or sources without full-text articles (conference abstracts, letters, comments).

Data extraction

Data were extracted independently by two reviewers, and discrepancies were resolved by a third reviewer on the research team. For each eligible article, the following information was extracted: the first author, publication year, number of included studies, study design, number of cases/controls or total participants, type of exposure (ie, dietary folate (folate from foods alone), total folate (folate from foods and supplements), FA supplement or blood folate), type of comparison, outcome, study-specific summary risk estimates (OR, RR, HR or standardised mean difference) together with the corresponding CIs.

In addition, the following data were extracted from the original studies included in eligible meta-analyses: the first author, publication year and region, study design, follow-up time (cohort study and RCT), number of cases/controls or total participants, age and sex of participants and effect size and corresponding 95% CI.

Assessment of methodological quality

Two reviewers assessed the methodological quality of the included meta-analyses by using A Measurement Tool to Assess Systematic Review (AMSTAR)−2 tool. Discrepancies were resolved by consulting a third reviewer. AMSTAR-2 is a validated and reliable measurement tool for meta-analyses of both interventional and observational research.¹⁶ The tool consists of 16 domains, including a review plan, literature search strategy and conduct, presentation of the results and statistical analysis, assessment and discussion of the risk of bias, explanation and discussion of heterogeneity, publication bias and potential conflicts of interest. Seven items are considered critical domains. The methodological quality classification criteria: (1) no or one non-critical weakness was proposed as ‘high’; (2) more than one non-critical weakness was proposed as ‘moderate’; (3) one critical flaw with or without non-critical weaknesses was proposed as ‘low’ and (4) more than one critical flaw with or without non-critical weaknesses was proposed as ‘critically low’.¹⁶

Statistical analyses

For each association, the summary effect sizes with corresponding 95% CIs were recalculated by using the DerSimonian and Laird random-effects model.¹⁷ Heterogeneity was assessed by the I² statistics. I² values below 25% or between 25% and 50% indicated low or moderate heterogeneity, respectively. Conversely, I² values exceeding 50% or 75% signalled significant or considerable heterogeneity in the data.¹⁸ We calculated the 95% prediction interval (PI), which accounts for between-study heterogeneity and quantifies the uncertainty in estimating the effect size for future studies investigating the same association.¹⁹ The small-study effects (ie, potential publication bias) were examined by Egger’s test and contour-enhanced funnel plots. ²⁰P value <0.10 was interpreted as statistically significant evidence suggesting the presence of small-study effects. In addition, we conducted an excess significance test to assess discrepancies between the observed and expected number of statistically significant studies (defined as p<0.05) using a χ² test. P value <0.10 was considered as the presence of excess significance bias. All analyses were performed using R V.4.3.1, using the meta package (V.7.0–0) and the metafor package (V.4.6–2).

Subgroup and sensitivity analyses

Subgroup analyses were performed using data from the meta-analyses of observational studies, where available, to assess the consistency of associations across different populations on the basis of sex, geographic region, alcohol consumption, tumour subsite, cancer stage, follow-up duration and FA fortification period.

If the meta-analysis pooled original studies included both cohort and case-control studies, we repeated the main analysis restricted to cohort studies to assess robustness. Further sensitivity analyses were conducted in meta-analyses of RCTs, including the exclusion of small-sized studies (n<100).

Evaluation of the quality of evidence

The quality of evidence of the included meta-analyses was evaluated using the Grading of Recommendations Assessment, Development and Evaluation (GRADE) approach.²¹ The evidence grade was classified as high, moderate, low or very low. RCTs start with high-quality evidence, and observational studies start with low-quality evidence. Limitations in study quality, inconsistency, uncertainty about directness, imprecise or sparse data or reporting bias can lower the grade of evidence, whereas a large effect or dose-response gradient can upgrade it.²¹ Two reviewers independently rated the quality of evidence, and any discrepancies were resolved by contacting a third reviewer.

We further assessed the quality of evidence for associations identified in meta-analyses of observational studies using criteria applied in previously published umbrella reviews.^{13 22} Statistically significant associations (p<0.05) from random-effects models were stratified into four evidence classes: convincing (I), highly suggestive (II), suggestive (III) or weak (IV). P value ≥0.05 indicated no statistically significant association. The detailed grading criteria are provided in the online supplemental table S2.

Patient and public involvement

Patients and/or the public were not involved in the design, conduct, reporting or dissemination plans of this research.

Results

Study selection

Following the protocol, we identified 375 articles after removing duplicates. The full literature search process is shown in figure 1. 58 articles were retrieved for full-text assessment. Of these, 53 articles were excluded, and the reasons for their exclusion are provided in the online supplemental table S3. The updated search results were presented in online supplemental eFigure 1 and table S4. Finally, five meta-analyses were included in this umbrella review.23,27

Figure 1. Flow diagram of study selection. RCT, randomised controlled trial.

Characteristics of included meta-analyses

The characteristics of all eligible studies are provided in table 1. Meta-analyses included in this umbrella review were published between 2017 and 2023. Among them, eight associations from four meta-analyses of observational studies were reported, and two associations from one meta-analysis of RCTs were reported. The number of participants ranged from 367 to 2 766 913, and the number of cases ranged from 160 to 29 351, with six associations having >1000 cases.

Table 1. Characteristics of included meta-analyses in the umbrella review.

Study	Exposure	Population	Outcome	Number and study design	Metric	Cases, n	Participants, n	AMSTAR-2
Fu et al23	Total folate	General	CRC	23 cohorts	RR	29 351	2 766 913	High
	Dietary folate	General	CRC	13 cohorts	RR	22 033	1 969 002
	FA supplement	General	CRC	Seven cohorts	RR	14 273	1 045 976
Burr et al24	FA supplement	Patients with IBD	CRC	Five cohorts + five C-C	HR	760	4451	Moderate
Sun et al25	RBC folate	General	CRA	Eight C-C	OR	774	1989	Critically low
Shiao et al26	Circulating folate	General	CRC	Nine C-C + eight NC-C	OR	5137	12 500	Critically low
	Circulating folate	General	CRA	Three C-C + one NC-C	OR	1051	2179
	Circulating folate	General	CRP	Three C-C	OR	160	367
Moazzen et al27	FA supplement	General	CRC	SIx RCTs	RR	1022	33 522	Moderate
	FA supplement	Patients with CRA history	CRA	Four RCTs	RR	444	1315

C-C: case-control study; NC-C: nested case-control study; RBC: red blood cell; AMSTAR, A Measurement Tool to Assess Systematic Reviews; C-C, case-control study; CRA, colorectal adenoma; CRC, colorectal cancer; CRP, colorectal polyps (adenomatous and hyperplastic polyps); FA, folic acid; HR, hazard ratio; IBD, inflammatory bowel disease; NC-C, nested case-control study; OR, odd ratio; RBC, red blood cell; RCT, randomised controlled trial; RR, risk ratio.

The methodological quality assessment using AMSTAR-2 revealed that one meta-analysis was of high quality and two meta-analyses were of moderate quality, and the remaining two were rated as critically low quality. The detailed methodological quality assessment can be found in figure 2.

Figure 2. Results of the methodological quality assessment of included meta-analyses using AMSTAR-2. AMSTAR-2, A Measurement Tool to Assess Systematic Reviews-2; MA, meta-analysis; PICO, population intervention/exposure comparator outcome; RoB, risk of bias. ^*critical domains.

Description and summary of associations

As shown in table 2, 10 associations from five meta-analyses were recalculated using random-effects models. Four associations from meta-analyses of observational studies demonstrated nominal statistical significance (p<0.05), while only one reached statistical significance at p<10^-6. Three of the four associations targeted the general population: higher intakes of total folate (RR 0.84; 95% CI 0.80 to 0.90; GRADE: moderate), dietary folate (RR 0.88; 95% CI 0.81 to 0.96; GRADE: low) and FA supplement (RR 0.83; 95% CI 0.77 to 0.90; GRADE: low), all of which indicated a reduced risk of CRC. The remaining one association targeted the patients with inflammatory bowel disease (IBD): higher intake of FA supplement (HR 0.71; 95% CI 0.53 to 0.96; GRADE: low) suggested a lower risk of CRC. FA supplement, however, failed to demonstrate statistical significance in the RCT-based meta-analysis, regardless of whether participants were from the general population (RR 0.93; 95% CI 0.77 to 1.12; GRADE: moderate) or had a prior adenoma diagnosis (RR 0.93; 95% CI 0.68 to 1.25; GRADE: low). None of the remaining four associations between blood folate levels and CRC risk achieved statistical significance (p>0.05).

Table 2. Associations between folate exposures and colorectal cancer risk.

Exposure	Population	Outcome	Comparison	Random effect size (95% CI)	P value	I2 %	GRADE	EC
Meta-analyses of observational studies
Total folate23	General	CRC	Highest vs lowest	0.84 (0.80 to 0.90)	<1×10-6	16.6	Moderate	I
Dietary folate23	General	CRC	Highest vs lowest	0.88 (0.81 to 0.96)	0.003	34.3	Low	IV
FA supplement23	General	CRC	Highest vs lowest	0.83 (0.77 to 0.90)	<1×10-5	0	Low	III
FA supplement24	Patients with IBD	CRC	Yes vs No	0.71 (0.53 to 0.96)	0.026	25.4	Low	IV
RBC folate25	General	CRA	Case vs control	1.71 (0.30 to 9.86)	0.548	98.3	Very low	No
Circulating folate26*	General	CRC	Case vs control	0.88 (0.73 to 1.07)	0.202	84.9	Very low	No
Circulating folate26*	General	CRA	Case vs control	0.83 (0.67 to 1.02)	0.081	32.4	Low	No
Circulating folate26*	General	CRP	Case vs control	1.05 (0.90 to 1.23)	0.547	0	Low	No
Meta-analyses of randomised controlled trials
FA supplement27	General	CRC	Treatment vs control	0.93 (0.77 to 1.12)	0.452	60.7	Moderate	NA
FA supplement27	Patients with CRA history	CRA	Treatment vs control	0.93 (0.68 to 1.25)	0.604	71.5	Low	NA

^*Plasma or serum level of folate.

CRA, colorectal adenoma; CRC, colorectal cancer; CRP, colorectal polyps (adenomatous and hyperplastic polyps); EC, evidence class (observational studies); FA, folic acid; GRADE, Grading of Recommendations Assessment, Development and Evaluation; IBD, inflammatory bowel disease; NA, not applicable; RBC, red blood cell.

Four associations (4/10) had significant or considerable heterogeneity (I²>50%) (online supplemental table S5). The effect sizes of the largest study were statistically significant (p<0.05) for four associations (4/10). Only one association (1/10) showed a 95% PI excluding the null value. Excess significance bias was found for one association, and small-study effects were not identified (online supplemental eFigure 2).

In meta-analyses of observational studies, one association met convincing evidence criteria for CRC risk reduction with higher total folate intake, and one association (FA supplement) reached suggestive evidence in the general population. The remaining associations were either weak evidence (two associations) or no evidence (four associations). In meta-analyses of RCTs, two associations were identified as moderate and low quality, respectively. The detailed GRADE scores for each meta-analysis are presented in the online supplemental table S6.

Subgroup analyses

Subgroup analyses were performed in the meta-analyses of observational studies; however, meta-analyses of RCTs were not conducted due to the limited number of eligible original studies. An inverse association between dietary folate intake and CRC risk in the general population was observed in the following subgroups: females (RR: 0.90; 95% CI 0.83 to 0.97), individuals in the USA (RR 0.83; 95% CI 0.75 to 0.91), those with alcohol intake ≥15 g/day (RR 0.79; 95% CI 0.71 to 0.89), colon-specific cases (RR 0.82; 95% CI 0.70 to 0.96), and studies with ≥15 years of follow-up (RR 0.86; 95% CI 0.78 to 0.95) (figure 3A). In addition, this association persisted in the USA both before (RR 0.79; 95% CI 0.64 to 0.98) and after (RR 0.84; 95% CI 0.75 to 0.93) the implementation of mandatory FA fortification. The protective effect of total folate intake against CRC remained statistically significant across subgroups stratified by sex (male and female), region (Europe and USA), tumour subsite (colon, proximal colon and distal colon) and follow-up duration (≥15 years or <15 years) (figure 3B). No statistically significant associations were observed after stratification by alcohol consumption status. Besides, this association was observed exclusively in the USA after the implementation of mandatory FA fortification (RR 0.82; 95% CI 0.77 to 0.89). In the general population, FA supplement has a statistically significant protective effect against CRC in the USA (RR 0.84; 95% CI 0.77 to 0.91) and for follow-up periods both less than (RR 0.82; 95% CI 0.74 to 0.90) and greater than (RR 0.85; 95% CI 0.75 to 0.96) 15 years (figure 3C). In the patients with IBD, FA supplement intake was associated with a lower CRC risk in the USA (RR 0.65; 95% CI 0.44 to 0.95) and prefortification period (RR 0.41; 95% CI 0.20 to 0.85) (figure 3D). Subgroup analyses of circulating folate levels demonstrated statistically significant associations with CRC in Asia (OR 0.72; 95% CI 0.52 to 0.99), advanced-stage tumours (III/IV) (OR 1.95; 95% CI 1.18 to 3.22) and colon-specific anatomical subsites (OR 1.61; 95% CI 1.11 to 2.34), while associations with CRA were restricted to US populations (OR 0.78; 95% CI 0.66 to 0.92) (figure 3E,F). The detailed GRADE scores for subgroup analyses are provided in online supplemental table S7.

Figure 3. Subgroup analyses of associations between folate exposures and colorectal cancer risk (A) Associations between dietary folate intake and CRC risk in the general population; (B) Associations between total folate intake and CRC risk in the general population; (C) Associations between FA supplement intake and CRC risk in the general population; (D) Associations between FA supplement intake and CRC risk in patients with IBD; (E) Associations between circulating folate level and CRC risk in the general population; (F) Associations between circulating folate level and CRA risk in the general population. CRA, colorectal adenoma; CRC, colorectal cancer; FA, folic acid; IBD, inflammatory bowel disease; NE, not estimated because the number of studies (<10) too small to test for small-study effects, or the number of studies (<3) too small to test for a 95% prediction interval; NP, not pertinent because the estimated number is larger than observed; O/E, observed/expected number of studies with significant results. PI, prediction interval.

Sensitivity analyses

After excluding case-control studies, no statistically significant association was observed between FA supplement and CRC risk in patients with IBD (RR: 0.77; 95% CI 0.47 to 1.27; GRADE: very low), and the direction of the association between circulating folate levels and CRC risk in the general population was altered (RR: 1.11; 95% CI 0.89 to 1.37; GRADE: very low). After excluding small-scale studies, we observed a non-statistically significant but elevated risk of CRA recurrence in individuals with a history of adenomas (RR: 1.05; 95% CI 0.86 to 1.29), with the strength of evidence being upgraded from low to high (online supplemental tables S8, S9).

Discussion

The present umbrella review included five published meta-analyses containing 10 summary risk estimates of the associations between folate exposures and CRC risk. In meta-analyses of observational studies, four statistically significant associations were identified. Among the general population, moderate-quality evidence supported an inverse association between total folate intake and CRC incidence, while low-quality evidence suggested inverse associations for both dietary folate intake and FA supplement intake with CRC risk. In patients with IBD, low-quality evidence indicated an inverse association between FA supplement intake and CRC incidence. Nevertheless, the FA supplement failed to demonstrate statistical significance in meta-analyses of RCTs.

High folate intake conveys a protective effect against CRC in the general population. The effect of folate on CRC is influenced by many factors. The results of the stratified analyses showed that dietary folate intake exerts no chemopreventive efficacy in non-drinkers or low-alcohol consumers (<15 g/day), yet confers significant CRC risk reduction in moderate-heavy drinkers (≥15 g/day). A similar trend was observed in relation to total folate intake, although this did not attain statistical significance. Excessive alcohol leads to DNA hypomethylation through significant reductions in S-adenosylmethionine (SAM) levels, impairing folate absorption and the one-carbon cycle and consequently promoting cancer progression.²⁸ Elevated folate intake compensates for alcohol-induced depletion of methyl donors in chronic consumers, restoring hepatic SAM homeostasis and maintaining one-carbon metabolic flux, thereby attenuating ethanol-mediated DNA hypomethylation and associated carcinogenic pathways through epigenetic regulation.

The hierarchical analysis also indicated that folate’s preventive effect is more pronounced in the colon than in the rectum, especially in the distal colon. It is speculated that the colon epithelium highly expresses the reduced folate carrier and the proton-coupled folate transporter, whereas the density of these transporters is lower in rectal cells, leading to differences in folate uptake efficiency.²⁹ Intestinal bacteria may also have an impact on folate bioavailability.^{30 31} Further research is required to determine the effect of folate on the different locations of CRC.

One of the most controversial and speculative applications of FA supplement is its potential impact on cancer prevention. The present umbrella review indicated a preventive effect of FA supplement in meta-analysis of cohort studies; however, this effect is not significant in meta-analysis of RCTs. The duration of follow-up is a crucial factor in the interpretation of the inverse associations we observed between folate intake and CRC risk, because certain studies have suggested a substantial latency period for the protective effect of FA supplement.^{6 32} It is estimated that the adenoma-carcinoma sequence progresses over a period of at least 10 years.³³ In the current umbrella review, RCTs were all conducted with a follow-up duration of under 10 years, while the vast majority of the cohort studies were conducted with a follow-up duration exceeding 10 years (only one study was conducted for 8 years). This may explain the null effects in a meta-analysis of RCTs of FA supplements in general population.

Notably, an elevated level of circulating folate was observed to have a provoking effect on advanced-stage tumours (III/IV); and sensitivity analysis from meta-analysis of RCTs also observed an adverse effect of FA supplements on CRA recurrence among individuals with previously resected CRA (a surrogate CRC endpoint, which is a precancerous lesion).³⁴ These results are in accordance with the hypothesis that folate exerts dual modulatory effects on colorectal carcinogenesis.^{35 36} Folate is preventive in the absence of paraneoplastic lesions; once early lesions are established, FA supplements appear to promote carcinogenesis. In a standard chemo-carcinogenic rodent CRC model, excessive folate intake (>20 times the basal daily dietary requirement) has been demonstrated to accelerate the onset and progression of CRC.³⁷ Furthermore, two animal models (Apc^Min and Apc+/−x Msh2−/− mice) have also revealed that moderate dosages of FA supplement (4–10 times the basal daily dietary requirement) after the establishment of neoplastic foci in the colorectal mucosa promote rather than suppress colorectal carcinogenesis.^{38 39}

Regarding the link between FA supplement and an elevated probability of CRA recurrence, one possible explanation is that a substantial proportion of patients with early precursor lesions could have been included in these trials despite undergoing resection of adenoma at baseline. This is because endoscopy has a miss rate and incomplete resection of colorectal polyps.⁴⁰ Supplemental FA may accelerate the proliferation and growth of these residual adenomas. On the other hand, genetic variation in the 3p25.2 locus (in the region of Synapsin II (SYN2)/tissue inhibitor of metalloproteinase 4 (TIMP4)) might modify the link between FA supplement and CRC risk. Variant rs150924902 (located upstream to SYN2) shows the strongest interaction: FA supplement was associated with decreased CRC risk among those carrying the TT genotype but increased CRC risk among those carrying the TA genotype.⁴¹ However, genotypes have not yet been investigated in the included studies, and they may also affect the results.

One potential mechanism underlying the folate-mediated promotion of CRC growth is its ability to provide nucleotide precursors to rapidly replicating neoplastic cells. Rapidly dividing malignant cells require a considerable quantity of nucleotides for DNA replication, because nucleotides are the fundamental units of DNA.⁴² Folate can stimulate nucleotide synthesis, giving neoplastic cell proliferation and growth advantages.⁴³ Besides, incorporating methyl groups into DNA at promoter CpG islands of tumour suppressor genes is an epigenetic modification that can silence gene expression.⁴⁴ Folate can possibly promote CRC through enhancing the DNA methylation process and thus suppressing tumour suppressor gene expression.⁴⁵ However, folate status appears to have the opposite role in normal tissues compared with its impact on established neoplasms, in which folate sufficiency in normal tissues may suppress carcinogenesis. The biological plausibility of the concept since folate plays a critical role in DNA synthesis, repair, methylation and critical enzyme regulation.⁴⁶

Given the aforementioned results, it can be reasonably inferred that for folate to be a safe and effective chemoprevention against CRC, an appropriate dosage of folate intervention should be administered in individuals free of any signs of neoplastic foci and relevant genotypes. It should be noted that genetic testing and identification of early lesions in the general population are both challenging endeavours. Considering that an appropriate dosage and timing of FA intervention are vital for ensuring the safety and efficacy of chemoprevention, well-designed animal studies determining the effects of FA supplement on CRC development and progression are urgently needed.

This study should also be interpreted cautiously, as it has potential limitations. First, the present umbrella review consists of only published meta-analyses; individual studies might be omitted if they were not evaluated in these meta-analyses. Second, the results of this review from observational studies may be influenced by selection and recall bias. Finally, the umbrella review’s credibility is directly dependent on the included meta-analyses, as well as indirectly on the original studies. Bias in the original studies could not be controlled.

Conclusions

The umbrella review findings suggest that dietary folate and total folate intake have a possible beneficial effect in the primary prevention of CRC. It is recommended that FA supplements be used cautiously, particularly in individuals with pre-existing neoplastic lesions. Further high-quality studies are necessary to draw a definitive consensus regarding the appropriate doses, target populations and long-term health effects and to provide public health recommendations.

Data availability statement

Data are available upon reasonable request.