by Jack Norris, RD

Contents

• Summary
• Introduction
• European Food Safety Authority
  • Table 1. Upper Limit (UL) for supplemental folate
  • Prostate cancer and elevated folate levels
  • Table 2. EFSA Folate Studies: Summary of Appendix C Evidence Tables and Figures 10–16
• Umbrella reviews post-EFSA report
• Vegrim 2022: High-dose FA with anti-seizure medications and childhood cancer
• Vegrim 2025: High-dose FA, anti-seizure medication, and cancer in women
• Unmetabolized folic acid
• Conclusion
• Footnote A: Folic acid reduces neural tube defects
  • Evidence for NTD-prevention from supplements
  • Evidence for NTD-prevention from food fortification
• Bibliography

Disclaimer: While I’ve done my best to represent the true state of the evidence, and nothing here represents a fringe position, this article doesn’t constitute medical advice and shouldn’t be used to override instructions from your physician.

Note: I used AI for the literature review, data analysis, claim verification, and editing of this article. Last updated: April 2026.

Summary

Having reviewed the evidence, I don’t think healthy adults need to avoid multivitamins containing folic acid at the typical dose of 400 µg/day.

Folic acid is the synthetic form of folate (vitamin B9) used in fortified foods and supplements. It can prevent neural tube defects, which are serious birth defects of the brain and spinal cord. Because the neural tube closes just 28 days after conception, often before someone knows they’re pregnant, adequate folic acid intake before and in the earliest weeks of pregnancy is critical. Women of childbearing age who don’t regularly eat fortified foods like bread, pasta, and rice are at risk of inadequate intake.

Some practitioners advise avoiding multivitamins that contain folic acid due to concerns about cancer. Almost all standard multivitamins contain folic acid, so this isn’t a trivial recommendation.

The European Food Safety Authority (EFSA, 2023), in a comprehensive review, set the upper intake level for folic acid at 1,000 µg per day. This limit was established not because cancer risk begins at that level, but to prevent folic acid from masking the early signs of vitamin B12 deficiency, a separate and well-characterized concern. The standard multivitamin dose of 400 µg is well below this threshold.

On cancer specifically, the evidence is mixed, voluminous, and difficult to summarize. Randomized trials and large prospective studies generally show no increased risk of cancer from folic acid intakes, including supplements at typical doses. Some studies find elevated risk with high blood folate levels, but these are difficult to interpret because people with undetected cancer may already have altered folate metabolism, making it look as though high folate caused the cancer when the reverse may be true. The EFSA concluded the evidence was insufficient to establish a causal relationship between folic acid intake and cancer risk at or below 1,000 µg per day.

For a healthy adult, a standard multivitamin containing 400 µg of folic acid appears safe based on the current evidence, and for women who could become pregnant, it may be important.

Introduction

Folate is a vitamin, sometimes referred to as vitamin B9, found naturally in foods like leafy green vegetables, beans, citrus fruits, and liver. It plays a critical role in the earliest stages of fetal development, when the neural tube, the embryonic structure that becomes the brain and spinal cord, is forming and closing, a process completed by just 28 days after conception, often before a woman knows she’s pregnant.

Folate isn’t a stable molecule and can’t be reliably used in fortified foods and supplements, so a synthetic version, folic acid, is used instead. Because many pregnancies are unplanned, and because the neural tube closes so early, relying on supplements after someone has learned they’re pregnant is too late. Fortifying staple foods like bread, pasta, rice, and corn tortillas ensures that many more women of childbearing age will automatically maintain adequate levels, without requiring any change in behavior. See Footnote A for a short summary of the evidence that folic acid reduces the risk of neural tube defects (NTDs).

Folate’s importance extends beyond pregnancy. It also plays a role in regulating homocysteine, an amino acid that, at elevated levels, has been associated with cardiovascular disease and cognitive decline, though whether homocysteine is a cause or simply a marker of inadequate B vitamin status remains unclear.

Folate must be changed to 5-MTHF in order to function as a co-enzyme in the methylfolate cycle. People who have a genetic variant (MTHFR 677C→T) don’t convert folate to 5-MTHF at the same rate as people who don’t. People with two copies of the variant, TT homozygotes, have red blood cell folate concentrations roughly 16% lower than those with no copies, even when consuming the same amount of folate (Tsang, 2015). People with one copy fall in between. The homozygous prevalence in the United States is 19.4% in Mexican Americans, 11.6% in non-Hispanic whites, and 1.2% in non-Hispanic blacks (Yang, 2012, Table 2). To my knowledge, no studies have specifically examined NTD rates in offspring of TT homozygous women before and after the era of folic acid fortification.

Vitamin B12 plays a role in folate metabolism. Without adequate vitamin B12, food folate can become “trapped” in the methylfolate cycle, where it’s unable to perform its function of facilitating DNA synthesis. In contrast, folic acid bypasses the need for vitamin B12 for DNA synthesis; in so doing, folic acid can mask early signs of a B12-deficiency, though this is thought to occur only with folic acid intakes above 1,000 µg (EFSA, 2023). More information about the methylfolate trap can be found in Folate, B12, and the Methylfolate Trap.

While there are good reasons for people to ensure they get extra folate via folic acid in fortified foods and supplements, some practitioners have concerns about multivitamins that contain folic acid due to folic acid’s associations with cancer. Since almost all multivitamins contain folic acid, removing them as an option could increase the risk of NTDs in newborns, as most periconceptional women are unlikely to take the time and expense to find a multivitamin with 5-MTHF, the preferred alternative form.

The concern about cancer isn’t completely without merit—some studies have found an association, especially with regard to blood levels. This article reviews the research examining the link between folic acid and cancer. Given the extensive body of research, including multiple meta-analyses of meta-analyses, I can’t review it all. Instead, I’m starting with the European Food Safety Authority’s 2023 report.

European Food Safety Authority

The EFSA (2023) reviewed the evidence on potential harm from folate and folic acid supplementation to set an upper intake level. Table 1 provides the amounts of folic acid considered by the European Food Safety Authority to be unlikely to mask B12 deficiency for different age groups.

Table 1. Upper Limit (UL) for supplemental folate
Age group	UL males and females (µg/day)
4–6 months	200
7–11 months	200
1–3 years	200
4–6 years	300
7–10 years	400
11–14 years	600
15–17 years	800
Adults	1,000
Pregnant women	1,000
Lactating women	1,000
Supplemental folate includes folic acid from fortified foods and supplements, and supplements of 5-MTHF-glucosamine and l-5-MTHF-Ca. Source: EFSA, 2023, Section 4

Much of the data the EFSA examined concerned associations between folate status and colorectal and prostate cancer. Table 2 lists the studies they included in their analysis, drawn from the tables in their Appendix C. The EFSA excluded many other studies due to methodological concerns.

As shown in Table 2, folic acid intake, whether measured through diet, supplements, or controlled trials, was mostly associated with null findings or a lower risk of colorectal cancer. The most controlled evidence comes from randomized trials of adenomas, which are precancerous polyps. Two low-risk-of-bias RCTs using 500–1,000 µg/day found no increased risk of adenoma. One trial (Cole 2007/Passarelli 2019) at 1,000 µg/day found a statistically significant increased risk of one specific polyp subtype (sessile serrated adenomas/polyps; HR 1.94), though all other adenoma outcomes in that same trial were null.

Blood levels of folate told a somewhat different story, trending toward null or mildly increased colorectal cancer risk, a distinction worth noting because plasma folate reflects total body folate status, including diet and underlying metabolic factors, not just supplement use, making it harder to interpret as a direct effect of supplementation.

The EFSA concluded: “The Panel notes that evidence from RCTs on the relationship between folic acid supplementation and risk of recurrence of colorectal adenomas is inconsistent…The Panel considers that the available [body of evidence] from intervention studies is insufficient to conclude on a positive and causal relationship between dietary intake of folate and risk of CRC (p. 38).”

For prostate cancer, findings were mostly null, though among results that reached statistical significance, increased risk appeared more often than decreased risk. As with colorectal cancer, the EFSA found the evidence insufficient to establish harm at or below 1,000 µg per day, and called for more research, writing, “Further investigation of the relationship between high folate intake and the risk of cancer is needed, including colorectal and prostate cancer. Additional research is needed on the relationship between high folate intake and the risk of [sessile serrated adenomas/polyps] (p. 53).”

Importantly, EFSA did not set the 1,000 µg upper limit because that is where cancer risk begins. The limit was established primarily to avoid masking a B12 (cobalamin) deficiency: “[T]he Panel considers that the risk of progression of cobalamin-dependent neurological symptoms in cobalamin-deficient patients should be considered the most serious adverse effect of ‘high’ folic acid intake and used as the critical effect to establish an UL for folic acid (p. 51).”

Prostate cancer and elevated folate levels

When comparing studies that measured dietary folate intake against studies that measured blood folate levels, a pattern emerges, especially for prostate cancer: Higher blood folate levels tend to be associated with an increased cancer risk, while dietary intake studies find little or no association with higher folate/folic acid intakes.

One explanation for this discrepancy is reverse causality: prostate tumors may alter the body’s folate metabolism, raising blood folate levels. Men with undetected cancer at the time of blood draw would systematically show up in the higher folate categories, making it look like high folate caused the cancer when the cancer may have already been present. This problem would be especially pronounced in nested case-control studies, where a high number of participants might have had undiagnosed tumors when their blood was drawn, compared to controls, who by definition did not develop cancer.

Mendelian randomization, using genetic variants that naturally increase folate levels, can help untangle this. Burrows et al. (2020; preprint, not peer-reviewed) applied this approach and found no evidence that being genetically predisposed to higher serum folate levels increased prostate cancer risk; results pointed slightly toward protection (OR 0.87, 95% CI 0.71–1.06).

For colorectal cancer, the genetic analysis trended toward increased risk: OR 1.18 (95% CI 0.64–2.18) in the Genetic and Epidemiology of Colorectal Cancer Consortium cohort, and OR 1.45 (95% CI 0.85–2.47) in the UK Biobank (Burrows, 2020). This result should be read alongside the broader body of observational evidence, in which dietary folate intake is mostly associated with null or reduced colorectal cancer risk, and blood folate shows inconsistent results. Note that Burrows et al. didn’t measure blood folate levels.

Table 2. EFSA Folate Studies: Summary of Appendix C Evidence Tables and Figures 10–16
Study	Type	Parameter / Exposure	Outcome	Finding	RoB
Figure 10 / Appendix C.2.1b. RCTs: FA Supplementation and Recurrent Colorectal Adenomas
Cole, 2007 / Passarelli, 2019 Aspirin/Folate Polyp Prevention Study (AFPPS); USA	RCT	1,000 µg FA/d; 3 + 3–5 yrs	Recurrent Colorectal Adenomas (any, advanced, ≥3, SSA/P)	any adenoma 1.21 (0.99–1.47); advanced adenoma 1.20 (0.73–1.97); ≥3 adenomas 1.58 (0.87–2.86) (↑) SSA/P 1.94 (1.02–3.68)	Low
Logan, 2008 UkCAP; UK / Denmark	RCT	500 µg FA/d; 3 yrs	Recurrent Colorectal Adenomas (any, advanced)	1.07 (0.85–1.34) any adenoma; 0.98 (0.68–1.40) advanced	Low
Wu, 2009 NHS/HPFS FA prevention trial; USA	RCT	1,000 µg FA/d; 3–6.5 yrs	Recurrent Colorectal Adenomas	0.87 (0.65–1.16)	Low
Jaszewski, 2008 § USA	RCT	5,000 µg FA/d; 3 yrs	Mean number of adenomas per patient	(↓) Fewer adenomas in FA group (0.36 vs. 0.82; p=0.025)	Mod
Appendix C.2.1a. RCT: FA Supplementation and Colorectal Cancer
Qin, 2017 CSPPT; China	RCT	800 µg FA + enalapril vs. enalapril alone; 4.5 yrs	Colorectal Cancer	2.17 (0.82–5.70); p=0.117	Low
Figure 11 / Appendix C.2.2a. Observational Studies: Plasma/Serum Folate and Colorectal Cancer
Eussen, 2010 EPIC; DK, SE, FR, GR, DE, IT, NL, ES, GB	NCC	pF (quintiles); 3.6 yrs	Colorectal Cancer	p-trend 0.44	Low
Neuhouser, 2015 WHI-OS; USA	NCC	pF and rbcF (quartiles); ~15 yrs	Colorectal Cancer	rbcF p-trend 0.63; pF p-trend 0.80	Low
Rossi, 2006 1969 Busselton Health Survey; Australia	PC	sF and rbcF (quartiles); 23 yrs	Colorectal Cancer morbidity	per 1 µg/L decrease in rbcF 1.19 (0.93–1.54); p=0.18	Mod
Otani, 2008 JPHC; Japan	NCC	pF (quartiles per m/f); 11.5 yrs	Colorectal Cancer	p-trend 0.88 (men), 0.63 (women)	Low
Lee, 2012 NHS, HPFS, PHS; USA	NCC	pF (quartiles; pre-fortification samples); 24 yrs	Colorectal Cancer	(↑) Q4 vs. Q1 1.47 (1.07–2.01) p-trend 0.17; (↑) post-fortification era subgroup: Q2–4 vs. Q1 2.56 (1.09–6.02)	Low
Van Guelpen, 2006 Northern Sweden Health and Disease Cohort (MONICA/VIP/MSP); Sweden	NCC	pF (quintiles); 4.2 yrs	Colorectal Cancer	(↑) Q3 vs. Q1 2.00 (1.13–3.56) p-trend 0.325	Mod
Le Marchand, 2009 Multiethnic Cohort Study; USA	NCC	pF (quartiles; post-fortification era); ~10 yrs	Colorectal Cancer	Q4 vs. Q1 0.61 (0.33–1.13) p-trend 0.097	Mod
Gylling, 2014 Northern Sweden Health and Disease Study (VIP/MSP); Sweden	NCC	pF (tertiles); 11-24 yrs	Colorectal Cancer	(↑) T2 vs. T1 1.62 (1.08–2.42) p-trend 0.32	Mod
Geijsen, 2020 COLON / EnCoRe / CORSA / ColoCare; Netherlands, Austria, Germany, USA	PC	p/sF (tertiles) and UMFA (tertiles; 15% w/detectable FA); 3.7 yrs	Recurrent Colorectal Cancer	pFA T3 1.14 (0.81–1.60); (↑) UMFA T3 3.12 (1.22–8.00) p-trend 0.03	Mod
Weinstein, 2008 The Wei Study (ATBC); Finland	NCC	sF (quintiles; male smokers); 17 yrs	Colorectal Cancer	p-trend 0.68	Mod
Cho, 2015 NHS, HPFS; USA	NCC	UMFA (3 categories), ~16-20 yrs	Colorectal Cancer	UMFA G3 1.12 (0.81–1.55) p-trend 0.32	Low
Shrubsole, 2009 Shanghai Women’s Health Study; China	NCC	pF (tertiles, women); ~9 yrs	Colorectal Cancer	T3 1.2 (0.8–1.7) p-trend 0.30	Mod
Takata, 2014 Shanghai Men’s Health Study; China	NCC	pF (tertiles, men); 8 yrs	Colorectal Cancer	T3 1.33 (0.90–1.98) p-trend 0.15	Mod
Figure 12 / Appendix C.2.2b & C.2.2c. Observational Studies: Total Folate Intake and Colorectal Cancer (PCs and NCCs)
Glynn, 1996 ATBC Cancer Prevention Study; Finland	NCC	sF, dietary folate (quartiles, male smokers); 5-8 yrs	colon / rectal cancer	sF: null for colon and rectal; dietary: (↑) colon Q3 0.34 (0.13–0.88) p-trend 0.15	Mod
Kato, 1999 New York University Women’s Health Study; USA	NCC	sF, dietary folate (quartiles, women); 9 yrs	Colorectal Cancer	(↓) sF: inverse Q4 0.52 (0.27–0.97) p-trend 0.04; dietary: null	Mod
Zschäbitz, 2013 WHI-OS; USA	PC	Dietary, FA, and total folate (quartiles post-menopausal women); 11 yrs	Colorectal Cancer	Total folate intake: Q4 0.90 (0.74–1.10) p-trend 0.51	Low
Flood, 2002 BCDDP follow-up cohort; USA	PC	Total folate intake (quintiles; women); 8.5 yrs	Colorectal Cancer	Q5 1.01 (0.75–1.35) p-trend 0.67	Mod
Lee, 2011 NHS, HPFS; USA	PC	Dietary, FA, and total folate (quintiles); ~24 yrs	Colorectal Cancer	total folate Q5 0.85 (0.68–1.08) p-trend 0.07	Low
Razzak, 2012 Iowa Women’s Health Study; USA	PC	Total folate intake (quartiles; women); ~18 yrs	Colorectal Cancer	Q4 0.95 (0.76–1.20) p-trend 0.46	Mod
Gibson, 2011 NIH-AARP Diet and Health Study; USA	PC	Dietary, FA, and total folate (multiple categories); 9.1 yrs	Colorectal Cancer (pre- and post-fortification)	(↓) Post-fortification: highest total folate 0.70 (0.58–0.84) p-trend <0.001; Pre-fortification: null	Low
Le Marchand, 2005 Multiethnic Cohort Study; USA	NCC	Total folate intake (tertiles); Follow-up: NR	Colorectal Cancer by MTHFR genotype	(↓) Inverse for CC and TT genotypes (0.69 and 0.39 respectively at highest tertile); null for CT; p-interaction 0.22	Mod
Roswall, 2010 Diet, Cancer and Health Study; Denmark	PC	Total folate and supplemental FA intake (quartiles); 9-13 yrs	colon and rectal cancer	Colon: total folate Q4 0.67 (0.46–1.00) p-trend 0.23; Rectal: total folate Q4 1.06 (0.67-1.70) p-trend 0.46	Low
Wang, Wu (2021) † NHS; USA	PC	Total folate intake (quintiles, pre- and post-fortification); 36 yrs	Colorectal Cancer	Q5 0.92 (0.73–1.15); per 400 µg DFE/d 0.97 (0.88–1.07) p-trend 0.50	Low
Stevens, 2011 Cancer Prevention Study II (CPS-II) Nutrition Cohort; USA	PC	Total, dietary, FA (quintiles); ~8 yrs	Colorectal Cancer	(↓) total folate Q5 0.81 (0.66–0.99) p-trend 0.047	Low
Zhang, 2006 Women’s Health Study; USA	PC	Total folate intake (quintiles; women); 10.1 yrs	Colorectal Cancer	Q5 1.16 (0.76–1.79) p-trend 0.46	Mod
Kim, 2010 † Pooled meta-analysis (11 US/Dutch cohorts); USA / Netherlands	PCs	Total folate intake (quintiles); 6-15 yrs	Colorectal Cancer	(↓) Q5 0.85 (0.77–0.95) p-trend 0.02	Low
Figure 13 / Appendix C.2.2c. Observational Studies: Folate from Natural Sources and Fortified Foods and Colorectal Cancer (subset of C.2.2c studies)
Figure 13 plots results from the C.2.2c studies above, isolating the dietary folate exposure (natural + fortified, excluding supplements). None showed evidence of a positive association for this exposure type.
Appendix C.2.2d. Observational Study: Plasma Folate and Recurrent Colorectal Adenomas
Martínez, 2006 WBF and UDCA trials (used as cohort); USA	PC	pF (quartiles; total population and by multivitamin use); 2.6-3 yrs	Recurrent Colorectal Adenomas	(↓) Q4 0.74 (0.56–0.98) p-trend <0.01	Mod
Figure 14 / Appendix C.2.2e & C.2.2f. Observational Studies: Total Folate Intake and Incident/Recurrent Colorectal Adenomas
Martínez, 2004 WBF trial (used as cohort); USA	PC	pF and total folate intake (quartiles); 3 yrs	Recurrent Colorectal Adenomas	(↓) Total folate Q4 0.61 (0.42–0.89) p-trend 0.01; pF Q4 0.66 (0.46–0.97) p-trend 0.04	Mod
Baron, 1998 The Polyp Prevention Study; USA	PC	Total folate intake (quartiles), 10-14 yrs	Recurrent Colorectal Adenomas	Q4 1.11 (0.69–1.78) p-trend 0.57	Mod
Murphy, 2008 The Polyp Prevention Trial (used as cohort); USA	PC	Total folate intake (quartiles); 10-14 yrs	Recurrent Colorectal Adenomas	Q4 0.91 (0.67–1.23) p-value NR	Mod
He, 2018 ‡ NHS 1–2, HPFS; USA	PC	Total folate intake (quartiles); 18-20 yrs	SPs and conventional adenomas	SPs: null p-trend 0.79; (↓) Conventional adenomas: Q4 0.93 (0.87–0.99) p-trend 0.002	Low
Lee, 2011 NHS, HPFS; USA	PC	Total folate, dietary, FA (quintiles); ~24 yrs	Colorectal adenomas	(↓) Total folate Q5 0.73 (0.57–0.94) p-trend 0.005; (↓) Dietary folate Q5 0.71 (0.59-0.85) p-trend 0.008; (↓) FA Q5 0.75 (0.65–0.87) p-trend <0.001	Low
Appendix C.3.1. RCT: FA Supplementation and Prostate Cancer
Figueiredo, 2009 Aspirin/Folate Polyp Prevention Study (AFPPS); secondary analysis; USA	RCT	1,000 µg FA/d; 3 + 3–5 yrs	Prostate Cancer	(↑) 2.58 (1.14–5.86)	Low
Figure 15 / Appendix C.3.1 & C.3.2a. Intervention and Observational Studies: FA/Plasma Folate and Prostate Cancer
Weinstein, 2003 ATBC Study; Finland	NCC	sF (quartiles, smokers); FU: NR	Prostate Cancer	Q4 1.20 (0.74–1.94) p-trend 0.52	Mod
Johansson, 2008 EPIC; 7 European countries	NCC	p/sF (quintiles); 14 yrs	Prostate Cancer	(↑) Q4 1.62 (1.12–2.34) p-trend 0.46	Low
Essén, 2019 Swedish Apolipoprotein Mortality Risk; Sweden	PC	sF (3 categories); 13 yrs	Prostate Cancer	high vs. normal 0.73 (0.48–1.10) p-trend 0.07	Mod
Hultdin, 2005 Northern Sweden Health and Disease Cohort; Sweden	NCC	pF (quartiles)	Prostate Cancer	Q4 1.30 (0.74–2.24) p-trend 0.17	Mod
Rossi, 2006 1969 Busselton Health Survey; Australia	PC	sF, rbcF (quartiles); 4.9 yrs	Prostate Cancer	rbcF per 100 µg/L decrease 1.20 (0.96–1.52) p=0.10	Mod
Beilby, 2010 Wittenoom Mine & Mill former workers; Australia	NCC	sF (tertiles); 14 yrs	Prostate Cancer	T3 1.09 (0.48–2.46) p-trend 0.83	Mod
de Vogel, 2013 JANUS cohort; Norway	NCC	sF (quintiles); 15.6 yrs	Prostate Cancer	Q5 1.15 (0.97–1.37) (↑) p-trend 0.04	Low
Figure 16 / Appendix C.3.2b. Observational Studies: Total Dietary Folate Intake and Prostate Cancer
Weinstein, 2006 ATBC Study (used as cohort); Finland	PC	Total folate intake (quintiles; smokers); 5-8 yrs	Prostate Cancer	Q5 0.96 (0.81–1.15) p-trend 0.84	Mod
Roswall, 2013 Diet, Cancer and Health Cohort Study; Denmark	PC	Total and supplemental folate (quartiles); 14.3 yrs	Prostate Cancer	total Q4 0.87 (0.71–1.07) p-trend 0.23; supplemental Q4 0.93 (0.73–1.17) (↓) p-trend 0.0213	Low
Stevens, 2006 Cancer Prevention Study II Nutrition Cohort; USA	PC	Total and dietary folate (quintiles); 8-9 yrs	Prostate Cancer	(↑) Q4 1.11 (1.01–1.22) p-trend 0.35	Low
Finding key: (↓) Inverse = higher folate associated with lower risk (↑) Positive = higher folate associated with higher risk RoB key: EFSA grade for risk of bias; Tier 1 (Low) or Tier 2 (Mod – moderate) Model note: All results reflect the most fully adjusted model reported in each study. † Note on Kim et al. (2010) and Wang, Wu, et al. (2021): Excluded from Figure 12 for analytical comparability reasons (pooled cohort design and 36-year follow-up complexity, respectively). Both are included in EFSA’s assessment. ‡ Note on He et al. (2018): Excluded from Figure 14 because it assessed serrated polyps (SPs) as a distinct endpoint which isn’t directly comparable to conventional adenoma recurrence outcomes in the other plotted studies. The study is included in EFSA’s assessment. § Note on Jaszewski et al. (2008): Excluded from Figure 10 because its primary outcome (mean adenomas per patient) was not directly comparable to the risk ratio–based outcomes of the other three plotted RCTs. It is included in EFSA’s assessment. Source: EFSA, 2023
Abbreviations DFE – dietary folate equivalent FA – folic acid HPFS – Health Professionals Follow-up Study Mod – moderate risk of bias (EFSA Tier 2) MTHFR – methylenetetrahydrofolate reductase NCC – nested case-control study NHS – Nurses’ Health Study NR – not reported PC – prospective cohort study pF – plasma folate rbcF – red blood cell folate RCT – randomized controlled trial RoB – risk of bias sF – serum folate SP – serrated polyp SSA/P – sessile serrated adenoma/polyp UMFA – unmetabolized folic acid; FA that hasn’t been converted to another form

Umbrella reviews post-EFSA report

Two umbrella reviews (reviews of meta-analyses) have been published since the EFSA report.

Li et al. (2025) focused specifically on colorectal cancer and reached conclusions generally consistent with the EFSA: Dietary and total folate intake were associated with reduced CRC risk in observational studies, while folic acid supplementation showed no significant effect in RCTs. The only meta-analysis included in Li’s review that was more recent than the EFSA report was by Fu et al. (2023), which pooled data from seven prospective cohorts covering over one million participants and found folic acid supplement use associated with a lower risk of colorectal cancer (RR 0.83; 95% CI 0.77–0.90). Li et al. rated Fu’s methodology as high quality but graded the underlying evidence for this specific association as low quality (Table S6), reflecting limitations in the original cohort studies.

Yoo et al. (2026) conducted an umbrella review encompassing 168 unique associations across multiple cancer types. Their most credible findings, rated ‘highly suggestive,’ found a protective effect of folate on esophageal cancer, colorectal cancer, total cancer, and childhood brain and spinal cord tumors.

Vegrim 2022: High-dose FA with anti-seizure medications and childhood cancer

Some anti-seizure medications interfere with folate metabolism; because of this, women of childbearing age have often been prescribed high doses of folic acid (1,000 to 5,000 µg/day) to prevent NTDs.

A 2022 study by Vegrim et al., published in JAMA Neurology, examined cancer risk among children born to more than 3.3 million mothers in Denmark, Norway, and Sweden between 1997 and 2017. Among the 27,784 children born to mothers with epilepsy, those whose mothers had filled prescriptions for high-dose folic acid (≥1 mg/day) during pregnancy had a higher risk of childhood cancer compared with children of mothers with epilepsy who had not used high-dose folic acid (HR 2.7, 95% CI 1.2–6.3). In absolute terms, the cancer risk was 1.4% in the exposed group and 0.6% in the unexposed group. No increased risk was found in children of mothers without epilepsy.

The UK Teratology Information Service (UKTIS, 2025) reviewed these findings and concluded that while the study was of high quality, several important limitations precluded firm conclusions: the near-complete overlap between high-dose folic acid use and anti-seizure medication (93.2% of the exposed group used both) made it impossible to disentangle the two exposures; the number of cancer cases was small, producing wide confidence intervals; the long lag between prenatal exposure and cancer diagnosis left room for unmeasured confounding; and no other studies have investigated the same association. UKTIS didn’t alter its guidance recommending high-dose folic acid for women with epilepsy taking anti-seizure medications.

In 2024, the American Academy of Neurology, American Epilepsy Society, and Society for Maternal-Fetal Medicine issued a joint practice guideline that cited the Vegrim findings (citation e4) and recommended “at least 0.4 mg/day” of folic acid for people with epilepsy on anti-seizure medications, declining to endorse higher doses and calling for future studies to clarify optimal dosing. The guideline noted that folic acid supplementation at this level is likely associated with higher IQ and possibly associated with reduced autistic traits in children born to people with epilepsy on anti-seizure medications, making the cancer association a genuine clinical balancing act rather than a straightforward contraindication (Pack, 2024).

Vegrim 2025: High-dose FA, anti-seizure medication, and cancer in women

A Scandinavian cohort study (Vegrim, 2025) followed nearly 1.5 million women for a median of 5–7 years and found an association between high-dose prescription folic acid (≥1 mg/day) and cancer after a 6-month lag sensitivity analysis (HR 1.1, 95% CI 1.02–1.2; eTable 3). Note that “high-dose” here refers to the strength of the prescription rather than sustained intake: A woman who filled one prescription and stopped is counted in the same exposed group as one who took 5 mg/day for years. The total amount of folic acid accumulated over the follow-up period, estimated from the number of prescriptions filled, showed no dose-response relationship with cancer risk, weakening the case for causation.

For overall cancer, this translates to a number needed to harm (NNH) of approximately 1 in ~1,220 women over the study’s median 5-year follow-up period (calculated by applying the lag-adjusted HR of 1.1 to the unexposed incidence rate of 164 per 100,000 person-years, giving an estimated exposed rate of 180.4, an absolute risk increase of 16.4 per 100,000 person-years × 5 years = 82 per 100,000; NNH = 100,000 / 82 ≈ 1,220; lag-adjusted incidence rates were not reported so this is an approximation).

The only subsite for which they found a higher risk of cancer after sensitivity and multiple finding corrections was for a high-dose prescription of folic acid (≥1 mg/day) associated with a doubling of non-Hodgkin lymphoma risk (HR 1.9, 1.3–2.9). To put this in perspective, the NNH is ~4,545 (95% CI: ~2,500–11,000) per 5 years, meaning the absolute risk remains small even if the association is causal (calculations: 1 / (5-year absolute risk increase): incidence rate difference of 7.7 − 3.3 = 4.4 per 100,000 person-years × 5 years = 22 per 100,000; NNH = 100,000 / 22 ≈ 4,545).

Women taking anti-seizure medication should talk to their doctor about the level of folic acid that’s safe for them. It’s unlikely a multivitamin with 400 µg of folic acid would be of concern.

Unmetabolized folic acid

Unmetabolized folic acid (UMFA) is simply folic acid that hasn’t yet been converted to its active forms. In countries with folic acid fortification of foods, and for people taking folic acid supplements, UMFA is more commonly detectable in the blood, a routine consequence of intake that has prompted researchers to ask whether it has any health consequences.

Troen et al. (2006) found that among 105 healthy postmenopausal women, those with detectable UMFA in their fasting blood had ~23% lower natural killer (NK) cell activity than those without detectable folic acid. The association was dose-dependent in older women (60–75 years), with NK cell activity declining significantly as UMFA concentrations rose (p-trend=0.002). UMFA was detectable in 78% of fasting participants, suggesting that circulating unmetabolized folic acid is a routine consequence of folic acid intake at population-level exposures. Since NK cells are a frontline defense against both viral infections and nascent tumor cells, a sustained reduction in their effectiveness would be concerning.

These findings have several limitations. The study was cross-sectional, so causality cannot be established. The sample was small (105 women). The authors performed multiple comparisons across several folate measures and subgroups without correcting for the false discovery rate. Applying Benjamini-Hochberg correction across the nine trend tests reported in the paper, the p-trend=0.002 result in older women narrowly survives (BH threshold: 0.0056), but the headline finding of 23% lower NK cytotoxicity across all women (p=0.04) does not. The surviving result also comes from a post-hoc subgroup of roughly 45 women: the age stratification was not a pre-specified hypothesis but emerged from sequential exploration of the data, as the authors describe in the paper, which means the true number of implicit comparisons is larger than the nine used in the correction, and the p-value correspondingly less meaningful than it appears.

A subsequent randomized controlled trial by Paniz et al. (2017) found reduced NK cell count and killing ability in Brazilian adults supplemented with folic acid, but at a dose of 5,000 µg per day, well above the 1,000 µg upper intake level established by EFSA and far outside the range relevant to someone taking a daily multivitamin.

The most direct test of whether UMFA actually increases cancer risk in humans comes from three studies reviewed by the EFSA:

Cho et al. (2015), using 20 years of follow-up data from the Nurses’ Health Study and Health Professionals Follow-up Study, measured circulating UMFA and found no association with colorectal cancer risk overall (OR 1.12, 95% CI 0.81–1.55). Among people with detectable UMFA, those with the MTHFR CT or TT genotypes and with higher UMFA had an increased risk of CRC (RR 2.20, 95% CI 1.22–3.94). However, the authors flag this as potentially due to chance, and the supplementary data support that view with more than 20 subgroup comparisons tested and no multiplicity correction. One caveat: blood samples were collected before folic acid fortification, so UMFA levels were lower than those in the current population. The authors acknowledge that this may have contributed to the null result, though most detectable UMFA in the study came from supplement use at doses comparable to those in later studies, which somewhat limits the concern.

Geijsen et al. (2020) found that among people who already had colorectal cancer, having detectable UMFA wasn’t associated with recurrence (HR 1.18, 95% CI 0.84–1.66). However, among the 15% of participants with detectable UMFA, those with the highest levels compared to the lowest had a higher risk of recurrence (HR 3.12, 95% CI 1.22–8.00). This was a secondary finding based on a small number of cases and shouldn’t be generalized to a healthy population. The EFSA rated it “moderate risk of bias” primarily due to inadequate control for confounding and unreported attrition. Reverse causality cannot be excluded, since the follow-up was short, 3.7 years, and people with more aggressive recurrent disease may have had higher UMFA for reasons unrelated to supplementation.

The European Food Safety Authority, in its 2023 review, was aware of these concerns and evaluated the UMFA evidence directly, but concluded it was insufficient to factor into setting the upper intake level, while calling for further research into whether UMFA specifically affects biological pathways leading to adverse health effects (EFSA, 2023).

Note that secondary genotype analyses are typically exploratory; until a study is designed from the outset to test whether a specific genotype modifies UMFA-related cancer risk, such findings shouldn’t be used to advise people with any particular genotype to avoid folic acid supplements.

Conclusion

My personal assessment of the EFSA’s and subsequent findings is that folic acid supplementation at doses found in typical multivitamins has no impact on, or might even prevent, some cancers in a healthy population, while some types of cancer may increase blood folate levels. Although no authoritative body has explicitly stated this as their position, none recommends avoiding folic acid at typical multivitamin doses, which is consistent with this assessment.

Footnote A: Folic acid reduces neural tube defects

The evidence that folic acid reduces the risk of neural tube defects comes from two distinct lines of research: supplementation trials and natural experiments in national food fortification programs.

Evidence for NTD-prevention from supplements

In the pre-fortification era, several studies demonstrated that folic acid supplementation reduces the risk of NTDs. The Medical Research Council’s randomized double-blind trial showed that 4,000 µg/day of folic acid (alone) reduced NTD recurrence by 72% in high-risk women who had previously had an affected pregnancy (MRC Vitamin Study Research Group, 1991). A randomized controlled trial then found that periconceptional multivitamin supplementation containing 0.8 mg of folic acid reduced the incidence of first-occurrence NTDs compared with a control group (Czeizel, 1992).

Before China implemented a national fortification program, a large prospective cohort study of nearly 250,000 pregnant women in China found that daily periconceptional supplementation with just 400 micrograms of folic acid alone reduced NTD rates by 79% in a high-prevalence region and 41% in a lower-prevalence region, with highly compliant women in both regions converging on similarly low NTD rates regardless of their starting point (Berry, 1999, Table 4).

Evidence for NTD-prevention from food fortification

Botto et al. (2006) analyzed birth defect trends across 15 population-based registries in 11 countries through 2003. In regions where folic acid food fortification was mandated, the United States, Canada, and Australia, NTD rates fell significantly, ranging from roughly 10% to 35%, depending on the registry. In contrast, European countries that issued supplementation recommendations without fortification saw no meaningful change in NTD rates, effectively serving as a real-world control group demonstrating that recommendations alone are insufficient at the population level.

A 2010 CDC Grand Rounds report extended and reinforced these findings. United States’ blood folate levels changed from 21% of the population having deficient serum folate before fortification to under 1% afterward, providing a plausible mechanism for the concurrent 36% drop in NTD prevalence (from 10.8 to 6.9 per 10,000) recorded through 2006. The same report noted a persistent disparity: Hispanic women remained at significantly greater risk for NTD-affected pregnancies than non-Hispanic white women, even after general fortification, and specifically identified corn masa flour fortification as the targeted solution, projecting it would increase folic acid intake among Mexican-American women by 20% while minimally affecting other groups. On January 1, 2026, a law in California went into effect requiring corn tortillas to be fortified with folic acid, with other states to follow in June (CDC, 2010; CNN, 2026).

Bibliography

Berry RJ, Li Z, Erickson JD, et al. Prevention of neural-tube defects with folic acid in China. China-U.S. Collaborative Project for Neural Tube Defect Prevention. N Engl J Med. 1999 Nov 11;341(20):1485-90.

Botto LD, Lisi A, Bower C, et al. Trends of selected malformations in relation to folic acid recommendations and fortification: an international assessment. Birth Defects Res A Clin Mol Teratol. 2006 Oct;76(10):693-705.

Burrows K, Kazmi N, Haycock P, Tsilidis KK, Martin RM, Lewis SJ. Mendelian randomisation study exploring the associations of serum folate with pan and site-specific cancers. bioRxiv [Preprint]. 2020 Jun 13. [cited 2026 Apr 27]

Cho E, Zhang X, Townsend MK, Selhub J, Paul L, Rosner B, Fuchs CS, Willett WC, Giovannucci EL. Unmetabolized folic acid in prediagnostic plasma and the risk of colorectal cancer. J Natl Cancer Inst. 2015 Sep 15;107(12):djv260.

CNN. Corn tortillas in California now must contain folic acid. More states are looking at it. March 30, 2026.

Cordero A, Mulinare J, Berry RJ, Boyle C, Dietz W, Johnston R Jr, Leighton J, Popovic T. CDC Grand Rounds: Additional opportunities to prevent neural tube defects with folic acid fortification. MMWR Morb Mortal Wkly Rep. 2010;59(31):980–984.

Czeizel AE, Dudas I. Prevention of the first occurrence of neural-tube defects by periconceptional vitamin supplementation. N Engl J Med. 1992;327(26):1832–1835.

EFSA Panel on Nutrition, Novel Foods and Food Allergens (NDA Panel); Turck D, Bohn T, Castenmiller J, et al. Scientific opinion on the tolerable upper intake level for folate. EFSA J. 2023 Nov 13;21(11):e08353.

Fu H, He J, Li C, Deng Z, Chang H. Folate intake and risk of colorectal cancer: a systematic review and up-to-date meta-analysis of prospective studies. Eur J Cancer Prev. 2023 Mar 1;32(2):103-112.

Geijsen AJMR, Ulvik A, Gigic B, et al. Circulating folate and folic acid concentrations: associations with colorectal cancer recurrence and survival. JNCI Cancer Spectr. 2020 Jul 7;4(5):pkaa051.

Li T, Yin L, Li Y, Huang L, Zhang K, Wang H, Xu D, Yan J, Huang G. Folate exposures and risk of colorectal cancer: an umbrella review of meta-analyses of observational studies and randomised controlled trials. BMJ Open. 2025 Nov 28;15(11):e103637.

MRC Vitamin Study Research Group. Prevention of neural tube defects: results of the Medical Research Council Vitamin Study. Lancet. 1991 Jul 20;338(8760):131-7.

Pack AM, Oskoui M, Williams Roberson S, et al. Teratogenesis, perinatal, and neurodevelopmental outcomes after in utero exposure to antiseizure medication: practice guideline from the AAN, AES, and SMFM. Neurology. 2024 Jun;102(11):e209279.

Paniz C, Bertinato JF, Lucena MR, et al. A daily dose of 5 mg folic acid for 90 days is associated with increased serum unmetabolized folic acid and reduced natural killer cell cytotoxicity in healthy Brazilian adults. J Nutr. 2017 Sep;147(9):1677-1685.

Troen AM, Mitchell B, Sorensen B, Wener MH, Johnston A, Wood B, Selhub J, McTiernan A, Yasui Y, Oral E, Potter JD, Ulrich CM. Unmetabolized folic acid in plasma is associated with reduced natural killer cell cytotoxicity among postmenopausal women. J Nutr. 2006 Jan;136(1):189-94.

Tsang BL, Devine OJ, Cordero AM, et al. Assessing the association between the methylenetetrahydrofolate reductase (MTHFR) 677C>T polymorphism and blood folate concentrations: a systematic review and meta-analysis of trials and observational studies. Am J Clin Nutr. 2015 Jun;101(6):1286-94.

UK Teratology Information Service (UKTIS). High-dose folic acid in women with epilepsy and offspring cancer risks. UKTIS One-Page Summary. February 2025.

Vegrim HM, Dreier JW, Alvestad S, et al. Cancer risk in children of mothers with epilepsy and high-dose folic acid use during pregnancy. JAMA Neurol. 2022 Nov 1;79(11):1130-1138.

Vegrim HM, Dreier JW, Igland J, et al. High-dose folic acid use and cancer risk in women who have given birth: a register-based cohort study. Epilepsia. 2025 Jan;66(1):75-88.

Yang Q, Bailey L, Clarke R, et al. Prospective study of methylenetetrahydrofolate reductase (MTHFR) variant C677T and risk of all-cause and cardiovascular disease mortality among 6000 US adults. Am J Clin Nutr. 2012 May;95(5):1245-53.

Yoo S, Montazeri A, Bennett D, et al. Folate and global health umbrella review series, part 2: syntheses on cancers. J Glob Health. 2026 Feb 20;16:04102.