Scientific opinion on the tolerable upper intake level for folate

View full-text article in PMC

. 2023 Nov 13;21(11):e08353. doi: 10.2903/j.efsa.2023.8353

Reference Study name Country Study design Follow‐up Funding	Original cohort (N total) Exclusion criteria Study population (n)	Ascertainment of outcome	Exposure groups n/person‐years Exposure assessment method	Incident cases	Model covariates	Results
Clarke et al. (2007) The Oxford Healthy Aging Project UK PC 10 yrs Public	N = 2741 Population sampled: ≥65 yrs, residents in Oxford UK % lost to follow‐up: 79% n = 574 Sex (as % women): 60% Age (mean ± SD): 71.9 ± 5.2 MMSE (mean ± SD): 27.3 ± 2.7	Cognitive function assessed with the use of the MMSE on 3 occasions during 10 y	S‐holoTC, mean ± SD (pmol/L;): 73 ± 43 S‐cobalamin, mean ± SD (pmol/L): 280 ± 106 S‐Folate, mean ± SD (nmol/L): 15.8 ± 14.6 S‐holoTC by enzyme‐linked immunoabsorbent assay method modified for use on an automated analyser S‐cobalamin by automated chemiluminescence system S‐Folate by microbiological method	Outcome: change in MMSE score	Model 1: Sex, education, smoking, history of vascular disease, systolic blood pressure, APOE, genotypes Model 2: Model 1+ all markers of vitamin status (holoTC, tHcy, MMA, cobalamin, folate)	Change in cognitive decline (MMSE) per 100% increase in folate concentration ( β (SE) ): Model 1: 0.015 (0.012), p = NR Model 2: 0.004 (0.013), p = NR
Morris et al. (2012) Framingham Heart Study US PC 8 yrs Public	N = 5209 (examination 20, n = 1401) Population sampled: 2/3 of the adult population in Framingham, MA, US % lost to follow‐up: NR n = 549 Age (mean ± SD): 74.8 ± 4.6 MMSE (mean ± SD): 28.2 ± 2.1 MMSE <18 at examination 20 (1986–1990), non‐diabetic, normal kidney function, valid FFQ, baseline p‐cobalamin, p‐folate. Those with p‐cobalamin concentrations >701 pmol/L were excluded	MMSE every 2 years	p‐Folate (nmol/L): Q1 (ref): <5 Q2: 5–7.69 Q3: 7.7–12 Q4: 12.01–21.7 Q5: ≥21.75 Folate intake (μg/day) per quintile of p‐Folate (geometric mean (95% CI): Q1: 210 (190–232) Q2: 269 (248–291) Q3: 322 (296–350) Q4: 381 (348–416) Q5: 488 (445–536) Folic acid supplement users (%): Q1: 4.9 Q2: 5.5 Q3: 10 Q4: 23 Q5: 50 N (among those with p‐cobalamin <258 pmol/L) Q1 (ref): 60 Q2: 64 Q3: 44 Q4: 39 Q5: 24 126‐item, semiquantitative FFQ P‐cobalamin: non‐fasting, radioassay kit P‐folate: non‐fasting, microbial (L. casei) assay	Outcome: change in MMSE score	Analysis 1: adjusted for age, sex, educational achievement, baseline BMI, alcohol use versus nonuse Analysis 2: age, sex, educational achievement, baseline serum creatinine status, body mass index, smoking status	Change in MMSE score per quintile of plasma folate among those with low (<258 pmol) p‐cobalamin (β (95% CI): Analysis 1: β (95% CI) Q1 (ref): −0.32 (−0.44 to −0.21) Q2: −0.32 (−0.42 to −0.21), p = 0.94 Q3: −0.22 (−0.35 to −0.08), p = 0.25 Q4: −0.28 (−0.42 to −0.14), p = 0.67 Q5: −0.92 (−1.09 to −0.74), p < 0.001 P for trend <0.001 Change in MMSE score per quintile of plasma folate among those without low (≥258 pmol) p‐cobalamin (β (95% CI): Analysis 1: β (95% CI) Q1 (ref): −0,18 (−0.30 to −0.06) Q2: −0.14 (−0.27 to −0.01), p = 0.62 Q3: −0.17 (−0.27 to −0.06), p = 0.85 Q4: −0.14 (−0.24 to −0.03), p = 0.57 Q5: −0.14 (−0.23 to −0.05), p = 0.59 p for trend 0.48 Analysis 2: β (95% CI) Change in MMSE score per use of folate supplements among those with low (<258 pmol) p‐cobalamin (β (95% CI): Non‐supplement users: −0.15 (−0.11 to −0.22) Supplement users: −0.1 (−0.04 to −0.24) Change in MMSE score per use of folate supplements among those without low (≥258 pmol) p‐cobalamin (β (95% CI): Non‐supplement users: −0.34 (−0.27 to −0‐37) Supplement users: −0.78 (−0.57 to −1.0) (p for interaction between cobalamin status and folate supplementation = 0.002) *numbers extracted from figure using graphreader.com
Doets et al., 2014 Hordaland Homocysteine Study/Hordaland Health Study Norway PC 6 yrs Public	N = 2841 Population sampled: apparently healthy residents of Bergen (Norway), who participated both in the Hordaland Homocysteine Study in 1992–3 and in the Hordaland Health Study in 1997–9% lost to follow‐up: NR n = 2203 Sex (as % women): 55% Age (mean, 95% CI): 72.5 (71.5, 73.6) Modified MMSE (median, P5th–P95th): 12 (10–12)	Six cognitive tests were performed at the end of follow‐up: modified versions of the MMSE and of the Digit Symbol Test; a short form of the Block Design; the Kendrick Object Learning; an abridged version of the Controlled Oral Word Association Test; the Trail Making Test Part A. Principal component analysis was used to create a summary score for cognitive performance	At baseline p‐folate, median (P5th‐P95th) (nmol/L): 12.5 (8.7, 20.9) p‐cobalamin, median (P5th‐P95th) (pmol/L): 338 (196, 595) At end of follow‐up p‐folate, median (P5th‐P95th) (nmol/L): 15.8 (12.0, 34.0) p‐cobalamin, median (P5th‐P95th) (pmol/L): 339 (192, 651) 9.5% participants used B‐vitamin supplements P‐folate, p‐cobalamin: non‐fasting, microbiological assays	Outcome: cognitive score	Sex, education level, history of CVD /hypertension, apoE‐14 genotype and creatinine.	Multivariate quantile regression and ordinary least‐squares regression of the cognitive score: No associations of plasma folate (β = 0.019, SE = 0.031, p = 0.540) or plasma cobalamin (β = 0.020, SE = 0.033, p = 0.541), measured at baseline, with cognitive performance and no indication of an interaction between plasma folate and cobalamin (β = 0.024, SE = 0.028, p = 0.394)

Reference Study name Country Study design Follow‐up Funding

Original cohort (N total) Exclusion criteria Study population (n)

Ascertainment of outcome

Exposure groups n/person‐years Exposure assessment method

Incident cases

Model covariates

Results

Clarke et al. (2007)

The Oxford Healthy

Aging Project

10 yrs

Public

N = 2741

Population sampled: ≥65 yrs, residents in Oxford UK

% lost to follow‐up: 79%

n = 574 Sex (as % women): 60%

Age (mean ± SD): 71.9 ± 5.2

MMSE (mean ± SD): 27.3 ± 2.7

Cognitive function assessed with the use of the MMSE on 3 occasions during 10 y

S‐holoTC, mean ± SD (pmol/L;): 73 ± 43

S‐cobalamin, mean ± SD (pmol/L): 280 ± 106

S‐Folate, mean ± SD (nmol/L): 15.8 ± 14.6

S‐holoTC by enzyme‐linked immunoabsorbent assay method modified for use on an automated analyser

S‐cobalamin by automated chemiluminescence system S‐Folate by microbiological method

Outcome: change in MMSE score

Model 1:

Sex, education, smoking, history of vascular disease, systolic blood pressure,

APOE, genotypes Model 2:

Model 1+ all markers of vitamin status (holoTC, tHcy, MMA, cobalamin, folate)

Change in cognitive decline (MMSE) per 100% increase in folate concentration ( β (SE) ):

Model 1: 0.015 (0.012), p = NR

Model 2: 0.004 (0.013), p = NR

Morris et al. (2012)

Framingham Heart Study

8 yrs

Public

N = 5209 (examination 20, n = 1401)

Population sampled: 2/3 of the adult population in Framingham, MA, US

% lost to follow‐up: NR

n = 549

Age (mean ± SD):

74.8 ± 4.6

MMSE (mean ± SD): 28.2 ± 2.1

MMSE <18 at examination 20 (1986–1990), non‐diabetic, normal kidney function, valid FFQ, baseline p‐cobalamin, p‐folate.

Those with p‐cobalamin concentrations >701 pmol/L were excluded

MMSE every 2 years

p‐Folate (nmol/L):

Q1 (ref): <5

Q2: 5–7.69

Q3: 7.7–12

Q4: 12.01–21.7

Q5: ≥21.75

Folate intake (μg/day) per quintile of p‐Folate (geometric mean (95% CI):

Q1: 210 (190–232)

Q2: 269 (248–291)

Q3: 322 (296–350)

Q4: 381 (348–416)

Q5: 488 (445–536)

Folic acid supplement users (%):

Q1: 4.9

Q2: 5.5

Q3: 10

Q4: 23

Q5: 50

N (among those with p‐cobalamin <258 pmol/L)

Q1 (ref): 60

Q2: 64

Q3: 44

Q4: 39

Q5: 24

126‐item, semiquantitative FFQ

P‐cobalamin: non‐fasting, radioassay kit

P‐folate: non‐fasting, microbial (L. casei) assay

Outcome: change in MMSE score

Analysis 1: adjusted for age, sex, educational achievement, baseline BMI, alcohol use versus nonuse

Analysis 2: age, sex, educational achievement,

baseline serum creatinine status, body mass index, smoking status

Change in MMSE score per quintile of plasma folate among those with low (<258 pmol) p‐cobalamin (β (95% CI):

Analysis 1: β (95% CI)

Q1 (ref): −0.32 (−0.44 to −0.21)

Q2: −0.32 (−0.42 to −0.21), p = 0.94

Q3: −0.22 (−0.35 to −0.08), p = 0.25

Q4: −0.28 (−0.42 to −0.14), p = 0.67

Q5: −0.92 (−1.09 to −0.74), p < 0.001

P for trend <0.001

Change in MMSE score per quintile of plasma folate among those without low (≥258 pmol) p‐cobalamin (β (95% CI):

Analysis 1: β (95% CI)

Q1 (ref): −0,18 (−0.30 to −0.06)

Q2: −0.14 (−0.27 to −0.01), p = 0.62

Q3: −0.17 (−0.27 to −0.06), p = 0.85

Q4: −0.14 (−0.24 to −0.03), p = 0.57

Q5: −0.14 (−0.23 to −0.05), p = 0.59

p for trend 0.48

Analysis 2: β (95% CI)

Change in MMSE score per use of folate supplements among those with low (<258 pmol) p‐cobalamin (β (95% CI)*:

Non‐supplement users: −0.15 (−0.11 to −0.22)

Supplement users: −0.1 (−0.04 to −0.24)

Change in MMSE score per use of folate supplements among those without low (≥258 pmol) p‐cobalamin (β (95% CI)*:

Non‐supplement users: −0.34 (−0.27 to −0‐37)

Supplement users: −0.78 (−0.57 to −1.0)

(p for interaction between cobalamin status and folate supplementation = 0.002)

*numbers extracted from figure using graphreader.com

Doets et al., 2014

Hordaland Homocysteine Study/Hordaland Health Study

Norway

6 yrs

Public

N = 2841

Population sampled: apparently healthy residents of Bergen (Norway), who participated both in the Hordaland Homocysteine Study in 1992–3 and in the Hordaland Health Study in 1997–9% lost to follow‐up: NR

n = 2203

Sex (as % women): 55%

Age (mean, 95% CI): 72.5 (71.5, 73.6)

Modified MMSE (median, P5th–P95th): 12 (10–12)

Six cognitive tests were performed at the end of follow‐up: modified versions of the MMSE and of the Digit Symbol Test; a short form of the Block Design; the Kendrick Object Learning; an abridged version of the Controlled Oral Word Association Test; the Trail Making Test Part A. Principal component analysis was used to create a summary score for cognitive performance

At baseline

p‐folate, median (P5th‐P95th) (nmol/L):

12.5 (8.7, 20.9)

p‐cobalamin, median (P5th‐P95th) (pmol/L):

338 (196, 595)

At end of follow‐up

p‐folate, median (P5th‐P95th) (nmol/L):

15.8 (12.0, 34.0)

p‐cobalamin, median (P5th‐P95th) (pmol/L):

339 (192, 651)

9.5% participants used B‐vitamin supplements

P‐folate, p‐cobalamin: non‐fasting, microbiological assays

Outcome: cognitive score

Sex, education level, history of CVD /hypertension, apoE‐14 genotype and creatinine.

Multivariate quantile regression and ordinary least‐squares regression of the cognitive score:

No associations of plasma folate (β = 0.019, SE = 0.031, p = 0.540) or plasma cobalamin

(β = 0.020, SE = 0.033, p = 0.541), measured at baseline, with cognitive performance and no indication of an interaction between plasma folate and cobalamin (β = 0.024, SE = 0.028, p = 0.394)

Abbreviations: AD, Alzheimer disease; APOE, apolipoprotein E; CI, Confidence Interval; FFQ, food frequency questionnaire; FU, follow‐up; G, group; holoTC, holotranscobalamin; MMA, methylmalonic acid; MMSE, mini‐mental state examination; NA, Not assessed; NR, not reported; p, plasma; PC, prospective cohort; RoB, risk of bias; RR, Relative risk; S, serum; SE, standard error; SWE, Sweden; tHcy, total homocysteine; yrs, years.