Abstract
Periconceptional folate intake decreases the risk of pediatric acute lymphoblastic leukemia (ALL); however, the mechanism is not fully understood. We sought to identify sites of DNA methylation measured at birth both responsive to periconceptional folate and associated with lymphoblasts at the time of ALL diagnosis in a “meet in the middle” analysis. Folate-associated differentially methylated regions (DMRs) were identified from retrospectively collected periconceptional maternal folate intake (by dietary source) and epigenome-wide DNA methylation status from archived dried neonatal blood spots for 189 ALL cases and 205 healthy matched controls from the California Childhood Leukemia Study (1995–2008). Folate and lymphoblast-associated DMRs overlapped at 17 sites related to total folate intake, 13 for folate from food, 10 for natural foods, 13 for fortified foods and 18 for folate-containing supplements. The majority of overlapping DMRs were in a concordant direction of effect for supplemental folate (16/18, P = 0.001). Opposite direction of effect was identified among lower income participants for food (3/19, P = 0.004) and natural folate (5/37, P < 0.001), the latter of which was specific to Hispanic participants of low-income (9/31, P = 0.029). These results indicate that dietary folate, in particular from natural food sources, may reduce risk of ALL through modulation of early DNA methylation patterns.
Subject terms: Cancer epidemiology, Acute lymphocytic leukaemia, Cancer prevention, Cancer epigenetics, Risk factors
Introduction
The etiology of pediatric acute lymphoblastic leukemia (ALL) is not fully understood; however, the disease is likely to arise in the intrauterine period via acquired genetic factors as influenced by early life environmental exposures and inherited genetic susceptibility risk factors [1–3]. Maternal nutrition [4, 5], including dietary [6] and supplemental [7, 8] intake of folate, is associated with reduced risk of ALL. Folate intake in the pre- and periconceptional periods [9] demonstrates greater inverse association with the development of childhood ALL than serum levels at birth [10], highlighting periconceptional development as a crucial timepoint influencing leukemia risk. While single nucleotide polymorphisms (SNP) in folate regulatory genes in children do not associate with ALL risk [11], interaction between maternal SNPs [12] and folate intake may influence this relationship [13]. While folate intake reduces ALL risk across all populations, children of Hispanic backgrounds [9, 14] and low education levels [15] specifically demonstrate enhanced risk-reducing benefits of maternal folate consumption.
DNA methylation is a stable, heritable marker largely established during embryogenesis and influenced by genetic, environmental and stochastic influences. Aberrant DNA methylation is a hallmark of pediatric ALL at diagnosis [16], while early alterations in DNA methylation patterns show a potential association with the future development of leukemia [17, 18]. Maternal folate intake in the periconceptional period was demonstrated to affect DNA methylation patterns in offspring [19, 20], providing a potential mechanism by which dietary intake of nutrients modulates development of ALL [21]. We previously showed that the influence of periconceptional folate on DNA methylation varies by the dietary and supplemental sources, as well as ethnicity and socioeconomic status [22]. We further identified variation in DNA methylation responsiveness to maternal folate in pediatric ALL cases compared to healthy controls. The manner in which folate-responsive DNA methylation sites may contribute to alteration in DNA methylation patterns at diagnosis is not fully understood. Prior work using a meet-in-the-middle approach [23] has shown consistency in the direction of effect at folate-responsive DNA methylation sites at birth and at diagnosis in pediatric ALL cases. Here, we employ a meet-in-the-middle approach to identify the relationship between early-life DNA methylation patterns responsive to maternal folate intake by dietary or supplemental source and alterations in methylation in leukemic blasts at the time of diagnosis using resources from the California Childhood Leukemia Study (CCLS).
Materials/subjects and methods
Participant and sample acquisition
Participants were identified from the CCLS, a population-based case-control study [24]. Cases were identified from children (age 0–14 years) diagnosed with leukemia at California hospitals between 1995 and 2008. Controls were randomly selected from birth certificates from the California Department of Public Health Office of Vital Records and matched by birth date (to match the date of diagnosis for cases), sex, maternal race, and Hispanic identity at a 1:1 or 2:1 ratio. Whole-genome DNA methylation analysis was conducted on a subset of cases with pediatric ALL and matched controls using the Illumina Infinium Methylation 450k or EPIC BeadChip platform [20, 25]. Dried blood spots (DBSs) were obtained from peripheral blood heel-stick samples from birth maintained by the California Department of Public Health. DNA was isolated from DBS samples and sent for DNA methylation analysis as described elsewhere [20].
Leukemic blasts were obtained from diagnostic bone marrow samples from CCLS cases with B-lineage ALL, while purified control B-cell samples (CD19+CD34+) were obtained from fetal bone marrow samples as described previously [26]. Set A included 227 leukemic blast samples obtained from bone marrow examination at the time of ALL diagnosis, along with CD19+CD34+ sorted pre-B-cells derived from 6 individual participants fetal bone marrow samples. Set B samples included 37 leukemic blast samples and 10 distinct CD19+CD34+ sorted pre-B-cells from fetal bone marrow controls. Mononuclear cell purification and genomic DNA isolation was completed as noted elsewhere [25]. A subset of participants underwent whole-genome DNA methylation profiling using the Illumina Infinium Methylation 450k array (Set A) or the EPIC BeadChip array (Set B).
DNA methylation processing
Extracted DNA from DBS, leukemic blast and pre-B cell samples was modified by sodium bisulfite for conversion of unmethylated cytosines to uracil using the EZ DNA Methylation Kit (Zymo Research) for use in genome-wide methylation assessment. Samples were subsequently sent for methylation assessment using the Illumina Infinium Methylation EPIC BeadChip platform (Illumina) [20, 25]. Raw IDAT (intensity data file) files were imported into R (Version 4.0.0, The R Foundation) for data preprocessing and normalization using the openSesame pipeline from the SeSAMe package [27] in minfi [28] with the distribution of signal background calibrated by type I probe out-of-band signal. Probes with a detection P-value threshold >0.05 were masked from further analysis. NOOB background subtraction was performed, followed by removal of residual background. Nonlinear scaling was used to correct for dye balance. Probes and participants with >5% missing values were removed. Missing values were imputed using the impute.knn function in the R package impute. Chromosome X and Y probes were included in analysis.
Folate quantification
Retrospective maternal dietary history was obtained by a modified Block food-frequency questionnaire (mBFFQ) [29] as described elsewhere [9, 22, 30]. Briefly, the mBFFQ surveyed the frequency of intake for common foods and supplements in the year prior to pregnancy and was administered at the time of diagnosis or enrollment for controls. A Spanish language version of the mBFFQ which included seven additional Hispanic-specific foods was offered. Folate intake was quantified as dietary folate equivalents (DFEs) as a total amount of intake, from food sources only (including both fortified and naturally occurring dietary folate sources), and from supplemental folic acid sources only such as multi-vitamins and pre-natal vitamins. Supplemental DFEs were obtained from the exact listed amounts on folic acid and multivitamin supplements. DFE measurements accounted for folic acid fortification in grain products starting in 1998 in the United States.
Data analysis and statistical approach
DNA methylation beta values ranging from 0 (fully unmethylated) to 1 (fully methylated) were logit-transformed to M-values for statistical modeling and analysis. To identify probes significantly associated with periconceptional folate intake, multivariable linear regression was conducted independently for five folate sources (total, food, fortified, natural, and supplemental DFEs) controlling for array batch effect, ALL case status, sex, the top 10 principal components (PCs) for nucleated cell proportions and top 10 PCs for genetic ancestry. Reference-Free Adjustment for Cell Type Composition (ReFACTor) was used to estimate nucleated cell proportions [31], whereas genetic ancestry was estimated using the EPISTRUCTURE [32] in GLINT [33]. Supplemental DFEs were treated categorically (0, 194, 360, and 540 DFEs), whereas the remainder were treated as continuous variables. To identify probes significantly associated with leukemic blasts at the time of ALL diagnosis, multivariable linear regression was again employed independently for Set A samples (450k array) and Set B samples (EPIC array) controlling for ALL subtype and batch effect. Meta-analysis of the Set A and Set B linear regression results was conducted using METAL [34]. Differentially methylated regions (DMRs) were identified through Comb-p [35] using a seed P value < 0.05, maximum probe distance of 1000 base pairs, a minimum of 2 probes pe region, and a significance threshold of Šidák-corrected P value < 0.05 for both the folate regression analysis (individually by the 5 folate sources) and for the leukemia meta-analysis. Gene set enrichment analysis of overlapping regions for gene ontology and KEGG pathway terms was conducted using the missMethyl [36] package in R. Overlapping regions were identified using the genomic coordinates of DMRs identified in Comb-p. The direction of effect in each region was obtained by averaging regression coefficients for all probes contained within the DMR. Significant concurrent direction of effect was assessed by a binomial test using the number of concurrent and total overlapping regions. Subgroup analysis was conducted by Hispanic ethnicity of child and household annual income, defined as low (<$75 000) and high (≥$75 000), as well as by both Hispanic ethnicity and income status. Analysis was not completed in participants of high-income status stratified by Hispanic ethnicity due to a low sample size in these groups.
Results
Leukemia associated CpGs
Analysis of DNA methylation significantly associated with leukemic blasts was conducted in two subsets (Table 1). A total of 450,183 CpGs passed QC measures for analysis for Set A, and 700,796 for Set B. Multi-variate linear regression analysis was completed independently for both subsets controlling for ALL subtype and batch effect. In Set A, a total of 67,906 CpGs met threshold for significance (P < 0.05) after Bonferroni correction for multiple comparisons, while 27,728 CpGs were significant in Set B (Fig. 1). Meta-analysis of the linear regression results from Sets A and Set B was completed for 354,039 overlapping CpGs between the two arrays, resulting in 66,034 CpGs significantly associated with leukemic blasts after Bonferroni correction. This encompasses 60.1% of ALL-associated sites identified by Nordlund et al. (5652 of 9406 CpGs) [37]. The majority of these significant CpGs were hypomethylated in ALL samples (49,429, 74.9%) compared to hypermethylated (16,509, 25.1%). Genomic inflation λ for the three models were 11.31 for Set A, 4.00 for Set B, and 13.01 for the meta-analysis. To identify significant DMRs associated with leukemic blasts, the meta-analysis results were subsequently analyzed using Comb-p, resulting in 22 247 significant DMRs after Šidák correction for multiple comparisons (Fig. 2, Supplementary Table S1).
Table 1.
Cohort characteristics.
| Folate dataset | |
|---|---|
| Total subjects (N) | 394 |
| Age at enrollment (years;median, IQR) | 4.6 (4.5) |
| Male | 219 |
| Female | 175 |
| Case | 189 |
| Control | 205 |
| Hispanic ethnicity | |
| Hispanic | 175 |
| Non-Hispanic White | 160 |
| Income status | |
| Low-Income | 258 |
| High-Income | 137 |
| Hispanic, Low-Income | 150 |
| Hispanic, High-Income | 25 |
| Non-Hispanic White, High-Income | 78 |
| Non-Hispanic White, High-Income | 25 |
| Leukemia dataset | |
| Set A | |
| Leukemic blast samples | 227 |
| Bone marrow CD19+CD34+ pre-B cells | 6 |
| Subtype | |
| Hyperdiploid | 73 |
| t(12;21) | 52 |
| t(11;19) | 6 |
| MLL | 2 |
| Other | 73 |
| Unknown | 21 |
| Set B | |
| Leukemic blast samples | 37 |
| Fetal bone marrow controls | 10 |
| Subtype | |
| Hyperdiploid | 2 |
| t(12;21) | 11 |
| Unknown | 24 |
Folate dataset included acute lymphoblastic leukemia cases and controls from the California Childhood Leukemia Study (CCLS) with DNA methylation data from archived neonatal blood spots and survey-based periconceptional folate intake data. Leukemia dataset included pediatric ALL cases with DNA methylation data from lymphoblast samples at the time of leukemia diagnosis and CD34+CD19+ pre-B cells from healthy controls. Set A sample data was assessed using the Illumina 450k DNA methylation array, while Set B samples were assessed with the EPIC array.
IQR interquartile range.
Fig. 1. Lymphoblast-associated variation in DNA methylation.
A Quantile-quantile plots of regression results comparing DNA methylation in lymphoblasts obtained at the time of ALL diagnosis and CD19+CD34+ pre-B cells. Genomic inflation values are shown for Set A (λ = 11.31), Set B (λ = 4.00) and for meta-analysis of both sets (λ = 13.01). B Volcano plots showing the distribution of P-values by regression coefficients or Z-scores in the case of meta-analysis results. A total of 69,906 CpGs met threshold for significance (P < 0.05) after Bonferroni correction for multiple comparisons for Set A, 27,728 CpGs for Set B, and 66,034 CpGs after meta-analysis of both datasets. Assessment of differentially methylated regions (DMRs) using Comb-P identified a total of 22,247 significant DMRs from meta-analysis results.
Fig. 2. Overlapping differentially methylated regions.
Differentially methylated regions (DMRs) were identified using Comb-P from two epigenome-wide association studies (EWAS) assessing 1) the relationship between periconceptional folate intake by source and DNA methylation at birth, including total folate (all sources), food folate (natural and fortified sources), and supplemental folic acid, and 2) DNA methylation variation in lymphoblasts at the time of pediatric acute lymphoblastic leukemia (ALL) diagnosis compared to healthy bone marrow or purified S2-phase pre-B lymphocytes. In the latter analysis, data from the California Childhood Leukemia Study (CCLS) were obtained at two separate timepoints (Set A and Set B) and subsequently meta-analyzed. To identify epigenomic regions with DNA methylation variation associated with both periconceptional folate at birth and lymphoblasts at the time of ALL diagnosis, DMRs from each study were subsequently assessed for overlapping genomic regions using a ‘Meet in the Middle’ approach. Overlapping DMRs were further analyzed to identify concordance in the direction of DNA methylation (hypermethylation vs hypomethylation). Binomial test of the concordant versus total overlapping regions identified a significant bias toward a concordant direction of effect in DMRs associated with supplemental folic acid (16 of 18 DMRs, P = 0.001).
Folate-associated CpGs
A total of 394 participants with periconceptional folate intake survey data, including 189 ALL cases and 205 controls, underwent DNA methylation profiling from DBS samples obtained at birth using the EPIC array (Table 1). Mean folate intake by ALL case status and by demographic group are shown in Table 2. There was no significant difference between ALL cases and controls for total, food, fortified, and natural DFE categories. ALL cases within Hispanic (P = 0.016, Fisher exact test) and Hispanic low-income subjects (P = 0.038) demonstrated significantly lower supplemental folate intake. High-income subjects, in contrast, demonstrated significantly higher intake of supplemental DFEs in ALL cases (P = 0.020). Across demographic subgroups (Supplementary Table S2), Hispanic subjects had significantly higher intake of food, fortified, and natural DFEs compared to non-Hispanic White subjects (Two sample Student’s T Test, P = 3.55 × 10−5, 0.015, and 9.28 × 10−6, respectively). Low-income subjects had significantly lower intake of total DFEs compared to high-income subjects (P = 0.016), despite also having higher intake of natural DFEs (P = 0.001). Hispanic subjects of low-income had significantly higher intake of food and natural DFEs (P = 6.48 × 10−4 and 1.93 × 10−4). Multi-variable linear regression analysis, controlling for ALL case status, sex, batch effect, nucleated cell composition (from the top 10 principal components derived from ReFACTor analysis) and ethnic structure (from the top 10 principal components derived from EPISTRUCTURE analysis), was completed on 700 754 CpGs passing QC measures. Linear regression was completed for DFEs independently for 5 dietary folate sources, including total DFEs, food-derived DFEs, fortified food DFEs, naturally occurring food DFEs, and supplemental folic acid DFEs. Comb-P was subsequently used to identify significant DMRs for meet-in-the-middle analysis, resulting in 30 significant DMRs for total DFEs, 32 for food DFEs, 28 for fortified DFEs, 32 for natural DFEs, and 35 for supplemental DFEs (Fig. 2).
Table 2.
Folate distribution by demographic subgroup.
| All subjects | All (N = 394) | Cases (n = 188) | Controls (n = 206) | P |
|---|---|---|---|---|
| Total DFEs | 635.7 (391.9) | 636.9 (388.3) | 634.6 (396.1) | 0.954 |
| Food DFEs | 480.9 (276.8) | 480.8 (274.1) | 480.9 (279.9) | 0.999 |
| Fortified DFEs | 203.5 (193.9) | 199.9 (194.2) | 206.8 (194.0) | 0.725 |
| Natural DFEs | 277.4 (154.3) | 281.0 (156.0) | 274.1 (153.0) | 0.661 |
| Supplemental DFEs | ||||
| 0 | 281 | 134 | 147 | 0.893 |
| 194 | 27 | 12 | 15 | |
| 360 | 14 | 8 | 6 | |
| 540 | 72 | 34 | 38 | |
| Hispanic subjects | All (N = 174) | Cases (n = 82) | Controls (n = 92) | P |
| Total DFEs | 644.6 (397.0) | 595.9 (370.7) | 688.0 (416.4) | 0.125 |
| Food DFEs | 549.2 (320.3) | 547.4 (319.8) | 550.7 (322.5) | 0.945 |
| Fortified DFEs | 234.1 (217.4) | 238.8 (221.0) | 230.0 (215.2) | 0.791 |
| Natural DFEs | 315.0 (179.0) | 308.6 (177.7) | 320.8 (180.9) | 0.654 |
| Supplemental DFEs | ||||
| 0 | 142 | 75 | 70 | 0.016* |
| 194 | 9 | 5 | 4 | |
| 360 | 4 | 2 | 2 | |
| 540 | 19 | 3 | 16 | |
| Non-Hispanic White subjects | All (N = 160) | Cases (n = 75) | Controls (n = 85) | P |
| Total DFEs | 697.9 (408.1) | 697.9 (408.1) | 575.9 (370.9) | 0.051 |
| Food DFEs | 428.5 (208.5) | 428.5 (208.5) | 420.5 (226.8) | 0.817 |
| Fortified DFEs | 168.2 (149.2) | 168.2 (149.2) | 194.7 (184.6) | 0.316 |
| Natural DFEs | 260.3 (124.1) | 260.3 (124.1) | 225.7 (97.4) | 0.055 |
| Supplemental DFEs | ||||
| 0 | 101 | 42 | 59 | 0.053 |
| 194 | 11 | 3 | 8 | |
| 360 | 7 | 4 | 3 | |
| 540 | 41 | 26 | 15 | |
| High-income subjects | All (N = 137) | Cases (n = 58) | Controls (n = 79) | P |
| Total DFEs | 701.9 (405.3) | 766.0 (431.6) | 654.8 (380.8) | 0.12 |
| Food DFEs | 447.3 (260.8) | 437.7 (243.4) | 454.4 (274.2) | 0.708 |
| Fortified DFEs | 201.1 (199.8) | 190.7 (180.9) | 208.8 (213.4) | 0.593 |
| Natural DFEs | 246.2 (124.3) | 247.0 (137.7) | 245.6 (114.4) | 0.949 |
| Supplemental DFEs | ||||
| 0 | 73 | 24 | 49 | 0.02* |
| 194 | 15 | 6 | 9 | |
| 360 | 7 | 6 | 1 | |
| 540 | 42 | 22 | 20 | |
| Low-income subjects | All (N = 257) | Cases (n = 130) | Controls (n = 127) | P |
| Total DFEs | 600.4 (380.7) | 579.3 (354.1) | 622.0 (406.3) | 0.37 |
| Food DFEs | 498.7 (283.8) | 500.1 (285.5) | 497.4 (283.1) | 0.939 |
| Fortified DFEs | 204.7 (199.1) | 204.0 (200.5) | 205.5 (181.9) | 0.949 |
| Natural DFEs | 294.0 (165.9) | 296.1 (161.7) | 291.9 (170.8) | 0.837 |
| Supplemental DFEs | ||||
| 0 | 208 | 110 | 98 | 0.379 |
| 194 | 12 | 6 | 6 | |
| 360 | 7 | 2 | 5 | |
| 540 | 30 | 12 | 18 | |
| Hispanic low-income subjects | All (N = 149) | Cases (n = 74) | Controls (n = 75) | P |
| Total DFEs | 618.4 (381.0) | 570.0 (336.5) | 666.2 (417.0) | 0.123 |
| Food DFEs | 549.3 (309.7) | 543.7 (317.2) | 554.8 (304.1) | 0.8275 |
| Fortified DFEs | 229.7 (210.5) | 235.3 (228.6) | 224.4 (192.5) | 0.7572 |
| Natural DFEs | 319.6 (178.1) | 308.6 (168.2) | 330.4 (187.8) | 0.456 |
| Supplemental DFEs | ||||
| 0 | 129 | 69 | 60 | 0.038* |
| 194 | 6 | 3 | 3 | |
| 360 | 2 | 0 | 2 | |
| 540 | 12 | 2 | 10 | |
| Non-Hispanic White low-income subjects | All (N = 78) | Cases (n = 41) | Controls (n = 37) | P |
| Total DFEs | 589.9 (392.3) | 652.4 (404.6) | 520.7 (371.4) | 0.138 |
| Food DFEs | 424.3 (226.6) | 453.3 (236.1) | 392.1 (214.3) | 0.233 |
| Fortified DFEs | 181.9 (164.1) | 175.5 (154.5) | 189.1 (176.0) | 0.72 |
| Natural DFEs | 242.3 (125.3) | 277.8 (145.2) | 203.0 (84.5) | 0.006** |
| Supplemental DFEs | ||||
| 0 | 55 | 27 | 28 | 0.688 |
| 194 | 4 | 2 | 2 | |
| 360 | 4 | 2 | 2 | |
| 540 | 15 | 10 | 5 | |
Mean daily folate equivalent (DFE) levels assessed through the modified Block food frequency questionnaire are shown by folate source. Standard deviation is shown in parentheses. For Supplemental DFEs, subject counts for each intake level (0, 194, 360, 540) are shown. P values represent results of an two sample Student’s T test for continuous DFEs (Total, Food, Fortified, and Natural), and Fisher’s exact test for Supplemental DFEs.
* P < 0.05, ** P < 0.01.
Meet-in-the-middle analysis
Overlapping genomic coordinates from leukemic blast DMRs and each of the five sources of folate DMRs were analyzed to identify regions of DNA methylation significantly associated with both leukemic blasts and periconceptional folate intake (Fig. 2). Since periconceptional folate is inversely associated with ALL, we would predict that DNA methylation impacts of periconceptional folate at birth would be positively associated with the same DNA methylation marks in the tumor cell (and vice versa) – should DNA methylation be a means by which periconceptional folate may directly mediate ALL risk by influencing methylation in the opposite direction. A total of 17 (56.7%) DMRs overlapped for total DFEs, 13 (40.6%) for food DFEs, 10 (35.7%) for fortified DFEs, 13 (40.6%) for natural DFEs, and 18 (51.4%) for supplemental DFEs (Table 3). To identify common directionality of effect across these overlapping regions, coefficients from leukemic blast Set A and Set B and the folate regression analysis were averaged across probes contained within each DMR, resulting in a positive or negative directionality for each region. Instances in which directionality differed between Set A and Set B were subsequently omitted. Direction of effect was equivalent between leukemic blast and folate in 7 DMRs for total DFEs (binomial test P = 0.629), 6 for food DFEs (P = 1), 3 for fortified DFEs (P = 0.344), 4 for natural DFEs (P = 0.267), and 16 for supplemental DFEs (P = 0.001). Notably, for supplemental DFE, the significant concordance in direction of effect was positive for both the leukemic blast and folate analyses (Table 4). Gene set enrichment analysis of overlapping regions did not identify significant association with gene ontology or KEGG pathway terms (Supplementary Tables S3, 4).
Table 3.
Meet in the middle analysis overlapping regions.
| Concordant overlaps | Total overlaps | P | |
|---|---|---|---|
| All subjects | |||
| Total DFE | 7 | 17 | 0.6291 |
| Food DFE | 6 | 13 | 1 |
| Fortified DFE | 3 | 10 | 0.3438 |
| Natural DFE | 4 | 13 | 0.2668 |
| Supplemental DFE | 16 | 18 | 0.001312** |
| Hispanic subjects | |||
| Total DFE | 2 | 3 | 1 |
| Food DFE | 4 | 9 | 1 |
| Fortified DFE | 4 | 6 | 0.6875 |
| Natural DFE | 2 | 6 | 0.6875 |
| Supplemental DFE | 7 | 12 | 0.7744 |
| Non-Hispanic White subjects | |||
| Total DFE | 6 | 13 | 1 |
| Food DFE | 5 | 10 | 1 |
| Fortified DFE | 5 | 15 | 0.3018 |
| Natural DFE | 3 | 7 | 1 |
| Supplemental DFE | 4 | 11 | 0.5488 |
| Low-Income subjects | |||
| Total DFE | 6 | 13 | 1 |
| Food DFE | 3 | 19 | 0.004425** |
| Fortified DFE | 1 | 6 | 0.2188 |
| Natural DFE | 5 | 37 | 0.00000743*** |
| Supplemental DFE | 5 | 12 | 0.6072 |
| High-income subjects | |||
| Total DFE | 4 | 12 | 0.3877 |
| Food DFE | 8 | 11 | 0.2266 |
| Fortified DFE | 4 | 9 | 1 |
| Natural DFE | 1 | 1 | 1 |
| Supplemental DFE | 4 | 7 | 1 |
| Hispanic low-income subjects | |||
| Total DFE | 8 | 15 | 1 |
| Food DFE | 4 | 11 | 0.5488 |
| Fortified DFE | 3 | 4 | 0.625 |
| Natural DFE | 9 | 31 | 0.02945* |
| Supplemental DFE | 10 | 19 | 1 |
| Non-Hispanic White low-income subjects | |||
| Total DFE | 0 | 0 | – |
| Food DFE | 0 | 1 | 1 |
| Fortified DFE | 0 | 0 | – |
| Natural DFE | 0 | 2 | 0.5 |
| Supplemental DFE | 0 | 0 | – |
Total overlapping differentially methylated regions (DMRs) identified from the periconceptional folate epigenome-wide association study (EWAS) and acute lymphoblastic leukemia (ALL) EWAS are shown by dietary or supplemental source of folate. Concordant overlaps are those in which the mean direction of effect for the DMR (hypermethylated or hypomethylated) is consistent between the folate and ALL EWAS results. P-values for significant concordance or discordance in direction of effect obtained using a binomial test.
DFE dietary folate equivalent.
* P < 0.05, ** P < 0.01, *** P < 0.001.
Table 4.
Overlapping differentially methylated regions (DMRs) for all subjects.
| Chr | Gene name | Leukemic blast region | Folate region | Mean coefficient direction for region | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Start Pos | End Pos | Probes | Šidák P | Start Pos | End Pos | Probes | Šidák P | Folate DFE | Blast Set A | Blast Set B | ||
| Total DFEs | ||||||||||||
| 1 | LAX1 | 203733971 | 203734559 | 5 | 1.06E−10 | 203734167 | 203734559 | 7 | 1.04E−04 | − | − | − |
| 1 | SLC35F3 | 234366259 | 234367586 | 6 | 7.38E−08 | 234367322 | 234367586 | 4 | 0.04242 | − | + | + |
| 10 | GDF2 | 48416463 | 48417904 | 11 | 1.38E−06 | 48416839 | 48416977 | 6 | 0.002751 | − | + | + |
| 1 | LAX1 | 77417567 | 77417738 | 2 | 6.70E−04 | 77417567 | 77417738 | 2 | 0.01228 | − | + | + |
| 15 | UBE3A | 25683075 | 25685368 | 18 | 8.50E−30 | 25684818 | 25684849 | 3 | 0.0237 | + | − | − |
| 17 | C17orf98 | 36997420 | 36998108 | 9 | 3.97E−05 | 36997420 | 36997740 | 8 | 0.001372 | + | + | + |
| 19 | EVI5L | 7926080 | 7928773 | 8 | 1.59E−04 | 7928546 | 7928773 | 3 | 6.10E−04 | + | + | + |
| 19 | SHANK1 | 51170356 | 51172144 | 8 | 1.48E−06 | 51170137 | 51170356 | 3 | 0.01033 | + | + | + |
| 2 | MEIS1 | 66671478 | 66673985 | 10 | 7.41E−26 | 66672299 | 66672553 | 3 | 0.03284 | + | + | + |
| 3 | TUSC4;CYB561D2 | 50387146 | 50388910 | 18 | 2.44E−21 | 50388823 | 50388924 | 5 | 0.001263 | + | + | + |
| 6 | NA | 28583971 | 28584464 | 12 | 7.97E−14 | 28583971 | 28584121 | 9 | 0.02547 | − | + | + |
| 6 | RNF39 | 30038720 | 30039600 | 27 | 7.59E−13 | 30039132 | 30039600 | 14 | 1.16E−05 | − | + | + |
| 6 | PPT2;PRRT1 | 32115211 | 32123701 | 112 | 1.88E−27 | 32120584 | 32121475 | 27 | 4.54E−12 | + | + | + |
| 6 | PPP1R2P1 | 32847530 | 32847845 | 10 | 8.15E−04 | 32847530 | 32847845 | 11 | 0.03056 | − | + | + |
| 6 | CRISP2 | 49681178 | 49681391 | 7 | 1.51E−08 | 49681178 | 49681851 | 12 | 2.12E−07 | − | + | + |
| 7 | LOC100132707 | 154719422 | 154720452 | 4 | 4.07E−43 | 154720213 | 154720452 | 3 | 0.009585 | + | − | − |
| 9 | C9orf135 | 72435533 | 72436057 | 3 | 3.32E−07 | 72435785 | 72436057 | 3 | 3.87E−05 | − | + | + |
| Food DFEs | ||||||||||||
| 1 | LAX1 | 203733971 | 203734559 | 5 | 1.06E−10 | 203734167 | 203734559 | 7 | 7.11E−07 | − | − | − |
| 11 | NADSYN1 | 71209416 | 71210295 | 4 | 0.01948 | 71210210 | 71210295 | 2 | 0.00928 | − | − | − |
| 12 | KCNA6 | 4916913 | 4919230 | 11 | 2.81E−06 | 4918075 | 4919230 | 10 | 6.74E−14 | + | + | + |
| 1 | LAX1 | 1133074 | 1133378 | 5 | 5.28E−08 | 1133074 | 1133378 | 5 | 0.0123 | − | + | + |
| 2 | PLEKHH2;LOC728819 | 43903227 | 43903842 | 5 | 3.52E−04 | 43903470 | 43904011 | 6 | 6.66E−06 | − | + | + |
| 2 | LOC100287216;SH3RF3 | 109744523 | 109747264 | 17 | 1.01E−06 | 109746691 | 109747003 | 5 | 0.001069 | − | + | + |
| 3 | TUSC4;CYB561D2 | 50387146 | 50388910 | 18 | 2.44E−21 | 50388823 | 50388924 | 5 | 9.69E−08 | + | + | + |
| 4 | EIF4E | 99849011 | 99851211 | 19 | 3.99E−06 | 99850801 | 99851211 | 8 | 1.42E−05 | − | − | − |
| 4 | TACR3 | 104640197 | 104641548 | 11 | 3.86E−17 | 104640833 | 104641319 | 7 | 0.002247 | − | + | + |
| 6 | HLA-DQB2 | 32728862 | 32730299 | 21 | 4.93E−11 | 32729442 | 32729876 | 13 | 3.53E−09 | − | + | + |
| 6 | SLC17A5 | 74363018 | 74364700 | 10 | 1.32E−19 | 74364495 | 74364700 | 4 | 0.02705 | − | − | − |
| 7 | NA | 18125881 | 18127468 | 9 | 6.78E−06 | 18126902 | 18127112 | 4 | 0.02467 | − | + | + |
| 7 | LOC100132707 | 154719422 | 154720452 | 4 | 4.07E−43 | 154719422 | 154720452 | 5 | 2.47E−08 | + | − | − |
| Fortified DFEs | ||||||||||||
| 10 | ZNF33B | 43133935 | 43134302 | 4 | 1.84E−08 | 43133935 | 43134302 | 5 | 0.002341 | + | − | − |
| 10 | CYP2E1 | 135340445 | 135342218 | 12 | 0.004687 | 135340467 | 135340871 | 7 | 0.01295 | − | + | + |
| 11 | NADSYN1 | 71209416 | 71210295 | 4 | 0.01948 | 71210210 | 71210295 | 2 | 0.03506 | − | − | − |
| 10 | ZNF33B | 93614970 | 93617168 | 19 | 1.29E−17 | 93616894 | 93617168 | 12 | 5.84E−04 | + | − | − |
| 19 | SPIB | 50931222 | 50931875 | 7 | 6.51E−05 | 50931222 | 50931622 | 6 | 2.58E−05 | + | − | − |
| 2 | PLEKHH2;LOC728819 | 43903227 | 43903842 | 5 | 3.52E−04 | 43903470 | 43903842 | 5 | 0.003372 | − | + | + |
| 21 | NA | 45246252 | 45247038 | 5 | 1.48E−06 | 45246252 | 45246441 | 4 | 0.01049 | + | + | + |
| 6 | HLA-DQB2 | 32728862 | 32730299 | 21 | 4.93E−11 | 32729442 | 32729876 | 13 | 3.71E−07 | − | + | + |
| 6 | C6orf122;C6orf208 | 170190454 | 170191082 | 6 | 0.02188 | 170190914 | 170191082 | 4 | 0.01822 | + | + | + |
| 7 | HOXA4 | 27167828 | 27171401 | 25 | 2.23E−07 | 27170552 | 27171213 | 12 | 0.001559 | − | + | + |
| Natural DFEs | ||||||||||||
| 1 | ISG15 | 948740 | 948893 | 3 | 8.96E−35 | 948740 | 948893 | 5 | 9.35E−04 | + | − | − |
| 1 | LAX1 | 203733971 | 203734559 | 5 | 1.06E−10 | 203734167 | 203734559 | 7 | 8.60E−06 | − | − | − |
| 10 | GRID1 | 88022744 | 88023243 | 6 | 1.31E−07 | 88022859 | 88023185 | 4 | 0.01043 | − | + | + |
| 1 | ISG15 | 4916913 | 4919230 | 11 | 2.81E−06 | 4918075 | 4919230 | 10 | 6.74E−14 | + | + | + |
| 12 | TMPRSS12 | 51236582 | 51236768 | 6 | 5.42E−07 | 51236582 | 51236768 | 6 | 3.37E−05 | − | + | + |
| 12 | AGAP2;LOC100130776 | 58119543 | 58121004 | 8 | 1.48E−12 | 58119915 | 58120237 | 6 | 0.007224 | + | + | + |
| 2 | C2orf84 | 24396224 | 24398859 | 13 | 3.00E−09 | 24397539 | 24397871 | 7 | 0.01657 | − | + | + |
| 2 | LOC100287216;SH3RF3 | 109744523 | 109747264 | 17 | 1.01E−06 | 109746691 | 109747003 | 5 | 0.006027 | + | − | − |
| 22 | FAM118A | 45703899 | 45706726 | 18 | 5.42E−13 | 45704902 | 45705042 | 6 | 0.01679 | − | + | + |
| 4 | EIF4E | 99849011 | 99851211 | 19 | 3.99E−06 | 99850801 | 99851211 | 8 | 2.33E−04 | − | − | − |
| 5 | BHMT2;DMGDH | 78365255 | 78366076 | 7 | 3.20E−10 | 78365647 | 78365830 | 7 | 9.77E−04 | − | + | + |
| 6 | NA | 28583971 | 28584464 | 12 | 7.97E−14 | 28583971 | 28584155 | 11 | 0.001066 | − | + | + |
| 6 | PPP1R2P1 | 32847530 | 32847845 | 10 | 8.15E−04 | 32847530 | 32847845 | 11 | 0.009355 | − | + | + |
| Supplemental DFEs | ||||||||||||
| 10 | NA | 65732826 | 65733575 | 5 | 4.19E−04 | 65733273 | 65733575 | 4 | 0.02468 | + | + | + |
| 11 | CAT | 34460107 | 34461028 | 13 | 3.30E−22 | 34460107 | 34460557 | 11 | 3.66E−05 | + | − | − |
| 12 | PIWIL1 | 130821453 | 130824831 | 18 | 8.92E−22 | 130824015 | 130824529 | 5 | 2.25E−04 | + | + | + |
| 10 | NA | 104551553 | 104552397 | 7 | 2.64E−16 | 104551481 | 104551553 | 3 | 0.00429 | + | + | + |
| 15 | MIR548H4;NOX5;SPESP1 | 69221572 | 69223018 | 5 | 4.73E−07 | 69222895 | 69223018 | 5 | 9.18E−06 | + | + | + |
| 15 | ANKRD34C | 79574773 | 79576298 | 10 | 2.79E−10 | 79575334 | 79575634 | 2 | 0.02692 | + | + | + |
| 16 | ABAT | 8806359 | 8807043 | 12 | 9.74E−10 | 8806690 | 8807043 | 8 | 0.02621 | − | + | + |
| 18 | CABYR | 21718735 | 21719568 | 16 | 4.08E−06 | 21719352 | 21719568 | 6 | 1.76E−04 | + | + | + |
| 19 | SNAPC2 | 7982713 | 7986206 | 14 | 3.86E−08 | 7983877 | 7984171 | 5 | 0.01084 | + | + | + |
| 20 | GNAS | 57425515 | 57428473 | 58 | 1.79E−06 | 57427412 | 57427977 | 18 | 7.22E−07 | + | + | + |
| 3 | NA | 127633887 | 127634587 | 6 | 1.54E−08 | 127634188 | 127634587 | 7 | 0.003615 | + | + | + |
| 3 | NA | 133502564 | 133503437 | 6 | 5.23E−06 | 133502564 | 133502952 | 8 | 1.51E−04 | + | + | + |
| 4 | NA | 77341251 | 77342788 | 7 | 9.51E−09 | 77341841 | 77342104 | 4 | 0.02735 | + | + | + |
| 6 | NA | 28601271 | 28601519 | 10 | 4.66E−11 | 28601271 | 28601519 | 10 | 0.006544 | + | + | + |
| 7 | FOXK1 | 4762070 | 4763182 | 7 | 0.008631 | 4762236 | 4762371 | 2 | 9.11E−08 | − | − | − |
| 7 | ABCA13 | 48493792 | 48494734 | 7 | 3.34E−13 | 48494362 | 48494734 | 5 | 5.35E−05 | + | + | + |
| 8 | BAI1 | 143580770 | 143581070 | 3 | 0.04011 | 143580770 | 143581070 | 3 | 0.01311 | + | + | + |
| 8 | LYNX1 | 143858414 | 143859990 | 18 | 1.18E−09 | 143859410 | 143860090 | 10 | 1.15E−13 | + | + | + |
Overlapping differentially methylated regions (DMRs) identified using Comb-P on epigenome-wide association study (EWAS) for periconceptional folate and acute lymphoblastic leukemia (ALL) DMRs are shown by dietary or supplemental source of folate. Šidák P represents P-value after adjustment for multiple comparisons across the genome.
Chr chromosome, DFE dietary folate equivalent, Pos Position.
Subgroup analysis
Stratified assessment of Hispanic and non-Hispanic White participants demonstrates no significant directionality in overlapping regions (Supplementary Table S5, 6). In low annual income participants, there was a significant opposition in direction of effect for food DFEs (3 of 19 overlapping DMRs, P = 0.004) and natural DFEs (5 of 37 overlapping DMRs, P = 7.43 × 10−6, Table 5, Supplementary Table S7), while there was no significant directionality identified in high annual income participants (Supplementary Table S8). In Hispanic participants of low annual income, a significant opposition in direction of effect was again seen for natural DFEs (9 of 31 overlapping DMRs, P = 0.029, Table 5, Supplementary Table S9), while no significant directionality was seen in overlaps for non-Hispanic White participants of low annual income (Supplementary Table S10).
Table 5.
Overlapping regions between folate and leukemic blast analyses by Hispanic ethnicity and income status subsets with significantly discordant for direction of effect.
| Chr | Gene name | Leukemic blast region | Folate region | Mean coefficient for region | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Start Pos | End Pos | Probes | Šidák P | Start Pos | End Pos | Probes | Šidák P | Folate DFE | Blast Set A | Blast Set B | ||
| Low income subjects | ||||||||||||
| Food DFEs | ||||||||||||
| 1 | NA | 247802703 | 247803166 | 7 | 7.16E−08 | 247802703 | 247803033 | 7 | 0.01065 | − | + | + |
| 10 | PAOX | 135202522 | 135203200 | 5 | 3.38E−06 | 135202522 | 135203200 | 7 | 1.21E−08 | + | − | − |
| 1 | NA | 102702285 | 102702656 | 4 | 1.81E−10 | 102702414 | 102702656 | 3 | 2.20E−06 | − | + | + |
| 12 | TMPRSS12 | 51236582 | 51236768 | 6 | 5.42E−07 | 51236446 | 51236768 | 7 | 0.01819 | − | + | + |
| 13 | NA | 76444798 | 76445108 | 7 | 0.03747 | 76444798 | 76445108 | 8 | 0.03423 | − | + | + |
| 14 | DIO2 | 80676274 | 80677688 | 4 | 1.65E−04 | 80677608 | 80677847 | 4 | 2.34E−04 | − | + | + |
| 14 | MEG3 | 101290195 | 101294430 | 29 | 2.23E−07 | 101294430 | 101294623 | 2 | 0.01108 | − | + | + |
| 16 | ANKRD26P1 | 46603015 | 46604297 | 7 | 4.23E−12 | 46603015 | 46603321 | 6 | 0.03711 | − | + | + |
| 19 | C19orf6 | 1008643 | 1009949 | 4 | 0.001138 | 1009642 | 1009949 | 5 | 0.003159 | − | + | + |
| 2 | C2orf65 | 74875253 | 74876010 | 8 | 2.00E−15 | 74875253 | 74875561 | 8 | 0.008844 | − | + | + |
| 2 | LOC100287216;SH3RF3 | 109744523 | 109747264 | 17 | 1.01E−06 | 109746691 | 109747003 | 5 | 0.002443 | − | + | + |
| 6 | MAS1L | 29454623 | 29455256 | 7 | 0.008821 | 29454623 | 29454954 | 6 | 1.37E−05 | − | + | + |
| 6 | HLA-DQB2 | 32728862 | 32730299 | 21 | 4.93E−11 | 32729442 | 32729876 | 13 | 1.19E−04 | − | + | + |
| 6 | PPP1R2P1 | 32847530 | 32847845 | 10 | 8.15E−04 | 32847530 | 32847845 | 11 | 0.03313 | − | + | + |
| 7 | PON3 | 95025194 | 95027158 | 24 | 8.16E−15 | 95025855 | 95026211 | 13 | 6.53E−04 | − | + | + |
| 8 | SCARA5 | 27737078 | 27737223 | 2 | 3.01E−04 | 27737078 | 27737223 | 2 | 0.02614 | − | + | + |
| Natural DFEs | ||||||||||||
| 1 | GTF2B | 89356964 | 89358043 | 8 | 4.86E−21 | 89357908 | 89358043 | 5 | 0.02563 | − | + | + |
| 1 | NA | 117317838 | 117318232 | 5 | 0.01862 | 117317903 | 117318232 | 4 | 0.009964 | − | + | + |
| 1 | NA | 242220538 | 242220925 | 3 | 8.03E−04 | 242220538 | 242220925 | 3 | 4.56E−05 | − | + | + |
| 1 | GTF2B | 88022744 | 88023243 | 6 | 1.31E−07 | 88022744 | 88023243 | 6 | 4.86E−10 | − | + | + |
| 10 | SLC16A12 | 91295159 | 91296457 | 14 | 7.35E−16 | 91296252 | 91296457 | 3 | 0.03084 | − | + | + |
| 10 | PAOX | 135202522 | 135203200 | 5 | 3.38E−06 | 135202611 | 135203102 | 5 | 0.02039 | + | − | − |
| 13 | NA | 21900506 | 21901649 | 4 | 3.67E−05 | 21900392 | 21900810 | 4 | 0.00105 | − | + | + |
| 13 | TRPC4 | 38443634 | 38445267 | 12 | 5.54E−16 | 38445013 | 38445267 | 2 | 0.01753 | − | + | + |
| 14 | DIO2 | 80676274 | 80677688 | 4 | 1.65E−04 | 80677608 | 80677847 | 4 | 0.001305 | − | + | + |
| 19 | PPP1R13L | 45884900 | 45885922 | 3 | 3.66E−09 | 45885800 | 45886078 | 4 | 0.01325 | − | + | + |
| 2 | C2orf84 | 24396224 | 24398859 | 13 | 3.00E−09 | 24397539 | 24397871 | 8 | 1.58E−05 | − | + | + |
| 2 | GTF2A1L;STON1-GTF2A1L | 48844728 | 48845068 | 7 | 1.40E−07 | 48844728 | 48845068 | 7 | 0.002962 | − | + | + |
| 2 | BOLL | 198649771 | 198651576 | 16 | 5.23E−23 | 198649767 | 198650112 | 6 | 9.40E−04 | − | + | + |
| 2 | ANKMY1 | 241458886 | 241460664 | 8 | 1.27E−11 | 241458886 | 241459227 | 4 | 0.00544 | − | + | + |
| 20 | BLCAP;NNAT | 36147042 | 36150061 | 40 | 9.98E−06 | 36148375 | 36149271 | 31 | 7.56E−06 | − | + | + |
| 3 | PRR23A | 138725153 | 138725425 | 7 | 1.14E−14 | 138725153 | 138725425 | 8 | 4.33E−04 | − | + | + |
| 4 | RBM46 | 155702172 | 155703429 | 11 | 3.57E−06 | 155702760 | 155703429 | 5 | 1.92E−04 | − | + | + |
| 5 | NA | 131281008 | 131281574 | 6 | 0.004398 | 131281008 | 131281439 | 6 | 1.43E−04 | − | + | + |
| 6 | NA | 28583971 | 28584464 | 12 | 7.97E−14 | 28583971 | 28584155 | 11 | 3.33E−04 | − | + | + |
| 6 | NA | 28601271 | 28601519 | 10 | 4.66E−11 | 28601271 | 28601519 | 10 | 9.98E−04 | − | + | + |
| 6 | NA | 28828946 | 28832672 | 50 | 1.06E−14 | 28829171 | 28829640 | 10 | 2.86E−08 | − | + | + |
| 6 | TNXB | 32063394 | 32066121 | 42 | 4.45E−04 | 32063835 | 32064258 | 15 | 0.001266 | − | + | + |
| 6 | HLA-DQB2 | 32728862 | 32730299 | 21 | 4.93E−11 | 32729442 | 32729823 | 12 | 0.005971 | − | + | + |
| 6 | PPP1R2P1 | 32847530 | 32847845 | 10 | 8.15E−04 | 32847530 | 32847845 | 11 | 0.01474 | − | + | + |
| 6 | CRISP2 | 49681178 | 49681391 | 7 | 1.51E−08 | 49681178 | 49681851 | 12 | 2.68E−09 | − | + | + |
| 6 | SPACA1 | 88757302 | 88757878 | 6 | 5.18E−07 | 88757302 | 88757392 | 5 | 0.006156 | − | + | + |
| 6 | NA | 112688010 | 112688930 | 4 | 0.03221 | 112688010 | 112688607 | 3 | 0.002751 | − | + | + |
| 6 | C6orf174 | 127796287 | 127797286 | 7 | 1.71E−07 | 127796989 | 127797286 | 4 | 0.01374 | − | + | + |
| 7 | PON3 | 95025194 | 95027158 | 24 | 8.16E−15 | 95025836 | 95026248 | 15 | 1.75E−04 | − | + | + |
| 7 | REPIN1 | 150067807 | 150069890 | 7 | 2.33E−07 | 150069577 | 150069890 | 5 | 0.04007 | − | + | + |
| 7 | NA | 156400281 | 156400990 | 5 | 8.23E−05 | 156400711 | 156400990 | 5 | 0.01408 | − | + | + |
| X | FAM46D | 79590789 | 79591032 | 8 | 1.96E−08 | 79590789 | 79591032 | 8 | 2.42E−04 | − | + | + |
| Hispanic low income subjects | ||||||||||||
| Natural DFEs | ||||||||||||
| 1 | NAV1 | 201708522 | 201709675 | 9 | 1.16E−09 | 201708419 | 201708767 | 6 | 1.11E−04 | + | − | − |
| 1 | MFSD4 | 205560943 | 205561906 | 11 | 1.48E−23 | 205561090 | 205561401 | 6 | 0.00579 | + | − | − |
| 1 | NA | 242220538 | 242220925 | 3 | 8.03E−04 | 242220538 | 242220925 | 3 | 1.07E−04 | − | + | + |
| 1 | NAV1;NAV1 | 88022744 | 88023243 | 6 | 1.31E−07 | 88022859 | 88023243 | 5 | 1.70E−04 | − | + | + |
| 15 | LASS3 | 101084507 | 101085710 | 8 | 1.14E−16 | 101084980 | 101085388 | 8 | 7.31E−04 | − | + | + |
| 16 | UMOD;UMOD | 20359444 | 20361044 | 8 | 8.52E−06 | 20360206 | 20360497 | 4 | 0.001706 | − | + | + |
| 17 | WFIKKN2 | 48911141 | 48913424 | 21 | 7.52E−05 | 48912164 | 48912860 | 13 | 2.06E−06 | − | + | + |
| 19 | LYPD3;LYPD3 | 43969650 | 43970683 | 8 | 7.96E−04 | 43969650 | 43970048 | 7 | 5.29E−05 | − | + | + |
| 19 | LYPD5 | 44324467 | 44325008 | 11 | 1.02E−18 | 44324747 | 44325139 | 10 | 6.91E−04 | − | + | + |
| 19 | PPP1R13L | 45884900 | 45885922 | 3 | 3.66E−09 | 45885800 | 45886078 | 4 | 2.53E−05 | − | + | + |
| 2 | GTF2A1L;STON1-GTF2A1L | 48844728 | 48845068 | 7 | 1.40E−07 | 48844728 | 48845068 | 7 | 4.39E−06 | − | + | + |
| 20 | BLCAP;NNAT | 36147042 | 36150061 | 40 | 9.98E−06 | 36148375 | 36149231 | 30 | 4.08E−06 | − | + | + |
| 22 | FAM118A | 45703899 | 45706726 | 18 | 5.42E−13 | 45704675 | 45705265 | 8 | 2.87E−17 | − | + | + |
| 3 | MAGI1 | 65342216 | 65342971 | 5 | 5.21E−14 | 65342546 | 65342843 | 3 | 0.01418 | − | + | + |
| 3 | PRR23A | 138725153 | 138725425 | 7 | 1.14E−14 | 138725153 | 138725425 | 8 | 0.007163 | − | + | + |
| 4 | RBM46 | 155702172 | 155703429 | 11 | 3.57E−06 | 155702409 | 155703429 | 9 | 1.38E−07 | − | + | + |
| 4 | ODZ3 | 183721183 | 183721778 | 5 | 1.19E−04 | 183721565 | 183721778 | 4 | 0.02955 | − | + | + |
| 6 | NA | 28828946 | 28832672 | 50 | 1.06E−14 | 28829171 | 28829640 | 10 | 1.62E−05 | − | + | + |
| 6 | TNXB | 32063394 | 32066121 | 42 | 4.45E−04 | 32063595 | 32064956 | 37 | 2.86E−13 | − | + | + |
| 6 | NA | 90596146 | 90597591 | 9 | 5.81E−04 | 90597340 | 90597591 | 4 | 0.007835 | − | + | + |
| 7 | NA | 156400281 | 156400990 | 5 | 8.23E−05 | 156400711 | 156400990 | 5 | 0.02597 | − | + | + |
| 9 | NA | 90439478 | 90440066 | 5 | 6.88E−09 | 90439478 | 90439702 | 3 | 0.02084 | − | + | + |
Only regions discordant between folate and leukemic blast analyses are shown. Šidák P represents P-value after correction for multiple comparisons across the genome.
Chr chromosome, DFE dietary folate equivalent, Pos Position.
Discussion
These results identify overlapping regions of DNA methylation responsive to periconceptional folate at birth and within leukemic blasts at the time of pediatric ALL diagnosis. This includes 17 overlapping DMRs for total folate, 13 for food, 10 for fortified, 13 for natural, and 18 for supplemental folate intake. These findings indicate that sites of DNA methylation susceptible to periconceptional folate intake evident in a polyclonal birth blood cellular population are additionally associated with DNA methylation variation in leukemic cells proliferating from monoclonal precursors months to years later at the time of diagnosis at rates ranging from 35.7% to 56.7% by folate source. This implies the association between maternal folate intake and pediatric ALL risk may in part be reflected through alterations in DNA methylation identifiable at birth.
Interestingly, overlapping regions associated with dietary sources of folate (i.e. fortified and natural categories) showed a non-significant bias in opposing the direction of effect to that seen in leukemic blasts. This would indicate early variation in DNA methylation in response to dietary folate is driven in the opposite direction as that seen in leukemic blasts at diagnosis (i.e., hypomethylated at birth at sites found to be hypermethylated in leukemic blasts) and support a plausible protective mechanism for dietary folate. In addition, given the association between maternal dietary quality and reduction in pediatric ALL [5, 38], our findings related to dietary sources of folate may further reflect a biological mediation or manifestation of the beneficial effect of healthier diet on childhood leukemia risk.
These effects further varied by Hispanic ethnicity and income. While children with low annual income experienced opposing direction of effect for food and natural folate, there was no significant relationship with supplemental folate (i.e., vitamin pills). While there were no concordance or discordance in direction of effect in Hispanic or non-Hispanic White participants as a whole, Hispanic participants of low annual income presented a significant opposition in deflection of effect compared to natural folate. Here, while total folate intake was lower in Hispanic subjects, folate derived from food sources (both fortified and natural) was significantly higher in this group compared to non-Hispanic White subjects, who derived a larger proportion of total folate intake from supplements. Increased folate intake from natural sources may indicate an overall higher dietary quality, however this relationship is impacted by multiple cultural and socioeconomic factors within Hispanic subjects in California [39]. We previously demonstrated variation in the relationship between periconceptional folate and DNA methylation at birth by Hispanic ethnicity, annual income and education level [22] – specifically, these subgroups showed a higher magnitude of associations with supplemental folate. Our findings indicate that Hispanic children may benefit specifically from increased dietary folate in the pre/periconceptional period, and that this mechanism may in part be driven through early modifications in DNA methylation patterns. Notably, compared to all other racial/ethnic groups in the United States, Hispanic children have the highest risk of ALL [40], in addition to neural tube defects which are an additional folate-sensitive developmental morbidity [41].
In contrast to dietary folate, supplemental folate was significantly more likely to share a concordant direction of effect in overlapping regions to that of leukemic blasts. These results would imply folic acid supplementation drives DNA methylation toward the state seen in ALL at diagnosis. There are multiple potential explanations for the difference seen for dietary vs. supplemental folate. The bioavailability of naturally occurring folates in food is ~80% of that from folic acid containing supplements [42], presumably resulting in greater levels of folate available to the developing fetus. The availability of naturally occurring folates may be further reduced based on food handling and cooking practice. Thus, total folate exposure to the fetus from folic acid supplementation likely exceeds that of folate from food for the same measured intake from our survey data, which may impact the manner in which DNA methylation patterns are established. Similarly, the balance of other nutrients in food are likely to be highly varied in comparison to those of pre-natal and multi-vitamin supplements. Together, this would suggest potential variation in the mechanism by which folate impacts the early stages of ALL pathogenesis by folates found in food compared to folic acid from dietary supplements, and that excessively high doses of folic acid may contribute toward leukemia development. This finding would seem intriguing given Sydney Farber’s work demonstrating that folic acid accelerates the progression of acute leukemia [43], establishing anti-folates as the basis of early ALL treatment approaches [44]. However, epidemiological studies would contradict this premise, as folic acid supplementation shares a similar protective effect to that seen in dietary folate [15]. This relationship may further represent more complex gene-nutrient interactions in folate replete and depleted states [45], as higher supplemental folate levels generally coincided with higher total folate intake in our cohort. Variants in folate pathway genes leading to higher predicted serum folate levels are associated with reduced risk of ALL [14] – while dietary folate intake appears to mitigate this effect in Hispanic subjects, intake in Non-Hispanic White subjects did not alter their genetically determined risk of ALL. Notably, the mBFFQ assessment of supplements did not specifically distinguish between folic acid and 5-methyltetrahydrofolate containing products, however we anticipate relatively minimal contribution of the latter to the supplemental DFE variable since it was introduced to vital supplements after the majority of our study population was enrolled.
This study benefitted from a large sampling of participants from the diverse population in California with available perinatal DNA methylation data and detailed maternal periconceptional dietary and supplement calculations. We did not specifically control for genetic influence on DNA methylation patterns, or methylation quantitative trait loci (mQTL). However, none of the top overlapping regions (Tables 4 and 5) contained CpGs with known mQTL effect identified in prior investigations [46], reducing the likelihood of genetic influence confounding interpretation of our results. A limitation of this study is the reliance on retrospective mBFFQ survey in the assessment of folate intake, for which no formal assessment of reliability exists [5]. However, in this instance the risk of bias is reduced given knowledge of the relationship between folate and ALL risk is less widely known in the general population, which may serve to limit variation in recall between cases and controls [47]. Additionally, direct measurement of maternal folate may vary temporally [48] and may not be fully reflective of maternal stores accessible to the developing early fetus. In addition, our analysis of ALL leukemic blasts did not utilize deconvolution to control for the impact of normal bone marrow nucleated cells. However, the majority of cells in the diagnostic samples used will be made up of leukemic blasts, and the impact of any remaining non-leukemic mononucleated cells is unlikely to substantially impact our results, which is consistent with previous studies of similar design [37, 49, 50]. Our analysis did not specifically account for other types of vitamin intake which may influence ALL risk, and thus it is possible that some degree of the differences we identified between total, food and supplemental folate intake may represent more complex interactions between dietary quality and nutrient intake.
In summary, we identified linkage between regions of DNA methylation responsive to periconceptional folate and differentially methylated in lymphoblasts at the time of pediatric ALL diagnosis. These associations vary in number and epigenomic region depending on dietary source. The direction of effect in DNA methylation variation associated with folic acid from food and, in particular, naturally occurring folate in dietary sources had an inverse relationship to that of lymphoblasts for Hispanic participants of low annual income. These results suggest a unique benefit to increased dietary folate in Hispanic population as a means to reduce incidence of pediatric ALL. Given reported associations between maternal genetic variants in folate pathway genes [13] and ALL risk, these findings call for further investigation of the relationship between periconceptional folate, genetic variants and DNA methylation as it relates to pediatric ALL risk.
Supplementary information
Acknowledgements
The authors thank the families for their participation in the California Childhood Leukemia Study (formerly known as the Northern California Childhood Leukemia Study). The CCLS study was supported by National Institutes of Health grants R01ES009137, P42ES004705, R24ES028524, U24ES028524, P01ES018172, and P50ES018172, and from the United States Environmental Protection Agency under assistance agreements RD83451101 and RD83615901 for study recruitment, data collection and maintenance, genotyping, and/or DNA methylation work. Specific funding for this study was extended by the UK Children with Cancer grant #19-308. Our team extends appreciation to the Center for Advanced Research Computing at the University of Southern California for their invaluable computing resources (https://carc.usc.edu) and to the Southern California Environmental Health Sciences Center (P30ES007048) for core laboratory support, both significantly contributing to the research. Biospecimens and/or data used in this study were obtained from the California Biobank Program, (CBP requests #26 and #1531), Section 6555(b), 17 CCR. The California Department of Public Health is not responsible for the results or conclusions drawn by the authors of this publication. The authors thank Hong Quach and Diana Quach for support on DNA isolation and execution of the Illumina arrays. They thank Robin Cooley and Steve Graham of the California Department of Public Health for advice and logistical support. For recruitment of participants enrolled in the CCLS replication set, the authors gratefully acknowledge the clinical investigators at the following collaborating hospitals: University of California Davis Medical Center (Dr. Jonathan Ducore), University of California San Francisco (Drs. Mignon Loh and Katherine Matthay), Children’s Hospital of Central California (Dr. Vonda Crouse), Lucile Packard Children’s Hospital (Dr. Gary Dahl), Children’s Hospital Oakland (Dr. James Feusner), Kaiser Permanente Roseville (formerly Sacramento) (Drs. Kent Jolly and Vincent Kiley), Kaiser Permanente Santa Clara (Drs. Carolyn Russo, Alan Wong, and Denah Taggart), Kaiser Permanente San Francisco (Dr. Kenneth Leung), and Kaiser Permanente Oakland (Drs. Daniel Kronish and Stacy Month). Graphical abstract and Fig. 2 created in BioRender.
Author contributions
EMN was responsible for study design, conducting the analysis, interpretation of results and writing the manuscript. JLW was responsible for study design, interpretation of results and editing the manuscript. CM. AYK, LM organized procurement of study samples, collected data and prepared data for analysis, provided feedback on the report and edited the manuscript. All authors read and approved the final manuscript.
Funding
Open access funding provided by SCELC, Statewide California Electronic Library Consortium.
Data availability
All computer code generated for this analysis is available upon request from the authors. This study used biospecimens from the California Biobank Program, which prohibits uploading of genomic data (including genome-wide DNA methylation) and/or sharing of individual-level data obtained from these biospecimens under the statutory scheme of the California Health and Safety Code sections 124980(j), 124991(b) and (h), and 103850(a) and (d). Processed data files, including results from epigenome-wide association studies included in this manuscript are available as supplemental files.
Code availability
All computer code generated for this analysis is available upon request from the authors.
Competing interests
The authors declare no competing interests.
Ethics approval and consent to participate
This study was conducted in accordance with the Declaration of Helsinki. The institutional review boards of participating institutions and the State of California Health and Human Service Agency Committee for the Protection of Human Subjects (project number 2019-295) approved the study protocol. Written informed consent was obtained from a parent of participants, with assent from participants aged 7 years or older when appropriate.
Footnotes
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary information
The online version contains supplementary material available at 10.1038/s41375-026-02866-w.
References
- 1.de Smith AJ, Spector LG. In utero origins of acute leukemia in children. Biomedicines. 2024;12:236. [DOI] [PMC free article] [PubMed]
- 2.Bloom M, Maciaszek JL, Clark ME, Pui CH, Nichols KE. Recent advances in genetic predisposition to pediatric acute lymphoblastic leukemia. Expert Rev Hematol. 2020;13:55–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Wiemels JL, Walsh KM, de Smith AJ, Metayer C, Gonseth S, Hansen HM, et al. GWAS in childhood acute lymphoblastic leukemia reveals novel genetic associations at chromosomes 17q12 and 8q24.21. Nat Commun. 2018;9:286. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Abiri B, Kelishadi R, Sadeghi H, Azizi-Soleiman F. Effects of maternal diet during pregnancy on the risk of childhood acute lymphoblastic leukemia: a systematic review. Nutr Cancer. 2016;68:1065–72. [DOI] [PubMed] [Google Scholar]
- 5.Singer AW, Carmichael SL, Selvin S, Fu C, Block G, Metayer C. Maternal diet quality before pregnancy and risk of childhood leukaemia. Br J Nutr. 2016;116:1469–78. [DOI] [PubMed] [Google Scholar]
- 6.Bailey HD, Miller M, Langridge A, de Klerk NH, van Bockxmeer FM, Attia J, et al. Maternal dietary intake of folate and vitamins B6 and B12 during pregnancy and the risk of childhood acute lymphoblastic leukemia. Nutr Cancer. 2012;64:1122–30. [DOI] [PubMed] [Google Scholar]
- 7.Milne E, Royle JA, Miller M, Bower C, de Klerk NH, Bailey HD, et al. Maternal folate and other vitamin supplementation during pregnancy and risk of acute lymphoblastic leukemia in the offspring. Int J Cancer. 2010;126:2690–9. [DOI] [PubMed] [Google Scholar]
- 8.Thompson JR, Gerald PF, Willoughby ML, Armstrong BK. Maternal folate supplementation in pregnancy and protection against acute lymphoblastic leukaemia in childhood: a case-control study. Lancet. 2001;358:1935–40. [DOI] [PubMed] [Google Scholar]
- 9.Singer AW, Selvin S, Block G, Golden C, Carmichael SL, Metayer C. Maternal prenatal intake of one-carbon metabolism nutrients and risk of childhood leukemia. Cancer Causes Control. 2016;27:929–40. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Chokkalingam AP, Chun DS, Noonan EJ, Pfeiffer CM, Zhang M, Month SR, et al. Blood levels of folate at birth and risk of childhood leukemia. Cancer Epidemiol Biomarkers Prev. 2013;22:1088–94. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Metayer C, Spector LG, Scheurer ME, Jeon S, Scott RJ, Takagi M, et al. Folate metabolism and risk of childhood acute lymphoblastic leukemia: a genetic pathway analysis from the childhood cancer and leukemia International Consortium. Cancer Epidemiol Biomarkers Prev. 2024;33:1248–52. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Amigou A, Rudant J, Orsi L, Goujon-Bellec S, Leverger G, Baruchel A, et al. Folic acid supplementation, MTHFR and MTRR polymorphisms, and the risk of childhood leukemia: the ESCALE study (SFCE). Cancer Causes Control. 2012;23:1265–77. [DOI] [PubMed] [Google Scholar]
- 13.Lupo PJ, Dietz DJ, Kamdar KY, Scheurer ME. Gene-environment interactions and the risk of childhood acute lymphoblastic leukemia: exploring the role of maternal folate genes and folic Acid fortification. Pediatr Hematol Oncol. 2014;31:160–8. [DOI] [PubMed] [Google Scholar]
- 14.Xiang Y, Metayer C, Kogan SC, Ma X, Nickels E, Wiemels JL. The interrelationship between preconception folate nutritional intake and child genetic liability in the risk of childhood acute lymphoblastic leukemia. Cancer Epidemiol Biomarkers Prev. 2025;34:1415–24. [DOI] [PMC free article] [PubMed]
- 15.Metayer C, Milne E, Dockerty JD, Clavel J, Pombo-de-Oliveira MS, Wesseling C, et al. Maternal supplementation with folic acid and other vitamins and risk of leukemia in offspring: a Childhood Leukemia International Consortium study. Epidemiology. 2014;25:811–22. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Nordlund J, Syvänen AC. Epigenetics in pediatric acute lymphoblastic leukemia. Semin Cancer Biol. 2018;51:129–38. [DOI] [PubMed] [Google Scholar]
- 17.Nickels EM, Li S, Myint SS, Arroyo K, Feng Q, Siegmund KD, et al. DNA methylation at birth in monozygotic twins discordant for pediatric acute lymphoblastic leukemia. Nat Commun. 2022;13:6077. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Ghantous A, Nusslé SG, Nassar FJ, Spitz N, Novoloaca A, Krali O, et al. Epigenome-wide analysis across the development span of pediatric acute lymphoblastic leukemia: backtracking to birth. Mol Cancer. 2024;23:238. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Joubert BR, den Dekker HT, Felix JF, Bohlin J, Ligthart S, Beckett E, et al. Maternal plasma folate impacts differential DNA methylation in an epigenome-wide meta-analysis of newborns. Nat Commun. 2016;7:10577. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Gonseth S, Roy R, Houseman EA, de Smith AJ, Zhou M, Lee ST, et al. Periconceptional folate consumption is associated with neonatal DNA methylation modifications in neural crest regulatory and cancer development genes. Epigenetics. 2015;10:1166–76. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Timms JA, Relton CL, Rankin J, Strathdee G, McKay JA. DNA methylation as a potential mediator of environmental risks in the development of childhood acute lymphoblastic leukemia. Epigenomics. 2016;8:519–36. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Nickels EM, Li S, Morimoto L, Kang AY, de Smith AJ, Metayer C, et al. Periconceptional folate intake influences DNA methylation at birth based on dietary source in an analysis of pediatric acute lymphoblastic leukemia cases and controls. Am J Clin Nutr. 2022;116:1553–64. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Potter C, Moorman AV, Relton CL, Ford D, Mathers JC, Strathdee G, et al. Maternal red blood cell folate and infant Vitamin B 12 status influence methylation of genes associated with childhood acute lymphoblastic leukemia. Mol Nutr Food Res. 2018;62:e1800411. [DOI] [PubMed] [Google Scholar]
- 24.Ma X, Buffler PA, Layefsky M, Does MB, Reynolds P. Control selection strategies in case-control studies of childhood diseases. Am J Epidemiol. 2004;159:915–21. [DOI] [PubMed] [Google Scholar]
- 25.Lee ST, Muench MO, Fomin ME, Xiao J, Zhou M, de Smith A, et al. Epigenetic remodeling in B-cell acute lymphoblastic leukemia occurs in two tracks and employs embryonic stem cell-like signatures. Nucleic Acids Res. 2015;43:2590–602. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Lee ST, Xiao Y, Muench MO, Xiao J, Fomin ME, Wiencke JK, et al. A global DNA methylation and gene expression analysis of early human B-cell development reveals a demethylation signature and transcription factor network. Nucleic Acids Res. 2012;40:11339–51. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Zhou W, Triche TJ, Laird PW, Shen H. SeSAMe: reducing artifactual detection of DNA methylation by Infinium BeadChips in genomic deletions. Nucleic Acids Res. 2018;46:e123. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Aryee MJ, Jaffe AE, Corrada-Bravo H, Ladd-Acosta C, Feinberg AP, Hansen KD, et al. Minfi: a flexible and comprehensive Bioconductor package for the analysis of Infinium DNA methylation microarrays. Bioinformatics. 2014;30:1363–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Block G, Hartman AM, Naughton D. A reduced dietary questionnaire: development and validation. Epidemiology. 1990;1:58–64. [DOI] [PubMed] [Google Scholar]
- 30.Kwan ML, Jensen CD, Block G, Hudes ML, Chu LW, Buffler PA. Maternal diet and risk of childhood acute lymphoblastic leukemia. Public Health Rep. 2009;124:503–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Rahmani E, Zaitlen N, Baran Y, Eng C, Hu D, Galanter J, et al. Sparse PCA corrects for cell type heterogeneity in epigenome-wide association studies. Nat Methods. 2016;13:443–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Rahmani E, Shenhav L, Schweiger R, Yousefi P, Huen K, Eskenazi B, et al. Genome-wide methylation data mirror ancestry information. Epigenet Chromatin. 2017;10:1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Rahmani E, Yedidim R, Shenhav L, Schweiger R, Weissbrod O, Zaitlen N, et al. GLINT: a user-friendly toolset for the analysis of high-throughput DNA-methylation array data. Bioinformatics. 2017;33:1870–2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Willer CJ, Li Y, Abecasis GR. METAL: fast and efficient meta-analysis of genomewide association scans. Bioinformatics. 2010;26:2190–1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Pedersen BS, Schwartz DA, Yang IV, Kechris KJ. Comb-p: software for combining, analyzing, grouping and correcting spatially correlated P-values. Bioinformatics. 2012;28:2986–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Phipson B, Maksimovic J, Oshlack A. missMethyl: an R package for analyzing data from Illumina’s HumanMethylation450 platform. Bioinformatics. 2016;32:286–8. [DOI] [PubMed] [Google Scholar]
- 37.Nordlund J, Bäcklin CL, Wahlberg P, Busche S, Berglund EC, Eloranta ML, et al. Genome-wide signatures of differential DNA methylation in pediatric acute lymphoblastic leukemia. Genome Biol. 2013;14:r105. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Muñoz-Aguirre P, Denova-Gutiérrez E, Pérez-Saldivar ML, Espinoza-Hernández LE, Dorantes-Acosta EM, Torres-Nava JR, et al. Maternal dietary patterns and acute leukemia in infants: results from a case control study in Mexico. Front Nutr. 2023;10:1278255. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Harley K, Eskenazi B, Block G. The association of time in the US and diet during pregnancy in low-income women of Mexican descent. Paediatr Perinat Epidemiol. 2005;19:125–34. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Ward E, DeSantis C, Robbins A, Kohler B, Jemal A. Childhood and adolescent cancer statistics, 2014. CA Cancer J Clin. 2014;64:83–103. [DOI] [PubMed] [Google Scholar]
- 41.Canfield MA, Mai CT, Wang Y, O’Halloran A, Marengo LK, Olney RS, et al. The association between race/ethnicity and major birth defects in the United States, 1999-2007. Am J Public Health. 2014;104:e14–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Winkels RM, Brouwer IA, Siebelink E, Katan MB, Verhoef P. Bioavailability of food folates is 80% of that of folic acid. Am J Clin Nutr. 2007;85:465–73. [DOI] [PubMed] [Google Scholar]
- 43.Farber S, Cutler EC, Hawkins JW, Harrison JH, Peirce EC, Lenz GG. The action of pteroylglutamic conjugates on man. Science. 1947;106:619–21. [DOI] [PubMed] [Google Scholar]
- 44.Spain PD, Kadan-Lottick N. Observations of unprecedented remissions following novel treatment for acute leukemia in children in 1948. J R Soc Med. 2012;105:177–81. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Lucock M, Yates Z. Folic acid - vitamin and panacea or genetic time bomb? Nat Rev Genet. 2005;6:235–40. [DOI] [PubMed] [Google Scholar]
- 46.Li S, Mancuso N, Metayer C, Ma X, de Smith AJ, Wiemels JL. Incorporation of DNA methylation quantitative trait loci (mQTLs) in epigenome-wide association analysis: application to birthweight effects in neonatal whole blood. Clin Epigenet. 2022;14:158. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Infante-Rivard C. Jacques L. Empirical study of parental recall bias. Am J Epidemiol. 2000;152:480–6. [DOI] [PubMed] [Google Scholar]
- 48.Sobczyńska-Malefora A, Harrington DJ. Laboratory assessment of folate (vitamin B). J Clin Pathol. 2018;71:949–56. [DOI] [PubMed] [Google Scholar]
- 49.Nordlund J, Bäcklin CL, Zachariadis V, Cavelier L, Dahlberg J, Öfverholm I, et al. DNA methylation-based subtype prediction for pediatric acute lymphoblastic leukemia. Clin Epigenet. 2015;7:11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50.Nordlund J, Milani L, Lundmark A, Lönnerholm G, Syvänen AC. DNA methylation analysis of bone marrow cells at diagnosis of acute lymphoblastic leukemia and at remission. PLoS One. 2012;7:e34513. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
All computer code generated for this analysis is available upon request from the authors. This study used biospecimens from the California Biobank Program, which prohibits uploading of genomic data (including genome-wide DNA methylation) and/or sharing of individual-level data obtained from these biospecimens under the statutory scheme of the California Health and Safety Code sections 124980(j), 124991(b) and (h), and 103850(a) and (d). Processed data files, including results from epigenome-wide association studies included in this manuscript are available as supplemental files.
All computer code generated for this analysis is available upon request from the authors.