Table 2. Nutrition in early life and epigenetics.
*(Organised by, exposure, DNA methylation (epigenome wide, global methylation, imprinted genes, other genes).
| First author (year), country | Cohort, N (% female) | Early life variable(mean age ± SD (age range) | DNA methylation | Tissue | Mean age at epigenetic measure ± SD (age range) | Main result | Confounders |
|---|---|---|---|---|---|---|---|
| Maternal dietary intake / nutritional biomarker | |||||||
| Joubert (2016), Norway & The Netherlands [85] | MoBA: 1275 (NR) Generation R: 713 (NR) |
Plasma folate | Infinium Human Methylation450 BeadChip | Cord blood | Birth |
443 FDR-significant CpGs were differentially methylated in cord blood in relation to maternal folate. 48 CpGs met Bonferroni threshold (p<1.19x10-7). Selected loci from meta-analysis, Coef(SE), p: cg15908975(GRM8):-0.012(0.002),6.76x10-7 cg18574254(GRM8):-0.011(0.002),3.27x10-9 cg22591480(SLC16A12):-0.008(0.002),1.34x10-5 cg14920044(SLC16A12):-0.011(0.003), 4.31x10-6 cg24829292(OPCML):0.010(0.002),6.60x10-6 cg 22629528(OPCML):0.019(0.005),2.91x10-5 cg 26283170(OPCML):0.009(0.002), 1.30x10-5 cg 24804179(PRPH):-0.007(0.002), 8.05x10-6 cg 05775627(PRPH):-0.007(0.002),1.01x10-5 cg 16010628(PRPH):-0.005(0.001), 1.73x10-5 cg 05635274(PRSS21):0.009(0.002),4.77x10-6 cg 02296564(PRSS21):0.011(0.003),6.21x10-6 cg 22730830(PRSS21):0.013(0.003),3.99x10-6 cg 01232511(PRSS21):0.014(0.003), 1.23x10-5 cg 10612259(LHX1):-0.011(0.002), 9.10x10-8 cg 011965477(LHX1):-0.002(0.001), 2.09x10-5 cg 11775595(APC2):-0.015(0.003), 1.64x10-7 cg 14907738(APC2):-0.006(0.001), 8.57x10-6 cg 27150718(APC2):-0.009(0.002),5.81x10-7 cg 03165176(APC2):-0.012(0.003),1.44x10-5 cg 14559388(APC2):-0.003(0.001),4.98x10-6 cg 04624885(APC2):-0.010(0.002), 1.56x10-5 cg 19870717(APC2):-0.009(0.002),4.64x10-9 cg 16613938(APC2):-0.016(0.003),3.05x10-8 cg 23291200(APC2):-0.010(0.002),1.72x10-9 cg 13793157(KLK4):-0.009(0.002), 4.00x10-5 cg10078829(KLK4):-0.007(0.002), 1.84x10-5 |
Maternal age, education, smoking during pregnancy, parity, batch effects |
| Boeke (2012), US [86] | Project Viva, Periconceptional intake: 516 Second trimester intake: 484 (47.7) |
FFQ for B-vitamins (32 ± 5.1y) | LINE-1 using pyrosequencing | Cord blood | Birth |
0–4 weeks gestation, β = %5MC difference in LINE-1 methylation for increment in 1 SD in nutrient Methyl donor(Cumulative Index): β = -0.02 (-0.04, 0.01), p = 0.17 maternal vitamin B12 (μg/d): β = 0.01 (-0.06,0.08), p = 0.70 maternal betaine (mg/d): β = -0.04 (-0.11,0.03), p = 0.24 maternal choline (mg/d): β = -0.02 (-0.08,0.04), p = 0.45 maternal folate (μg/d)e: β = -0.03 (-0.10,0.03), p = 0.32 Second trimester, β = %5MC difference in LINE-1 methylation for increment in 1 SD for nutrients maternal vitamin B12: β = -0.02 (-0.09,0.06), p = 0.64 maternal betaine: β = -0.02 (-0.10,0.05), p = 0.50 maternal choline:β = -0.004 (-0.07,0.06), p = 0.98 maternal folate: β = -0.02 (-0.08,0.05), p = 0.61 |
Other methyl donors, child's sex, mother's age, race, smoking, pregnancy, weight gain, education, cadmium intake |
| Pauwels (2017), Belgium [87] | MANOE, 115(47.8) | FFQ for methyl donor intake & folic acid supplementation (31±3.6y) | Global DNA methylation using mass-spectrometry method & DNMT1, LEP, RXRA, IGF2 DMR using PCR | Cord blood | Birth (GAD 39.6±0.9w) |
Before pregnancy (n = 24) β(95%CI), p: LEP: Betaine: -0.13(-3.45, 3.19), p = 0.94 Choline: 1.48(-1.48, 4,45), p = 0.31 Folate: -0.33(-2.75, 2.09), p = 0.78 Methionine: 0.427 (0.01, 0.85), p = 0.048 DNMT1: Betaine:0.675(0.04, 1,31), p = 0.039 Choline:0.13(-0.52,0.78), p = 0.68 Folate:0.21(-0.3, 0.72), p = 0.40 Methionine: 0.04(-0.06, 0.14), p = 0.37 Second trimester (n = 89) β(95%CI), p: LEP: Betaine:-0.575(-1.16, 0.01), p = 0.05 Choline:-0.47(-0.95, 0.02), p = 0.058 Folate:-0.507(-0.89, -0.13), p = 0.009 Methionine: -0.06(-0.14, 0.02), p = 0.15 DNMT1: Betaine:-0.25(-0.58, 0.09), p = 0.15 Choline: -0.301(-0.57, -0.03), p = 0.03 Folate:-0.226(-0.45, -0.01), p = 0.045 Methionine: -0.04(-0.08, 0.009), p = 0.12 Third trimester (n = 89) β(95%CI), p: RXRA: Betaine: 0.35(-1.24, 1.94), p = 0.66 Choline:-0.935(-2.08, 0.21), p = 0.11 Folate:-1.001(-1.96, -0.04), p = 0.041 Methionine: -0.15(-0.35, 0.06), p = 0.16 DNMT1: Betaine:0.97(0.36, 3,67), p = 0.96 Choline:0.291(0.1. 0.84), p = 0.022 Folate:0.48(0.22, 1.06), p = 0.07 Methionine: 0.87 (0.74, 1.04), p = 0.12 Folic acid supplementation LEP CpG1 methylation > 6 months before conception vs. 3–6 months before conception: 34.6 ± 6.3% vs. 30.1 ± 3.6%, p = 0.011 LEP CpG3 methylation > 6 months before conception vs no supplement before conception: 16.2 ± 4.4% vs. 13.9 ± 3%, p = 0.036 RXRA mean methylation supplements during entire pregnancy vs. stopping in second trimester: 12.3 ± 1.9% vs. 11.1 ± 2%, p = 0.008 |
Maternal age, maternal BMI, maternal smoking before and during each trimester of pregnancy, gestational weight gain |
| Fryer (2009), UK [88] | 24 (58.3) | Folic acid supplementation during pregnancy (29.4±7y) | LINE-1 methylation using pyrosequencing | Cord blood | Birth |
Correlation with LINE-1 methylation: Maternal folic acid intake: β = 0.31, p = 0.15 Prescribed folic acid dose during pregnancy: β = 0.36, p = 0.31 |
Sex, GA maternal age, parity, and BMI and cord serum folate, plasma homocysteine |
| Haggarty (2013), UK [81] | 913 (46) | FFQ for folate intake, folic acid supplementation, RBC folate (30.5 (95%CI: 30.2–30.9y) | IGF2 (4 CpGs), PEG3 (7 CpGs), SNRPN (15q11, 4 CpGs) LINE-1 (4 CpGs) using pyrosequencing | Cord blood | Birth (GAD 3.95 (95%CI: 39.4, 39.6w)) |
LINE-1 methylation: maternal folate intake (100ug/d):β = 0.002 (-0.20,0.20), p = 0.98 maternal folate supplement use, yes/no (periconceptional):β = 0.05 (-0.25,0.35), p = 0.74 maternal folate supplement use, yes/no (first 12 weeks gestation): β = 0.16 (-0.23,0.55), p = 0.42 maternal folate supplement use, yes/no (after 12 weeks): β = -0.34 (-0.64,-0.04), p = 0.03 maternal RBC folate, 100 nmol/L:β = -0.13 (-0.20,-0.05), p = 0.001 PEG-3 methylation: maternal folate intake (100ug/d):β = 0.002 (-0.20,0.2), p = 0.44 maternal folate supplement use, yes/no (periconceptional):β = -0.02 (-0.40,0.37), p = 0.94 maternal folate supplement use, yes/no (first 12 weeks gestation): β = -0.47 (-0.86,-0.08), p = 0.02 maternal folate supplement use, yes/no (after 12 weeks): maternal RBC folate, 100 nmol/L: β = -0.02 (-0.10,0.06), p = 0.60 SNRPN methylation: maternal folate intake (100ug/d): β = 0.07 (-0.33,0.46), p = 0.74 maternal folate supplement use, yes/no (periconceptional): β = -0.22 (-0.36,0.81), p = 0.46 maternal folate supplement use, yes/no (first 12 weeks gestation): β = 0.39 (-0.37,1.15), p = 0.32 maternal folate supplement use, yes/no (after 12 weeks): β = -0.01 (-0.60,0.58), p = 0.97 maternal RBC folate, 100 nmol/L:β = 0.02 (-0.12,0.15), p = 0.82 IGF2 methylation: maternal folate intake (100ug/d): β = 0.23 (-0.21,0.67), p = 0.32 maternal folate supplement use, yes/no (periconceptional): β = 0.31 (-0.35,0.96), p = 0.36 maternal folate supplement use, yes/no (first 12 weeks gestation): β = -0.10 (-0.95,0.76), p = 0.83 maternal folate supplement use, yes/no (after 12 weeks): β = 0.68 (0.02,1.35), p = 0.04 maternal RBC folate, 100 nmol/L:β = 0.10 (-0.05,0.24), p = 0.18 |
|
| McKay (2012), UK [89] | The North Cumbria Community Genetics Project, Infant: 294 (48) Maternal: 121 |
Serum B12 (median 28.6y) | Global DNA methylation using LUMA & IGF2, IGFBP3, ZNT5 using pyrosequencing | Cord blood | Birth | Global DNA methylation correlated inversely with maternal vitamin B12 concentrations: β = 0.0002(0.0001), p = 0.06. After adjustment: serum B12:β = 0.00007 (0.00007), p = 0.29 |
Sex, GA, infant MTHFR genotype |
| Hoyo (2011), US [90] | NEST 428 (50) | Folic acid supplement before (n = 428) and during pregnancy (n = 223) (29 ± 6.2y) | IGF2 & H19 DMR using pyrosequencing | Cord blood | Birth |
Methylation % difference for folic acid supplement before pregnancy: IGF2 methylation: Moderate vs. non-users: 0.28, p = 0.76 High (i.e. prescribed & over the counter) vs. non-users: -1.15, p = 0.39 H19 methylation: Moderate vs. non-users:-1.96, p = 0.03 High vs. non-users: -2.76, p = 0.04 Methylation % difference for folic acid supplement during pregnancy: IGF2 methylation: Moderate vs. non-users:0.75, p = 0.59 High vs. non-users: 0.25, p = 0.93 H19 methylation: Moderate vs. non-users:-2.87, p = 0.02 High vs. non-users: -4.90, p = 0.05 |
Maternal education, race, mode of delivery, cigarette smoking, sex |
| Steegers-Theunissen (2009), The Netherlands [62] | HAVEN study controls 120(~58) | Folic acid supplementation during pregnancy 400 μg/day vs. no supplement | IGF2 (5 CPGs) using mass-spectrometry based method | Blood | 17 months |
Mean (SE) of IGF2 methylation in childhood without maternal exposure to folic acid n = 34 vs. exposed n = 86: 0.474(0.007) vs. 0.495(0.004), p = 0.014 Adjusted analysis: mean difference in IGF2 methylation 4.5% (1.8) with maternal exposure to folic acid vs unexposed, p = 0.014 |
Maternal education |
| Loke (2013), Australia [91] | PETS 95 twin pairs (55 MZ & 40 DZ) (~50%) | Folate and macronutrient intake | IGF2 and H19 DMRs using mass-spectrometry based method | HUVECs, (CBMCs and granulocytes); ectoderm (buccal epithelium) and extra embryonic ectoderm (placenta) | Birth (GAD median 37.0±1.94w) |
Difference (p) in absolute percentage methylation in all tissues combined All Assays combined Had folate: 0.50(0.44) Vitamin B12 (z-score): -0.23(0.24) Homocysteine(z-score): 0.27(0.29) Macronutrients (z-score): 0.37(0.17) H19 promoter DMR Had folate: -1.70(0.024) Vitamin B12 (z-score): -0.97(0.002) Homocysteine(z-score): 0.10(0.75) Macronutrients (z-score): 0.80(0.049) IGF2/H19 ICR Had folate: 0.40(0.69) Vitamin B12 (z-score): -0.23(0.54 Homocysteine(z-score): 0.40(0.29) Macronutrients (z-score): 0.20(0.050) IGF2 DMR0 Had folate: 0.90(0.46) Vitamin B12 (z-score):0.23(0.55) Homocysteine(z-score): 0.37(0.30) Macronutrients (z-score): 0.43(0.27) IGF2 DMR2 Had folate: 2.90(0.035) Vitamin B12 (z-score):0.27(0.63) Homocysteine(z-score): 0.17(0.72) Macronutrients (z-score): 0.10(0.77) Differences in coefficients between cell types Had folate: HUVECs vs buccal -4.5%; p = 0.026; Vitamin B12 z-score: Granulocytes vs buccal (2.1%; p = 0.004). No other difference found |
|
| Azzi (2014), France [43] | EDEN 254(NR) | FFQ for B-vitamins & supplementation (during pregnancy (29.8±4.4y)) | ZAC1 DMR using methylation-specific PCR | Cord blood | Birth (GA at birth 39.5±1.5) |
Spearman’s rank partial correlation coefficients Prior to pregnancy: Vitamin B2: 0.14 p = 0.04 Vitamin B3: 0.04, p = 0.60 Vitamin B6: 0.04, p = 0.49 Vitamin B9: 0.02, p = 0.74 Vitamin B12: 0.11, p = 0.08 Last 3 months of pregnancy: Vitamin B2: 0.11 p = 0.09 Vitamin B3: 0.08, p = 0.22 Vitamin B6: 0.04, p = 0.5 Vitamin B9: 0.04, p = 0.56 Vitamin B12: 0.02, p = 0.79 No association with folic acid supplementation and/or the use of a combination of micronutrients either prior to or during pregnancy (estimates not provided) |
|
| Obermann-Bors (2013), The Netherlands [64] | 120 (50) | Folic acid supplementation | LEP using mass-spectrometry based method | Blood | 17± 2.5m |
Variable, % absolute methylation change (SE), p No folic acid: 0.1(0.8) p = 0.91 |
Batch, correlation between 7 CpGs, |
| Adkins (2010), NR**[92] | 30 (NR) | Biomarkers on one carbon pathway | ~15,000 loci (Details not specified) | NR | Birth | Phosphatidyl choline was significantly correlated with newborn DNA methylation at a subset of loci | |
| Ba (2011), China [93] | 99 (48) | B-vitamin biomarker (27.8 ±5.3y) | IGF2 promoters using methylation-specific PCR | Cord blood | Birth (96% GAD 37-41w) |
Promoter P2: Mean change per SD of each characteristic (p): Maternal blood serum folate: 0.05 (0.47) Maternal blood serum vitamin B12: 0.09 (0.19) Promoter P3: Mean change per SD of each characteristic (p): Maternal blood serum folate: 0.049 (0.47) Maternal blood serum vitamin B12: -0.22 (0.001) |
Mother’s age, maternal prepregnancy BMI, weight gain during pregnancy, mother’s highest education level, parity, supplementation intake during pregnancy, birth weight and birth length, sex, and GA |
| Hoyo (2014), US [35] | NEST 496 (49.7) | Erythrocyte folate (first trimester) | IGF2, H19, DLK1, MEG3, PEG3, MEST, PEG10, SGCE, NNAT using pyrosequencing | Cord blood | Birth |
Erythrocyte folate quartiles β(SE): MEG3 methylation:β = -2.02 (0.58), p = 0.001 for Q4 vs Q1 NNAT methylation:β = -1.34 (0.73, p = 0.07 for Q3 vs Q1 PEG10/SEGCE methylation:β = -0.14 (0.33), p = 0.66 for Q4 vs Q1 MEG3-IG methylation:β = -0.68 (0.61), p = 0.27 for Q4 vs Q1 PLAG1 methylation:β = -1.01 (0.40), p = 0.01 for Q3 vs Q1 PEG3 methylation:β = 0.43 (0.22), p = 0.03 for Q2 vs Q1 PEG3/MEST methylation:β = 0.39 (0.44), p = 0.37 for Q4 vs Q1 H19 methylation:β = 0.09 (0.33), p = 0.78 for Q4 vs Q1 IGR2 methylation:β = -0.04 (0.43), p = 0.004 for Q2 vs Q1 |
Maternal race, sex, cigarette smoking, GAD, GA at blood draw, physical activity, pre-pregnancy BMI, and delivery route |
| McCullough (2016), US [94] | NEST 429 (50) | B-vitamin biomarkers (56% between 20-29y) | H19 MEG3 SGCE/PEG10 PLAGL1 DMR using pyrosequencing | Cord blood | Birth |
H19 methylation β(SE) serum B12:β = -0.41 (0.57), p = 0.48 for Q4 vs Q1 serum pyridoxal phosphate: β = -0.07 (0.63), p = 0.91 for Q4 vs Q1 serum 4-pyridoxic acid:β = -0.57 (0.61), p = 0.35 for Q4 vs Q1 serum homocysteine: β = 1.01 (0.59), p = 0.09 for Q2 vs Q1 MEG3 methylation β(SE) serum B12:β = -0.93 (0.85), p = 0.27 for Q4 vs Q1 serum pyridoxal phosphate: β = 3.24 (0.89), p<0.01 for Q4 vs Q1 serum 4-pyridoxic acid:β = 1.62 (0.87), p = 0.06 for Q4 vs Q1 serum homocysteine: β = 1.60 (0.87), p = 0.07 for Q4 vs Q1 SGCE/PEG10 methylation β(SE) serum B12:β = 0.47 (0.67), p = 0.48 for Q4 vs Q1 serum pyridoxal phosphate: β = -0.30 (0.81), p = 0.71 for Q4 vs Q1 serum 4-pyridoxic acid:β = 1.46 (0.74), p = 0.05 for Q2 vs Q1 serum homocysteine: β = 1.43 (0.77), p = 0.06 for Q2 vs Q1 PLAG1 methylation β(SE): serum B12:β = 1.79 (0.96), p = 0.06 for Q4 vs Q1 serum pyridoxal phosphate:β = -0.11 (1.04), p = 0.91 for Q4 vs Q1 serum 4-pyridoxic acid:β = -0.15 (0.99), p = 0.88 for Q4 vs Q1 serum homocysteine: β = 1.77 (0.97), p = 0.07 for Q3 vs Q1 |
GAD, GA at blood draw, maternal race/ethnicity, maternal smoking and pre-pregnancy body mass index |
| Dominguez-Salas (2014), The Gambia [95] | Keneba Cohort 126 (43) | One-carbon metabolism biomarkers (18-45y) | Metastable epialleles: BOLA3, LOC654433, EXD3, ZFVE28 using methylation-specific amplification microarray and pyrosequencing. RBM46, PARD6G, ZNF678 using pyrosequencing | Blood lymphocytes (n = 126), Hair follicle (n = 87) |
3.6 ±0.9m |
Effect sizes are 1)standardised β coefficient for change in mean DNA methylation (combined MEs) per 1 SD of the predictor and 2) odds ratio per change in predictor: Peripheral blood lymphocyte: serum folate nmol/ l: β = 0.02(-0.07,0.12), OR = 1.03 (0.90,1.17), p = 0.62 serum vitamin B2 1/EGRAC:β = 0.09 (0.00,0.19), OR = 1.19 (0.98,1.46), p = 0.05 serum vitamin B12 pmol/l: β = 0.03 (-0.07,0.14), OR = 1.04 (0.91,1.19), p = 0.54 serum active vitamin B12 pmol/l: β = -0.04 (-0.16,0.07), OR = 0.98 (0.87–1.11), p = 0.45 serum choline umol/l: β = -0.01 (-0.12,0.09), OR = 0.95 (0.80–1.12), p = 0.80 serum betaine umol/l: β = 0.05 (-0.10,0.20), OR = 1.03 (0.89–1.19), p = 0.49 serum dimethyl glycine umol/l:β = -0.06 (-0.16,0.04), OR = 0.95 (0.86,1.04), p = 0.21 serum betaine/dimethyl glycine:β = 0.08 (-0.02,0.17), OR = 1.05 (0.97,1.14), p = 0.11 serum S-adenosylmethionine nmol/l: β = -0.06 (-0.17,0.05), OR = 0.79 (0.58,1.08), p = 0.28 serum S-adenosylhomocysteine nmol/l: β = -0.09 (-0.18,0.01), OR = 0.88 (0.75,1.02), p = 0.07 maternal serum S-adenosylmethionine/S-adenosylhomocysteine: β = 0.06 (-0.03,0.15), OR = 1.08 (0.92,1.27), p = 0.18 serum methionine umol/l: β = 0.07 (-0.03,0.18), OR = 1.19 (0.90,1.56), p = 0.18 serum homocysteine umol/l: β = -0.14 (-0.23,-0.05), OR = 0.80 (0.68,0.93), p = 0.003 maternal serum vitamin B6 nmol/l: β = -0.16 (-0.27,-0.04), OR = 0.82 (0.71,0.94), p = 0.005 serum cysteine umol/l: β = -0.19 (-0.31,-0.07), OR = 0.45 (0.30,0.68), p = 0.002 Hair follicle: serum folate nmol/l: β = 0.01 (-0.11,0.13), OR = 1.00 (0.86,1.16), p = 0.81 serum vitamin B2 1/EGRAC:β = 0.11 (0.00,0.22), OR = 1.22 (0.97,1.53), p = 0.04 serum vitamin B12 pmol/l: β = 0.08 (-0.06,0.23), OR = 1.06 (0.88,1.26), p = 0.25 serum active vitamin B12 pmol/l:β = -0.03 (-0.18,0.13), OR = 1.00 (0.85,1.18), p = 0.75 serum choline umol/l: β = 0.01 (-0.13,0.14), OR = 0.96 (0.77,1.19), p = 0.91 serum betaine umol/l: β = 0.13 (-0.07,0.32), OR = 1.06 (0.88,1.28), p = 0.19 serum dimethyl glycine umol/l:β = -0.02 (-0.15,0.11), OR = 0.97 (0.86,1.09), p = 0.79 serum betaine/dimethyl glycine:β = 0.06 (-0.06,0.18), OR = 1.04 (0.94,1.15), p = 0.34 serum S-adenosylmethionine nmol/l: β = -0.05 (-0.19,0.09), OR = 0.85 (0.57,1.27), p = 0.48 serum S-adenosylhomocysteine nmol/l:β = -0.12 (-0.25,0.01), OR = 0.84 (0.69,1.03), p = 0.06 serum S-adenosylmethionine/S-adenosylhomocysteine: β = 0.09 (-0.03,0.22), OR = 1.15 (0.93,1.41), p = 0.13 serum methionine umol/l: β = 0.00 (-0.13,0.14), OR = 0.99 (0.70,1.38), p = 0.96 serum homocysteine umol/l: β = -0.15 (-0.27,-0.03), OR = 0.82 (0.67,1.00), p = 0.02 maternal serum vitamin B6 nmol/l: β = -0.12 (-0.26,0.02), OR = 0.86 (0.73,1.02), p = 0.08 serum cysteine umol/l: β = -0.20 (-0.36,-0.04), OR = 0.43 (0.25,0.72), p = 0.01 |
|
| Rerkasem (2015), Thailand [59] | 249(NR) | 24-hour food recall & FFQ in each trimester | LINE-1 and Alu using COBRA | Blood | 20y |
% Total methylation, r, p(FDR) Maternal Protein intake 1st trim: Alu: 0.18, p = 0.46 LINE-1: -0.11, p = 0.75 Maternal Protein intake 2ndt trim: Alu: -0.08, p = 0.61 LINE-1: 0.08, p = 0.61 Maternal Protein intake 3rd trim: Alu: 0.04, p = 0.78 LINE-1: 0.06, p = 0.78 Maternal CHO intake 1st trim: Alu: 0.05, p = 0.81 LINE-1: -0.05, p = 0.82 Maternal CHO intake 2nd trim: Alu: 0.01, p = 0.87 LINE-1: -0.05, p = 0.88 Maternal CHO intake 3rd trim: Alu: 0.07, p = 0.74 LINE-1:0.06, p = 0.73 Maternal fat intake 1st trim: Alu: -0.11, p = 0.64 LINE-1: -0.22, p = 0.46 Maternal fat intake 2nd trim: Alu: -0.09, p = 0.87 LINE-1: -0.007, p = 0.98 Maternal fat intake 3rd trim: Alu: -0.17, p = 0.09 LINE-1:0.006, p = 0.96 Maternal energy intake 1st trim: Alu: 0.03, p = 0.82 LINE-1: -0.11, p = 0.54 Maternal energy intake 2nd trim: Alu: -0.02, p = 0.92 LINE-1: -0.03, p = 0.91 Maternal energy intake 3rd trim: Alu: -0.008, p = 0.92 LINE-1:0.05, p = 0.88 |
|
| Drake (2012), UK [61] | The Motherwell Cohort, 34(64) | FFQ (early ≤20w & late pregnancy >20w) | HSD2 (promotor region), exon 1(C) and 1(F) of GR (exon 1(C) and 1(F)), IGF2 DMRs using pyrosequencing | Blood | 40 (0.12y) |
Correlation of mean GR exon 1F methylation during late pregnancy Meat/w: r = 0.48, p = 0.009 Fish/w: r = 0.38, p = 0.048 Veg/w: r = 0.67, p<0.001 Bread/w: r = -0.49, p = 0.009 Potato/w: r = -0.39, p = 0.04 Methylation was increased at a specific CpG sites in HSD2 with increased meat (r = 0·42, p = 0·03) and fish r = 0·40, p = 0·04) intake in late pregnancy. Other results not presented |
Sex, BMI, birth weight. |
| Godfrey (2011), UK [96] | PAH 78 (NR) | FFQ (GA 15w) | eNOS, SOD1, IL8, P13KCD, RXRA using pyrosequencing | Cord blood | Birth | Higher methylation of RXRA but not of eNOS was associated with lower maternal CHO intake. Maternal fat and protein intake were not associated with RXRA methylation. No estimates for other nutrients/genes | |
| Simpkin (2015), UK [58] | AIRES 1018 (51) | Serum selenium & vitamin D | Infinium Human Methylation450 BeadChip to estimate Horvath epigenetic age | Cord blood & blood | Birth, 7.5y, 17.1y |
Correlations between early life variable and age acceleration: Maternal selenium & AA at birth:-0.103, p = 0.06 Maternal selenium & AA 7 years: -0.137, p = 0.009 Maternal selenium & AA at 17 years: 0.01, p = 0.84 Maternal vitamin D & AA at birth:-0.05, p = 0.20 Maternal vitamin D & AA at 17 years: -0.002, p = 0.95 Maternal vitamin D & AA at 17 years: -0.009, p = 0.82 |
Cell-type composition |
| Early life dietary intake / nutritional biomarker | |||||||
| Simpkin (2015), UK [58] | AIRES 1,018 (51) | Breastfeeding | Infinium Human Methylation450 BeadChip to estimate Horvath epigenetic age | Cord blood & blood | Birth, 7.5y, 17.1y |
Correlations between early life variable and age acceleration: Breastfeeding & AA at birth: r = 0.035, p = 0.30 Breastfeeding & AA at 7 years: r = -0.010, p = 0.76 Breastfeeding & AA at 17 years: r = 0.026, p = 0.43 |
Cell-type composition |
| Rossnerova (2013), Czech Republic [97] | Asthmatics:100 (45). Controls:100(45) |
Breastfeeding | Infinium Human Methylation27 BeadChip | Blood | 11.6±2y | Breastfeeding was associated with overall DNA methylation, but no statistical test performed | |
| Obermann-Borst (2013), The Netherlands [64] | 120 (50) | Breastfeeding | LEP using mass-spectrometry based method | Blood | 17± 2.5m |
% absolute methylation change (SE), p Duration breast feeding: -0.6 (0.2), p = 0.04 |
Batch, correlation between 7 CpGs, birth weight, growth rate, smoking, BMI, GA, sex folic acid |
| Tao (2013), US [65] | 639 (100) breast cancer cases | Breastfeeding | E-cadherin, p16 and RAR-β2, using PCR | Breast tumour tissue | 57.5y ±11.3 |
OR (95%CI) for methylation breastfed yes (ref) vs no E-cadherin Premenopausal group: 1.21(0.50,2.93) Postmenopausal group:1.06(0.64,1.77) P16 Premenopausal group: 2.75(1.14,1.67) Postmenopausal group:0.79(0.49,1.26) RAR-β2, Premenopausal group: 1.18(0.53,2.62) Postmenopausal group: 1.30(0.83–2.04) |
Age, education, race, oestrogen receptor status |
| Wijnands (2015), UK [98] | 120 (41.7) | Breastfeeding & lipid biomarkers | LEP & TNFα using mass-spectrometry based method | Blood | 17±2.5m |
%Absolute methylation change (i.e. methylation change per SD change in biomarker (SE)) TNFα Total cholesterol: -1.0(0.5), p = 0.036. (Additional adjustment for HDL attenuated the results p = 0.07) Triglycerides: 0.1(0.5), p = 0.773 HDL-cholesterol:-1.2(0.5), p = 0.013. (Adjustment for maternal HDL slightly attenuated the association p = 0.08) LDL- cholesterol:-0.8(0.5), p = 1.00 %Absolute methylation change (β(SE)) LEP Total cholesterol:-0.6(0.3), p = 0.11 Triglycerides: 0.1 (0.4), p = 0.71 HDL-cholesterol:-3.4 n(1.5), p = 0.02. (Adjustment for maternal HDL slightly attenuated the association p = 0.041) LDL- cholesterol: -1.7 (1.5), p = 0.25 Bonferroni correction attenuated to nonsignificant estimates TNFα methylation was not associated with duration of breastfeeding. LEP methylation was significantly associated with duration of breastfeeding: -0.6 (95%CI -1.19, -0.01) per increment in breastfeeding duration category |
Bisufite batch |
| Fryer (2011), UK [25] | 12 (92) | Plasma homocysteine (birth) | Infinium Human Methylation27 BeadChip | Cord blood | Birth | Two clusters were identified following unsupervised hierarchical clustering to identify underlying methylation β-value across samples. Plasma homocysteine was lower (p = 0.038) in cluster B. There was no difference in serum folate (estimates not presented). 298 CpGs associated with plasma homocysteine (p<0.05) | |
| Fryer (2009), UK [88] | 24 (58.3) | Plasma homocysteine & serum folate (birth) | LINE-1 methylation using pyrosequencing | Cord blood | Birth |
Correlation with LINE-1 methylation: Cord plasma homocysteine: β = -0.69, p = 0.001 (p = 0.004 following adjustment) Cord serum folate: β = 0.21, p = 0.34 |
Sex, GA, maternal age, parity, BMI, serum folate, and maternal folic acid intake |
| McKay (2012), UK [89] | The North Cumbria Community Genetics Project 294 (48) | RBS folate & serum B12 (GA 39.5 ± 1.4w) | Global DNA methylation using LUMA & IGF2, IGFBP3, ZNT5 using pyrosequencing | Cord blood | Birth | Methylation of the IGFBP3 locus inversely correlated with infant vitamin B12 concentration (r = -0.16, p = 0.007) | Sex, GA, infant MTHFR genotype |
| Nafee (2009), UK** [31] | 24(NR) | Homocysteine (birth) | LINE-1 | Cord blood | Birth | LINE-1 methylation levels were inversely correlated with cord blood homocysteine (p = 0.01, r = -0.688) | |
| Perng (2012), Columbia [50] | BSCC 568(53.7) | Erythrocyte folate, plasma vitamin B12, vitamin A ferritin (an indicator of iron status), serum zinc concentrations (5-12y) | LINE-1 using pyrosequencing | Blood | (5-12y) |
LINE-1 methylation β(95%CI) & Erythrocyte Folate (nmol/L), All ptrend = 0.51: Q1: n = 139, ref Q2: n = 139, -0.03(-0.18, 0.11) Q3: n = 139, 0.01(-0.14, 0.16) Q4: n = 139, 0.04(-0.11, 0.19) LINE-1 methylation β(95%CI) & plasma B12 (pmol/L), All ptrend = 0.51: Q1: n = 137, ref Q2: n = 136, -0.04(-0.19, 0.11) Q3: n = 134, 0.06(-0.22, 0.09) Q4: n = 136, -0.12(-0.28, 0.04) LINE-1 methylation β(95%CI) & serum zinc (umol/L), All ptrend = 0.60: Q1: n = 140, ref Q2: n = 142, 0.014(-0.14, 0.16) Q3: n = 141, 0.07(-0.08, 0.23) Q4: n = 141, 0.02(-0.14, 0.18) Adjusted: LINE-1 methylation β(95%CI) & plasma ferritin (ug/L), All ptrend = 0.22: Q1: n = 141, ref Q2: n = 139, -0.16(-0.31, -0.01) Q3: n = 143, -0.08(-0.24, 0.07) Q4: n = 141, -0.13(-0.28, 0.03) LINE-1 methylation β(95%CI) & plasma vitamin A (umol/L), All ptrend = 0.006: <0.700: ref 0.70–1.05: -0.07(-0.24, 0.10) ≥:1.050, -0.19(-0.36, -0.02) |
Sex, vitamin A, CRP, maternal BMI, household socioeconomic position |
| Ba (2011), China [93] | 99 (48) | B-vitamin biomarkers (96% GAD 37-41w) | IGF2 2 promoters using methylation-specific PCR | Cord blood | Birth (96% GAD 37-41w) |
Promoter P2: Mean change per SD of each characteristic (p): cord blood serum folate: 0.18 (0.07) cord blood serum vitamin B12: -0.03 (0.75) Promoter P3: Mean change per SD of each characteristic (p): cord blood serum folate:-0.03 (0.77) cord blood serum vitamin B12: -0.04 (0.60) |
Mother's age, maternal pregnancy BMI, weight gain during pregnancy, mother's highest education level, parity, supplementation intake during pregnancy, Newborn's birth weight and birth length, Newborn's sex and GA |
| Haggarty (2013), UK [81] | 913 (46) | RBS folate (GAD: 39.5 (95%CI: 39.4, 39.6w)) | IGF2 (4 CpGs), PEG3 (7 CpGs), SNRPN (15q11, 4 CpGs) LINE-1 (4 CpGs) using pyrosequencing | Cord blood | Birth |
LINE-1 methylation: cord RBC folate 100 nmol/L: β = -0.08 (-0.12,-0.03), p = 0.001 PEG-3 methylation: cord RBC folate 100 nmol/L: β = -0.11 (-0.16,-0.05), p<0.001 SNRPN methylation: cord RBC folate 100 nmol/L: β = -0.002 (-0.09,0.09), p = 0.96 IGF2 methylation: cord RBC folate 100 nmol/L: β = 0.11 (0.02,0.20), p = 0.02 |
|
| Voisin (2015), Greece [99] | Greek Healthy Growth Study, Obese: 35 (68) Normal weight: 34 (66) |
24-hour recall for %energy from fat, cholesterol intake, MUFA/SFA, PUFA/SFA & MUFA+PUFA (~10y) | Infinium Human Methylation27 BeadChip | Blood | ~10y | The methylation levels of one CpG island shore and four sites were significantly correlated with total fat intake. No significance was found for cholesterol intake. The methylation levels of 2 islands, 11 island shores and 16 sites were significantly correlated with PUFA/SFA; of 9 islands, 26 island shores and 158 sites with MUFA/SFA; and of 10 islands, 40 island shores and 130 sites with (MUFA+PUFA)/SFA Top 10 most significant CpG sites/islands (Gene, Coefficient, adjusted p: %Energy from fat GPS1: -0.0135, p = 0.006 TAMM41: 0.00987, p = 0.006 TAS2R13: -0.0118, p = 0.012 MZB1: 0.0145, p = 0.023 TXNIP: 0.0148, p = 0.043 MUFA/SFA: ALDH3A2: -0.289, p = 0.00097 MYLK3: -0.238, p = 0.00363 LOC642852: –0.317, p = 0.00364 TPPP2: –0.309, p = 0.00364 RXFP2: -0.262, p = 0.00364 TMEM80: -0.245, p = 0.00364 SEMA3G: 0.28, p = 0.00388 VCAM1: -0.259, p = 0.00482 KRT73: -0.245, p = 0.00496 KRTCAP2: -0.30, p = 0.0051 PUFA/SFA CBR1: 1.28, p = 4.02e–06 RBCK1: 0.687, p = 2.3e–05 ABHD16A: -0.302, p = 7.18e–05 KRT23: -0.326, p = 0.00536 PDE3A: -0.274, p = 0.0066 NCOA1: -0.42, p = 0.00722 PCED1A: -0.41, p = 0.00914 MRPL13: 0.308, p = 0.00914 AKR7A2: 0.237, p = 0. 00914 FAM154A:- 0.357, p = 0.0193 (MUFA+PUFA)/SFA MRPL13: 0.186, p = 0.000952 NCOA1: -0.233, p = 0.00308 PCED1A:- 0.213, p = 0.00308 CCNA2:- 0.126, p = 0.00308 LCE1B:- 0.254, p = 0.00352 ALDH3A2: -0.176, p = 0.00352 MYLK3: -0.166, p = 0.00352 GBP7: -0.175, p = 0.00352 DGKI:- 0.178, p = 0.00352 DNTTIP: 0.148, p = 0.00352 |
Tanner stage, cell-type composition |
| De La Rocha (2016), Mexico [100] | 49 (55) | Serum fatty acids | Global DNA methylation using total 5-methyldeoxycytosine | Blood | Lactating infant (89.6±68.2d) |
Change in %methylation per one % increase in FA serum C20:4 (arachidonic acid): β = 0.08, p = 0.04 serum C20:5 (eicosapentaenoic acid): β = 0.099, p = 0.04 No significant associations with other fatty acids (data not shown in main paper) |
Age, birth weight, normalised weight gain |
| Lee (2012), US [26] | THREE, 141 (~47) | Serum copper,(87% GAD ≥37w) | NFIX, FAPGE, MSRB3 using pyrosequencing | Cord blood | Birth |
Association(95% CI) with serum copper ug/dl in cord blood: NFIX: β = 0.13 (0.06,0.20) RAPGE: β = -0.10 (-0.16,-0.05) MSRB3: β = -0.15 (-0.21,-0.08) |
Batch effects |
| Famine / Seasonality | |||||||
| Tobi (2015), The Netherlands [101] | Dutch Hunger Winter 885 (54) (Exposure during gestation:348,Periconceptional 74,Time-controls:160, Family-controls: 303) |
Famine | Infinium HumanMethylation450 BeadChip | Blood | 58.9±.5y | Famine vs. time-and family controls: % methylation (95%CI) Famine in 1–10 weeks gestation (n = 73) cg20823026 (FAM150B/TMEM18): 2.3 (1.5–3.1), p = 3.1x10-8 cg10354880 (SLC38A2):0.7(0.5,0.9), p = 5.9x10-7 cg27370573 (PPAP2C): 2.7(1.7,3.7), p = 3.6x10-7 cg11496778(OSBPL5/MRGPRG): -2.3(-3.1, -1.5), p = 2.1x10-7 Famine in 11–20 weeks gestation (n = 123): no significant cpgs Famine in 21–30 weeks gestation (n = 143): no significant cpgs Famine in 31- delivery (n = 128): no significant cpgs Any exposure to famine: cg15659713 (TACC1): 1.2(0.8,1.7), p = 2.0x10-7 cg26199857(ZNF385A): 2.0(1.3,2.7), p = 1.5x10-7 Conceived during extreme famine, but exposed for short period in gestation: cg23989336 (TMEM105):-3.5(-4.6,-2.3), p = 1.0x10-7 |
Age, sex, batch effects, cell heterogeneity, smoking status, current macronutrient and micronutrient intake and SEP |
| Finer (2016), Bangladesh [102] | 143(58) | Famine (postnatal exposure 1-2y or exposure during gestation or unexposed) | Infinium HumanMethylation450 BeadChip 16 MEs: VTRNA2-1, PAX8, PRDM9, HLA-DQB2, PLD6, ZFP57, AKAP12, ATP5B, LRRC14B, SPG20, BOLA, RBM46, ZFYVE28, EXD3, PARD6G, ZNF678, ZFYVE28 |
Blood | Postnatal exposed: 31±0.4y Exposure during gestation:30±0.3y Unexposed: 28±0.3y |
Postnatal exposure n = 49 vs gestational exposure n = 40 vs unexposed n = 54 Genome-wide analyses No differences between groups at 5% FDR Targeted DNA methylation Methylation differences between groups seen in 6/16 MEs at p<0.05, driven by gestational exposure group:: VTRNA2-1, PAX8, PRDM9, ZFP57, BOLA, EXD3 z-score for mean methylation across all 16 MEs: gestational exposure: -0.24 postnatal exposure: -0.14 unexposed: -0.15 ANOVA p = 0.0003 |
Cell composition |
| Lumey (2012), The Netherlands [103] | Dutch Hunger Winter 947 (54) (Prenatal:350, Unexposed time controls:290, Unexposed same-sex sibling:307) |
Famine | LINE-1 & Sat-2 using pyrosequencing Global methylation using LUMA |
Blood | Prenatal exposure group:58.9 ±0.5y Unexposed time controls: 58.5 ±1.6y Unexposed same-sex siblings: 57.3± 6.3y |
Changes in DNA methylation (%units) in exposed vs. all non-exposed: Global methylation: Mean % (SD): 75.2% (4.7) B(95% CI): -0.15 (-0.49, 0.81), p = 0.63 LINE-1 methylation % (SD): Mean % (SD):77.1% (2.5) B(95% CI): -0.05(-0.33, 0.22), p = 0.70 Sat2 methylation % (SD): Mean % (SD):122.2 (56.2) B(95% CI): -0.51 (-7.38, 6.36), p = 0.88 |
Age, within family clustering |
| Heijmans (2008), The Netherlands [104] | Dutch Hunger Winter 244 (~54) (periconceptional:60, late gestation: 62, Unexposed same-sex sibling:122) |
Famine | IGR2 DMR (5 CpGs) using mass spectrometry-based method | Blood | Periconceptional group: 58.1±0.35y Late gestation group: 58.8± 0.4y Controls: 57.1± 5.5y |
Mean (SD) methylation in those periconceptionally exposed to famine vs. non-exposed siblings: Average: 0.488(0.047) vs. 0.515(0.055), p = 5.9x10-5 CpG1: 0.436(0.037) vs, 0.470(0.041), p = 1.5x10-4 CpG2 and 3: 0.451(0.033) vs. 0.473(0.055), p = 8.1x10-3 CpG4: 0.577(0.114 vs. 0.591(0.112), p = 0.41 CpG5: 0.491(0.061) vs. 0.529(0.068), p = 1.4x10-3 No difference in methylation of IGF2 DMR between a subset exposed in late gestation and unexposed siblings |
Age and family relations |
| Tobi (2014), The Netherlands [105] | Dutch Hunger Winter 48 (50) | Famine (early gestation) | 1.2M CpGs using RRBS | Blood | 58.1±0.35y |
Genomic annotation-centred analysis of differential methylation after famine (vs. unexposed sibling), pFDR: Genomic annotations** Non-CGI, ‘bona fide’ promoters: 0.026 Enhancers: 0.026 DNaseI/FAIRE-seq regions: 0.036 Middle exons: 0.036 Developmental enhancers type I: 0.036 ‘bona fide’ CGI shores: 0.053 Non-coding RNA: 0.053 Conserved regions: 0.053 CGI shores: 0.053 3’UTR: 0.085 Non genic CGI: 0.085 ‘Bonafide’ CGI border: 0.085 Developmental enhancer type II: 0.15 CGI: 0.15 Introns: 0.15 hESC bivalent chromatin domains: 0.28 Bonafide CGI: 0.32 Cell-type specific gene promoters: 0.32 First exons: 0.36 Promoters: 0.36 HSC bivalent chromatin domains: 0.36 Imprinted promoters: 0.36 ‘Bona fide’ CGI promoter: 0.37 CTCF insulators from CD4+ cells: 0.37 Imprinted DMRs: 0.37 Putative metastable epialles: 0.47 Variably methylated regions: 0.57 Promoters cancer genes: 0.63 Within the 5 annotations found to be significant, 181 DMRs were associated with prenatal famine pFDR<0.05 |
|
| Tobi (2009), the Netherlands [106] | Dutch Hunger Winter 244 (~54) (periconceptional:60, late gestation:62, unexposed same-sex sibling:122) |
Famine | GNASAS, GNAS A/B, MEG3, KCNQ1OT1, INSIGF and GRB10, IGF2R, IL10, TNF, ABCA1, APOC1, FTO, LEP, NR3C1 and CRH using mass-spectrometry based method | Blood | Periconceptional group: 58.1±0.35y Late gestation group: 58.8± 0.4y Controls: 57.1± 5.5y |
Within-pair differences divided vs sibling controls, p: Periconceptional exposure GNASAS: 0.24, 3.1x10-6 MEG3: 0.21, 8.0x10-3 (non-significant after Bonferroni correction) IL10: 0.37, 1.8x10-6 ABCA1: 0.21, 8.2x10-4 LEP: 0.24, 2.9x10-3 INSIGF: -0.61, 2.3x10-5 Non-significant for all other loci Late gestation exposure: No associations except for reduction in GNASAS: -0.26, 1.1x10-7 Non-significant for all other loci |
Family relatedness, bisulphite batch, age |
| Veenendaal (2012), The Netherlands [107] | Dutch Hunger Winter 759 (54%) (periconceptional:60, late gestation:62, unexposed same-sex sibling:122) |
Famine | PPARγ, GR1-C, PI3Kinase, LPL using PCR | Blood | 58±1y |
Methylation differences % (95%CI) for exposed vs. unexposed: Late gestation: GR: 0.60 (-16.39, 21.05) LPL: 11.01 (-5.35, 30.34) PI3Kinase: 6.18 (-42.25, 95.03) PPARγ: -2.37 (-14.53, 11.52) Mid-gestation GR: -5.26 (-22.04, 15.14) LPL: 12.08 (-5.45, 32.84) PI3Kinase: -32.36 (-64.33, 28.27) PPARγ: -8.70 (-20.63, 5.02) Early gestation: GR: 6.82 (-15.55, 35.12) LPL: 9.20 (-10.95, 34.04) PI3Kinase: -40.84 (-72.56, 27.38) PPARγ:-6.76 (-21.08, 10.30) No significant associations were found |
Maternal age, sex and parity |
| Waterland (2010), The Gambia [108] | The Keneba cohort 50 (50%) Conceived in rainy season:25, conceived in dry season:25 |
Famine | MEs: BOLA3, FLJ20433, PAX8, SLOTRK1, ZFYVE28 using pyrosequencing | Blood | Conceived in rainy season:6.61±2.73y Conceived in dry season:7.05±2.67y |
At all 5 MEs, DNA methylation was significantly higher among individuals conceived ruing the rainy season (i.e. hungry season): BOLA3: p = 0.03 FLJ20433: p = 0.03 PAX8: p = 0.02 SLOTRK1: p = 0.006 ZFYVE28: p = 0.002 Overall: p = 0.0001 Effect sizes were NR but highlighted as being large e.g. rainy season was associated with absolute methylation increments of over 10% at PAX8 and ZFYVE28 |
|
*Studies spanning more than one exposure may appear twice in the table;
** Abstract
AA: Age acceleration; ARIES: Accessible Resource for Integrated Epigenomic Study; BMI: Body Mass Index; BSCC: Bogotá School Children Cohort; CHO: Carbohydrate; CI: Confidence Interval; CBMCs: Cord Blood Mononuclear Cells; COBRA: Combined Bisulfite Restriction Analysis; D: Days; DMR: Differentially Methylated Region; DA: Dizygotic; FDR: False discovery rate; FFQ: Food frequency Questionnaire; GAD: Gestational Age at Delivery; HUVEC: Human Umbilical Vein Endothelial Cells; LUMA: Luminometric methylation assay); M: Months; MANOE: Maternal Nutrition and Offspring’s Epigenome Study; MoBA: Norwegian Moher and Child Cohort Study; MUFA: Monounsaturated fatty acid; MZ: Monozygotic; NEST: Newborn Epigenetics Study; NR: Not Reported; OR: Odds Ratio; PAH: Princess Anne Hospital Study; PETS: Peri/postnatal Epigenetic Twins Study PUFA: Polyunsaturated fatty acid; RBC: Red Blood Cell; SD: Standard Deviation; SE: Standard Error; SEP: Socioeconomic Position; SFA: Saturated Fatty Acid; THREE: Tracking Health Related to Environmental Exposures Study; W; Weeks Y: Year