Data S-2 List of Consumables Used Throughout the MEDE Studies

Electronic supplementary material

Data S-1 Meals composition

Data S-2 List of consumables used throughout the MEDE Studies.

Fig. S-1 Modeling characteristics of FIE-MS data from fasting and postprandial urine samples (negative ionization mode)

Fig. S-2 Modeling characteristics of FIE-MS data from fasting and postprandial urine samples (positive ionization mode)

Fig. S-3 Confirmation of the presence of hesperetin sulfate in urine samples collected 6 and 8h after consumption of the standardized breakfast

Table S-1 Anthropometric data from MEDE Study 1 and 2 subjects

Table S-2 Nutritional and inflammatory status markers1 from MEDE Study 1 and 2 subjects

Table S-3 Lifestyle data from MEDE Study 1 and 2 subjects

Table S-4 Eating, drinking, and urination behaviors data from MEDE Study 1 and 2 subjects

Table S-5 Effect of data transformation and normalization by metadata on data modeling characteristics

Table S-6 The average 35 top-ranked discriminatory features (positive ion mode, 100-510 m/z) between the ‘PRE’1 and fasting2 urine samples for the three separate studies.

Table S-7 The average 35 top-ranked discriminatory features (negative ion mode, 100-510 m/z) between the ‘PRE’1 and fasting2 urine samples for the three separate studies.

Table S-8 Top positive ion mode discriminatory features (m/z 100-550) between fasting and postprandial spot urine samples in rank order

Table S-9 Top negative ion mode discriminatory features (m/z 100-550) between fasting and postprandial spot urine samples in rank order

Data S-1 Meals composition

The standardized evening meal provided 2.9 MJ (24.6 % as protein, 35.8 % as fat and 39.6 % as carbohydrate) and the standardized breakfast 1.9 MJ (9.2 % as protein, 22.2 % as fat and 68.6 % as carbohydrate). Ready-prepared meals (‘Chicken in a pot’), still mineral water, orange juice, tea, croissants, corn flakes and semi skimmed milk were purchased from Sainsbury’s (Etherstone Avenue, Heaton, Tyne and Wear NE7 7GW, UK) and have been produced in the UK for Sainsbury’s Supermarkets Ltd (33 Holborn, London EC1N 2HT, UK), except for the corn flakes produced by Kellogg Marketing and Sales Company Ltd (Manchester M16 0PU, UK). The chocolate éclairs were donated by Marks & Spencer (Marks and Spence plc, PO Box 3339 Chester CH99 9QS, UK). Sugar sachets and skimmed milk portion pot were purchased in Costco (Mandela Way, Gateshead NE11 9DH, UK) and had been produced respectively in UK by Tate & Lyle (Thames Refinery, Silvertown, London E16 2 EW, UK) and in Holland for the Kenco Coffee Company (Cheltenham, GL50 3AE, UK).

Data S-2 List of consumables used throughout the MEDE Studies.

Jugs were purchased from Thermo Fisher Scientific (Bishopmeadow Road, Loughborough LE11 5RG). 24h urine containers were purchased from International Scientific Supplies Ltd (Richmond House, Canal Road, Bradford, West Yorkshire, BD2 1AL). Micro tubes were purchased from Sarstedt (68 Boston Road, Leicester LE4 1AW), Falcon® tubes from Dutscher Scientific UK Ltd (Dolphin Way, West Thurrock, RM19 1NZ), and vacutainer® tubes from Southern Syringe Services (Warren Bruce Road, Manchester, M17 1LB).

Product / Reference / Details / Brand
Jug with spout & moulded graduations PP 500 mL / FB55940 / Polypropylene / Fisherbrand
24h urine container / UR2401 / Polyethylene / ISS Ltd
1.5 mL Micro tube / 72.69 / Polypropylene, safety cap / Sarstedt
50 mL Falcon® tube / 352070 / Polypropylene, flat seal screw / Becton Dickinson
4 mL Lithium/Heparin vacutainer® tube / 367883 / Plastic, LH 68 I.U. / Becton Dickinson
5mL Gel vacutainer® tube / 367954 / Plastic / Becton Dickinson
2mL Fluoride/Oxalate vacutainer® tube / 367934 / Plastic, FX 5mg 4mg / Becton Dickinson
3mL EDTA vacutainer® tube / 367698 / Glass, K3E 7.5% 0.072 mL / Becton Dickinson
4.5mL EDTA vacutainer® tube / 367690 / Glass, K3E 15% 0.054 mL / Becton Dickinson

Fig. S-1 Modeling characteristics of FIE-MS data from fasting and postprandial urine samples (negative ionization mode)

Flow Infusion Electrospray-ionization Mass Spectrometry (FIE-MS) data (negative ion; m/z 100-510) generated from the ‘fasting’ spot urine samples and the postprandial urine samples collected after the consumption of the standardized breakfast (Study 1 after 3h, and Study 2 after 2h and 4h, where A denotes the first visit and B denotes the second visit and Study 3 after 1.5, 3 and 4.5h) were subjected to multivariate classification using 2 different algorithms. Statistical measures generated were: ‘Accuracy’, classification accuracy; ’AUC’, area under the receiver operating characteristic curve; and ‘Margin’, classification margin.

Fig. S-2 Modeling characteristics of FIE-MS data from fasting and postprandial urine samples (positive ionization mode)

Flow Infusion Electrospray-ionization Mass Spectrometry (FIE-MS) data (positive ion; m/z 100-510) generated from the ‘fasting’ spot urine samples and the postprandial urine samples collected after the consumption of the standardized breakfast (Study 1 after 3h, and Study 2 after 2h and 4h, where A denotes the first visit and B denotes the second visit and Study 3 after 1.5, 3 and 4.5h) were subjected to multivariate classification using 2 different algorithms. Statistical measures generated were: ‘Accuracy’, classification accuracy; ’AUC’, area under the receiver operating characteristic curve; and ‘Margin’, classification margin.

Fig. S-3 Confirmation of the presence of hesperetin sulfate in urine samples collected 6 and 8h after consumption of the standardized breakfast

(a) nano-flow FT-ICR-MS derived accurate mass of tentatively identified hesperetin sulfate [M-H]1- from a pool of 4 randomly selected volunteers; (b) FIE MS2 spectrum of m/z 381 showing the loss of m/z 80 (m/z 301.09 circled) corresponding to the loss of the sulfonate moiety; (c) FIE MS3 spectrum of m/z 301; (d) FIE MS2 spectrum of m/z 301 of synthetic hesperetin [M-H]1-.

Table S-1 Anthropometric data from MEDE Study 1 and 2 subjects

Men (n=14) / Women (n=10)
mean / CV% / mean / CV%
Age (y) / 38 / 48 / 60 / 7
Time since last period (y) / n/a / n/a / 10 / 49
Height (m) / 182 / 6 / 161 / 4
Weight (kg) / 79.6 / 9 / 64.7 / 12
BMI (kg/m2) / 24.2 / 9 / 24.9 / 12
Body fat content (%) / 19.1 / 27 / 32.5 / 18
Waist circumference (cm) / 88 / 7 / 83.8 / 11
Hips circumference (cm) / 103 / 4 / 102 / 7
Waist to Hips Ratio / 0.86 / 8 / 0.82 / 7
Systolic blood pressure (mm Hg) / 125 / 10 / 129 / 7
Diastolic blood pressure (mm Hg) / 76 / 10 / 81 / 5

Table S-2 Nutritional and inflammatory status markers1 from MEDE Study 1 and 2 subjects

1 Concentrations measured in fasting blood samples.

Men (n=14) / Women (n=10)
mean / CV% / mean / CV%
Plasma glucose (mmol/L) / 5.1 / 10 / 5 / 7
Serum insulin (mU/L) / 6.5 / 30 / 6.9 / 36
Insulin resistance index (HOMA) / 1.5 / 32 / 1.5 / 41
Serum triglycerides (mmol/L) / 1.1 / 49 / 1.2 / 24
Serum total cholesterol (mmol/L) / 4.6 / 19 / 6.3 / 7
Serum HDL cholesterol (mmol/L) / 1.5 / 31 / 1.9 / 24
Serum LDL cholesterol (mmol/L) / 2.6 / 18 / 3.8 / 17
Total/HDL cholesterol ratio / 3.2 / 20 / 3.5 / 24
Serum C-Reactive Protein (mg/L) / 1.9 / 164 / 1.2 / 91
Serum vitamin B12 (ng/L) / 324 / 32 / 382 / 34
Red cell folate (μg/L) / 216 / 35 / 215 / 59
Plasma selenium (μmol/L) / 1.06 / 23 / 1.11 / 17

Table S-3 Lifestyle data from MEDE Study 1 and 2 subjects

1 Expressed as energy expenditure due to physical activity

Men (n=14) / Women (n=10)
mean / CV% / mean / CV%
Daily Energy Intake (DEI: MJ) / 10.5 / 35 / 10.3 / 27
DEI as protein (%) / 15.5 / 13 / 16.1 / 19
DEI as fat (%) / 31.8 / 15 / 31.3 / 18
DEI as carbohydrates (%) / 52.8 / 10 / 52.6 / 15
DEI as alcohol (%) / 7.6 / 95 / 6.1 / 98
Calcium intake (mg/d) / 1183 / 38 / 1160 / 35
Vitamin C intake (mg/d) / 197 / 67 / 236 / 44
Iron intake (mg/d) / 18 / 48 / 20 / 24
Folate intake (μg/d) / 402 / 53 / 448 / 31
Carotene intake (mg/d) / 4 / 64 / 5.2 / 65
Non starch polysaccharide intake (g/d) / 25 / 63 / 30 / 35
Habitual physical activity (MJ)1 / 1.6 / 58 / 1.2 / 83
Pre-test day physical activity (MJ)1 / 1.6 / 77 / 0.6 / 68

Table S-4 Eating, drinking, and urination behaviors data from MEDE Study 1 and 2 subjects

1 Urine PRE samples; 2 Urine ‘0’ (fasting) samples.

Men (n=14) / Women (n=10)
mean / CV% / mean / CV%
Duration of standardized evening meal (min) / 21 / 31 / 24 / 34
Overnight water consumption (mL) / 347 / 40 / 321 / 38
Number of overnight voids of urine / 3 / 38 / 4 / 19
Volume (mL)1 / 1054 / 35 / 1385 / 18
Na (mmol/L)1 / 65 / 58 / 39 / 43
K (mmol/L)1 / 30 / 50 / 20 / 36
Creatinine (mmol/L)1 / 9.7 / 56 / 4 / 38
Na:creatinine ratio1 / 7.8 / 47 / 11.1 / 27
K:creatinine ratio1 / 3.5 / 42 / 5.7 / 26
Volume (mL)2 / 78 / 66 / 101 / 70
Na (mmol/L 2 / 75 / 46 / 69 / 35
K (mmol/L)2 / 67 / 39 / 82 / 40
Creatinine (mmol/L)2 / 17.2 / 49 / 12.9 / 60
Na:creatinine ratio2 / 6 / 68 / 6.7 / 56
K:creatinine ratio2 / 4.9 / 49 / 7.2 / 45

Table S-5 Effect of data transformation and normalization by metadata on data modeling characteristics

Principal Components-Linear Discriminant Analysis (PC-LDA) and Random Forest (RF) classification of data acquired by Flow Infusion Electrospray ionization Mass Spectrometry (100-510 m/z; positive and negative ion mode) analysis of fasting and postprandial urine samples from 12 individuals (Study 2). For PC-LDA, ‘time of urine collection’ was the class structure applied consisting of fasting, 2 and 4h postprandial, after data normalization as indicated in the table. Pair-wise Random Forest classification robustness comparisons were made between fasting and 2h postprandial samples only. Log10/TIC, logarithmic transformation of data and normalization using sample total ion count (TIC); * = Log10 of data prior to normalization; Tw, DF Eigenvalue; Margin, classification margin; ACC, classification accuracy; CRP, Serum C-Reactive Protein; HDL, High-density lipoprotein. Water intake: overnight intake for fasting samples, and test day intake for postprandial samples.

Table S-6 The average 35 top-ranked discriminatory features (positive ion mode, 100-510 m/z) between the ‘PRE’1 and fasting2 urine samples for the three separate studies.

Common mass signals found in two pair-wise comparisons or more are shaded; Common mass signals found in all three pair-wise comparisons are in bold; 1‘PRE’, pool of pre-test day over-night urine voids; 2fasting, spot urine sample after a 12h (minimum) fast.

Rank / Fasting vs PRE (+ve ion m/z)
Study 1 / Study 2 / Study 3
1 / 241 / 263 / 241
2 / 242 / 241 / 242
3 / 301 / 361 / 213
4 / 263 / 168 / 212
5 / 341 / 285 / 192
6 / 160 / 242 / 153
7 / 368 / 410 / 150
8 / 304 / 301 / 360
9 / 360 / 341 / 113
10 / 361 / 279 / 361
11 / 401 / 206 / 265
12 / 342 / 113 / 152
13 / 385 / 151 / 207
14 / 285 / 160 / 214
15 / 347 / 264 / 381
16 / 152 / 339 / 236
17 / 113 / 148 / 384
18 / 307 / 383 / 296
19 / 150 / 325 / 170
20 / 213 / 213 / 159
21 / 129 / 100 / 401
22 / 224 / 303 / 238
23 / 325 / 108 / 297
24 / 109 / 416 / 160
25 / 284 / 117 / 282
26 / 202 / 153 / 263
27 / 210 / 152 / 342
28 / 170 / 334 / 223
29 / 153 / 165 / 285
30 / 387 / 221 / 211
31 / 117 / 448 / 341
32 / 274 / 192 / 385
33 / 110 / 360 / 421
34 / 377 / 225 / 175
35 / 168 / 401 / 237

Table S-7 The average 35 top-ranked discriminatory features (negative ion mode, 100-510 m/z) between the ‘PRE’1 and fasting2 urine samples for the three separate studies.

Rank / Fasting vs PRE (-ve ion m/z)
Study 1 / Study 2 / Study 3
1 / 359 / 223 / 223
2 / 435 / 359 / 319
3 / 239 / 239 / 359
4 / 179 / 337 / 176
5 / 133 / 133 / 435
6 / 339 / 202 / 175
7 / 223 / 435 / 481
8 / 319 / 217 / 351
9 / 397 / 319 / 408
10 / 217 / 160 / 443
11 / 275 / 445 / 130
12 / 134 / 457 / 218
13 / 325 / 275 / 148
14 / 123 / 259 / 465
15 / 207 / 176 / 207
16 / 193 / 218 / 337
17 / 337 / 481 / 445
18 / 148 / 380 / 407
19 / 408 / 375 / 467
20 / 220 / 408 / 301
21 / 381 / 250 / 149
22 / 184 / 441 / 249
23 / 149 / 300 / 466
24 / 277 / 138 / 485
25 / 375 / 277 / 106
26 / 141 / 507 / 111
27 / 209 / 365 / 451
28 / 206 / 175 / 299
29 / 276 / 207 / 179
30 / 298 / 485 / 220
31 / 130 / 425 / 248
32 / 250 / 163 / 507
33 / 261 / 363 / 305
34 / 109 / 443 / 240
35 / 137 / 305 / 239