Towards objective measurements of habitual dietary intake patterns: comparing NMR metabolomics and food frequency questionnaire data in a population-based cohort

Table 4 Univariate statistics for metabolites that discriminated significantly between quartiles 1 (Q1) and 4 (Q4) in OPLS-DA models for relative Mediterranean Diet Score, Plant based Diet Index and Unhealthy Plant based Diet Index

Metabolite¹	¹H chemical shift region	rMediterranean Diet Score			Plant based Diet Index			Unhealthful Plant based Diet Index
Metabolite¹	¹H chemical shift region	Q1	Q4	q²	Q1	Q4	q²	Q1	Q4	q²
1,5-anhydrosorbitol	3.893	↑	↓	1.46E-3				↓	↑	2.55E-11
1,5-anhydrosorbitol + serine	3.996				↓	↑	0.0300
1,5-anhydrosorbitol + serine	4.005							↓	↑	6.69E-5
1,5-anhydrosorbitol + unidentified	3.367	↑	↓	2.20E-9				↓	↑	6.51E-9
2-hydroxyvalerate	0.931				↑	↓	3.02E-3
3-hydroxybutyrate	2.363				↑	↓	0.0120
acetate	1.935							↑	↓	1.35E-14
glutamine	2.454				↓	↑	2.17E-3
glutamine + unidentified	2.119				↓	↑	3.81E-3
lactate	1.34	↑	↓	2.34E-4				↓	↑	5.21E-3
lactate + proline	4.147	↑	↓	2.32E-4				↓	↑	3.05E-4
lipid	1.29							↓	↑	8.87E-3
NAAs/fatty acyl allylic protons from lipoprotein lipids	2.018							↓	↑	9.78E-3
phosphocholines, phosphoethanolamine + unidentified	3.239							↑	↓	2.10E-7

1 Mean chemical shift (ppm) for the bucket used for Mann-Whitney U-test. Selected if variables had loading scores − 0.1 ≥ w ≥ 0.1 and if they had among the 30 highest variable influence on projections values. 2 Mann-Whitney U-test. Significant Mann-Whitney U-test after FDR correction for 19 univariate tests (q < 0.05). ↑ = high, ↓ = low
Abbreviations: Q1, first quartile; Q4, last quartile

ISSN: 1475-2891