Table 3.
Plasma fatty acids demonstrating >twofold changes before and after a 4‐day, 51‐km cross‐country ski march
| Subpathway | Biochemical name | Fold change (POST/PRE) |
|---|---|---|
| Medium‐chain fatty acid | Caprate (10:0) | 2.07 |
| 10‐Undecenoate (11:1n1) | 2.10 | |
| Laurate (12:0) | 3.32 | |
| 5‐Dodecenoate (12:1n7) | 5.36 | |
| Long‐chain fatty acid | Myristate (14:0) | 4.12 |
| Myristoleate (14:1n5) | 5.70 | |
| Pentadecanoate (15:0) | 2.04 | |
| Palmitate (16:0) | 2.20 | |
| Palmitoleate (16:1n7) | 4.74 | |
| Margarate (17:0) | 2.73 | |
| 10‐Heptadecenoate (17:1n7) | 4.24 | |
| Nonadecanoate (19:0) | 2.07 | |
| 10‐Nonadecenoate (19:1n9) | 4.34 | |
| Arachidate (20:0) | 2.08 | |
| Eicosenoate (20:1) | 4.71 | |
| Erucate (22:1n9) | 3.42 | |
| Oleate/vaccenate (18:1) | 2.51 | |
| Polyunsaturated fatty acid (n3 and n6) | Stearidonate (18:4n3) | 3.16 |
| Docosapentaenoate (n3 DPA; 22:5n3) | 2.73 | |
| Linoleate (18:2n6) | 2.86 | |
| Linolenate [alpha or gamma; (18:3n3 or 6)] | 4.06 | |
| Adrenate (22:4n6) | 2.39 | |
| Docosadienoate (22:2n6) | 3.12 | |
| Dihomolinoleate (20:2n6) | 3.93 | |
| Fatty acid, branched | 13‐Methylmyristate | 2.22 |
| 15‐Methylpalmitate | 2.88 | |
| 17‐methylstearate | 2.49 | |
| Fatty acid, dicarboxylate | Adipate | 4.50 |
| 3‐Methyladipate | 3.59 | |
| Sebacate (decanedioate)a | 6.06 | |
| Dodecanedioate | 4.40 | |
| Tetradecanedioate | 3.59 | |
| Hexadecanedioate | 3.08 | |
| Octadecanedioate | 2.82 | |
| Fatty acid, monohydroxy | 3‐Hydroxyhexanoate a | 6.25 |
| 3‐Hydroxyoctanoate a | 5.02 | |
| 3‐Hydroxydecanoatea | 4.68 | |
| 3‐Hydroxysebacate | 10.04 | |
| 3‐Hydroxylaurate | 3.71 | |
| 5‐Hydroxyhexanoate | 2.68 | |
| 13‐HODE + 9‐HODE | 2.45 | |
| 9‐Hydroxystearate | 3.09 | |
| Fatty acid, acyl glutamine | Hexanoylglutamine | 13.80 |
Data are fold changes calculated using the mean for each time point. Repeated measures ANOVA (n = 24) used to examine main effect of time, diet, and their interaction on metabolites measure before (PRE) and after (POST) military training. P values were adjusted using the Benjamini–Hochberg correction (Q); main effect of time, Q < 0.10 for all. All diet by time interactions were not statistically significant (Q > 0.20). Metabolites in bold font are those with the strongest influence on prediction accuracy in the random forest analysis (see Fig. 1B). HODE, hydroxyl‐octadecadienoic acid.
Metabolite was >twofold higher in the cluster containing participants who finished the course late (POST2 vs. POST1, independent samples t test; Q < 0.10).