Skip to main content

Table 2 Summary of the characteristics of the 20 studies included in the review

From: Sex differences in the human metabolome

First author and year Country/ethnicity Subjects (total n) Sex (total n) Age/age groups Biologic matrix Analytical platform Statistics Quality score
Andraos S. 2021 [64] Australia 1166
1324
565 M
601 F
174 M
1150 F
11.4 ± 0.5
11.5 ± 0.5
46.2 ± 6.4
43.6 ± 4.8
years
(mean ± SD)
Plasma UHPLC–MS/MS Linear mixed-effects models
Pearson’s correlation
8
Caterino M. 2020 [35] Italy 291 152 M
139 F
(1–36 months)
1–6 months
7–12 months
13–24 months
25–36 months
Urine GC–MS Dunn’s test
Kruskal–Wallis test
Mann–Whitney U test
7
Caterino M. 2021 [36] Italy 311 174 M
137 F
48–72 h DBS LC–MS/MS Spearman’s correlation
PLS-DA
VIP
8
Chekmeneva E. 2018 [65] USA 132 N. F. 40–59 years Urine 1H NMR
DI-nESI-HRMS
UPLC–HRMS
PCA
OPLS-DA
10
De Paepe E. 2018 [66] Belgium 10 5 M
5 F
25–41 years Plasma
Urine
UHPLC–MS/MS CV-ANOVA
PCA
OPLS-DA
7
Dunn W. B. 2014 [38] UK 1200 N. F. (19–81 years)
 < 40 years
40–49 years
50–64 years
 > 64 years
Serum UPLC–MS (+)
UPLC–MS (−)
GC–MS
ANOVA
Kruskal–Wallis test
Mann–Whitney U test
PLS-DA
Random Forests
10
Fan S. 2018 [67] USA 120 60 M
60 F
N. F. Urine GC–MS Mann–Whitney U test
PLS-DA
PCA
6
Jovè M. 2016 [68] Spain 146 68 M
78 F
30–100 years Plasma LC–MS/MS (+)
LC–MS/MS (−)
ANOVA
PCA
Student’s t-test
7
Lau C.-H. E. 2018 [54] UK
France
Spain
Lithuania
Norway
Greece
(1192)
199
157
207
201
229
199
(54.6% M
45.4% F)
6–11 years Serum
Urine
1H NMR
FIA-MS/MS
LC–MS/MS
MWAS with multiple linear regression
PCA
Pearson’s correlation
8
Lawton K. A. 2008 [55] Caucasian
African-American
Hispanic
(269)
61
135
73
(131 M, 138 F) 
20M, 41 F
69 M, 66 F
42 M, 31 F
(20–65 years)
20–35 years
36–50 years
51–65 years
Plasma LC–MS (+)
LC–MS (−)
GC–MS
ANCOVA
ANOVA
9
Mittelstrass K. 2011 [24] Germany 3080
377
1452 M, 1552 F
197 M, 180 F
32–81 years
55–79 years
Serum LC–MS/MS PLS
Linear regression
Partial correlation analysis
7
Rist M. J. 2017 [56] Germany 200 99 M
101 F
36–80 years Plasma
Urine
1H NMR
FIA-MS/MS
GC(xGC)-MS
HILIC-MS/MS
UHPLC–MS/MS
Glmnet
PLS
SVM
8
Ruoppolo M. 2014 [15] Italy 76 35 M
41 F
20–45 years Serum HPLC
LC–MS/MS
SAM
Pearson’s correlation
Spearman’s correlation
6
Ruoppolo M. 2015 [57] Italy 3680 1856 M
1824 F
48–72 h DBS LC–MS/MS Mann–Whitney U test
Linear regression
PCA
8
Saito K. 2016 [58] Japan 60 15 M
15 F
15 M
15 F
25–35 years
25–35 years
55–64 years
55–65 years
Serum GC–MS
HILIC–MS (−)
UHPLC–MS (+)
UHPLC–MS (−)
Welch’s two-factor t-tests
PCA
9
Scalabre A. 2017 [59] France 90 66 M
24 F
(< 4 mo)
 < 1 mo
1 months
2 months
3 months
Urine 1H NMR PCA
O-PLS
CV-ANOVA
SRV
7
Slupsky C. M. 2007 [60] Canada 60 30 M
30 F
16–69 years Urine 1H NMR ANOVA
Mann–Whitney U test
PCA
PLS-DA
6
Thévenot E. A. 2015 [61] France 183 100 M
83 F
40.9 ± 10.3 years
(mean ± SD)
Urine UHPLC–MS(/MS) Mann–Whitney U test
O-PLS
Spearman’s correlation
7
Trabado S. 2017 [62] France 800 417 M
383 F
18–86 years
37.6 ± 17.2
(mean ± SD)
Plasma FIA-MS/MS
LC–MS/MS
CV-ANOVA
OPLS-DA
PCA
8
Vignoli A. 2018 [63] Italy 844 661 M
183 F
41.0 ± 12.0 years
(median ± SD)
Plasma 1H NMR PLS-DA
Wilcoxon rank-sum test
6
  1. Most of the categories/terms in the table were ordered alphabetically. Whereas not specified in the papers, the generic LC–MS/MS term was reported for targeted metabolomic analyses; ( +) and (-) refer to positive and negative ionization modes, respectively. Missing information are reported as N.F. (not found). mo = months, y = years, SD = standard deviation