Abstract
In contrast to the conventional spermiogram, metabolomics approaches give insights into the molecular composition of semen and may provide more detailed information on the fertility status of the respective donor. Given the intra-individual variability of spermiogram parameters between two donations, this study sought to elucidate the biological variability of the seminal plasma metabolome over an average period of 8 weeks. Two time-shifted semen samples from 15 healthy donors were compared by a targeted metabolomics approach utilizing the Biocrates AbsoluteIDQ p180 kit. Next to intraclass correlation coefficients (ICC), which represent a measure of reliability, coefficients of variation within individuals (CVW) and coefficients of variation between individuals (CVB) were calculated for each metabolite to demonstrate its stability. Furthermore, men were divided into two cohorts, a similar sperm concentration (SSC) and a differing sperm concentration (DSC) cohort, based on the observed variance in sperm concentration between the two semen donations. The ICC was higher in the SSC compared to the DSC cohort. The levels of 18 metabolites, primarily acylcarnitines, varied between the initial and subsequent donations. After subdivision into subgroups, only ornithine and phosphatidylcholine 40:5 exhibited differential levels between the two donations in the SSC group, compared to 14 metabolites in the DSC group. CVB was higher than CVW but both differed between the metabolite subclasses. Biogenic amines were identified as the least reliable analytes over time, exhibiting the highest CVW, compared to sphingomyelins, which demonstrated the highest reliability with the lowest variation. CVB was the highest for ether-bound glycerophosphatidylcholines and the lowest for amino acids.
Keywords: acylcarnitines, biogenic amines, mass spectrometry, metabolomics, semen, seminal plasma
INTRODUCTION
Seminal plasma with its distinctive composition serves not only as a source of nutrition for spermatozoa on their way to the ovum but also as a scavenger of free radicals to protect spermatozoa from oxidative stress.1 It is a mixture of secretions from the seminal vesicles, the prostate gland, the bulbo- and periurethral glands, as well as from the epididymis and the testicles.2 Its composition is species-specific.1 In order to neutralize the acidic conditions within the female reproductive tract, seminal plasma of humans contains basic amines, such as spermine, spermidine, putrescine, serotonin, alpha-aminoadipic acid (α-AAA), carnosine, creatinine, hydroxyproline, and phenethylamine. Other important components include (acyl)carnitines, which are relevant for energy production and have been linked to sperm performance, as well as vitamins, minerals, carbohydrates, lipids, and proteins.3
Metabolomic profiling of seminal plasma has become an integral part of research into male reproductive disorders over the last decade. The analysis of small, low-molecular-weight components downstream of genomics and proteomics at a specific time point allows for the rapid identification of altered biochemical processes, thereby providing a deeper understanding of the underlying molecular mechanisms of male infertility.4,5,6 Several studies have characterized the metabolic signature of seminal plasma from healthy and infertile men.3,7,8,9,10 Despite the identification of promising biomarkers for the diagnosis and treatment of male reproductive disorders, these studies are largely based a single sample per individual. However, in order to draw reliable conclusions from a single measurement, it is necessary to obtain information about the stability of a metabolic phenotype. Analytes exhibiting high intra-individual variability must be interpreted with caution, as they may introduce bias into the association between biomarkers and disease risk or simulate treatment effects.11 For the definition of meaningful reference ranges and appropriate cut-off values, it is essential to have an understanding of the biological variation of metabolite levels between and within individuals.12,13
Reports on repeated measurements of seminal plasma metabolites from different ejaculates of the same individual are only limited, and the respective studies were not designed to determine the reliability, but rather to investigate the effect of certain conditions on semen parameters and metabolite concentrations.14,15,16 In addition, the metabolome analyses in these studies were performed using nuclear magnetic resonance, a very insensitive analytical method that only detects highly abundant metabolites.
The present study was conducted to investigate the biological variability of the seminal plasma metabolome over an average interval of 8 weeks and to assess the reliability of targeted metabolite measurements by a mass spectrometric approach. Additionally, we sought to determine whether the intra-individual variation in sperm concentration affects the reproducibility of the measurements.
PARTICIPANTS AND METHODS
Ethical approval
The study was approved by the local Ethics Committee of Leipzig University (Leipzig, Germany; Approval No. 136-10-31052010) and conducted in accordance with the principles of the Declaration of Helsinki. Prior to their participation in the study, all study participants read and agreed to the study protocol, signed a written informed consent form, and confirmed that they were free of any sexually transmitted disease. All samples were processed and analyzed pseudonymously.
Study participants and sample collection
The present study is part of a larger investigation on seminal plasma metabolomics, which was conducted between June 2021 and September 2021 at the Andrology Unit of Leipzig University.
To investigate the reliability of metabolic profiles, a subset of 15 healthy participants was analyzed, having provided a second semen sample 3–12 weeks after the initial donation. Semen donors could be included in the present study if one or both samples exhibited semen parameters within the reference ranges defined by the 5th edition of World Health Organization (WHO) guidelines17 and no metabolic or cardiovascular diseases were present.
Routine semen analysis and further sample processing
Following liquefaction at 37°C for 30 min, an aliquot of the ejaculate was evaluated in accordance with the 5th WHO guidelines17 as previously described.18 The remaining semen was loaded onto a 45% and 90% discontinuous gradient of sperm separation medium (GM501 Gradient; Gynemed, Sierksdorf, Germany) and centrifuged at 600g for 20 min at room temperature (22°C) to separate the seminal plasma from sperm and other cells. The resulting supernatant consisting of seminal plasma was subsequently stored at −80°C until further analysis.
Targeted biomarker analysis
The targeted metabolomics approach was conducted on an API 5500 Q-Trap Triple Quadrupole Ion Trap (QTRAP) mass spectrometer (Sciex, Framingham, MA, USA) utilizing the AbsoluteIDQ p180 kit (Biocrates Life Sciences AG, Innsbruck, Austria). All measurements were performed in accordance with the manufacturer’s protocol as previously described.7,19 The samples were analyzed on three different 96-well plates.
Liquid chromatography (LC) was used for the separation of amino acids (AA) and biogenic amines (BA) prior to their quantification by mass spectrometry. Acylcarnitines (AC), lysophosphatidylcholines (LPC), diacyl-phosphatidylcholines (PC aa), acyl-alkyl-phosphatidylcholines (PC ae), sphingomyelins (SM), and the sum of hexoses were determined by flow injection analysis coupled with tandem mass spectrometry (FIA-MS/MS). Seven-point calibration curves were generated for AA and BA. The FIA method was based on a one-point calibration. Blanks and quality controls at different concentration levels were included on each plate. The limit of detection (LOD) was set to three times the values of the phosphate-buffered saline (PBS)-only-containing samples. Metabolite concentrations were calculated in µmol l−1.
Data preprocessing
Values below the LOD were replaced with the batch-specific LOD/2 for that particular metabolite (in order to avoid “division by zero errors”). Metabolites with more than 25% missing values were excluded from subsequent statistical analyses. To address the possibility of batch effects, the data underwent scaling using the 25%-trimmed mean plate-wise, a procedure suggested previously.20
Statistical analyses
The remaining 96 metabolites were analyzed for changes in concentration across both visits using the Wilcoxon signed-rank test. Additionally, we calculated Spearman’s rank correlation coefficients between spermiogram parameters and metabolite concentrations. The Mann–Whitney U test was used to test for differences in age, body mass index (BMI), and time between samplings between the both subgroups.
The reliability was expressed as the intraclass correlation coefficient (ICC), which was determined using a one-way analysis of variance (ANOVA) random-effects model on log-transformed data, as ICC(1) = (MSBS – MSWS)/(MSBS + [k-1] × MSWS), with MSBS as mean square between subjects, MSWS representing mean square within subjects and k as the number of measurements.21,22 For negative values, the ICC was set to zero.22 ICC values were evaluated as the following: ICC < 0.40 as poor reliability, 0.40 < ICC ≤ 0.50 as fair reliability, 0.51 ≤ ICC ≤ 0.74 as good reliability, and ICC > 0.74 as excellent reliability.23,24,25
To investigate the stability of the metabolomic phenotype, we calculated the coefficients of variation within individuals (CVW) and coefficients of variation between individuals (CVB) for each metabolite. These coefficients demonstrate the extent of variability relative to the mean. Low values indicate a good stability. Metabolites with a CVW less than 0.20 were assumed to be reasonably stable over time.
Furthermore, we investigated the impact of intra-individual variability in sperm concentration on the reliability of metabolite measurements. The data were divided into two groups: men with sperm concentrations that varied by less than 25% between the two semen donations were categorized as “similar sperm concentration (SSC)”, while the second group comprised men with sperm concentration variability greater than 25% (differing sperm concentration [DSC]). ICC, CVW, and CVB were calculated for each group.
All statistical analyses were performed using SPSS 29.0 (IBM Corp., Armonk, NY, USA) and Excel 2016 (Microsoft Corp., Redmond, WA, USA). The level of statistical significance was set to P < 0.05. Graphs were created using GraphPad Prism 10 (GraphPad Software, San Diego, CA, USA).
RESULTS
Study population
The seminal plasma of 15 healthy participants who provided a second semen sample between 3–12 weeks after the initial donation was subjected to analysis. Semen samples were obtained by masturbation and collected in sterile plastic containers after a period of sexual abstinence of 2–7 days. The recorded demographic characteristics are shown in Table 1. There were no differences in age (P = 0.439), BMI (P = 0.836), and time between the two semen donations (P = 0.511) between the SSC and the DSC groups. The results of the routine semen analysis were comparable between the two groups (Table 2).
Table 1.
Characteristics of study participants
| Characteristic | All (n=15) | SSC (n=6) | DSC (n=9) |
|---|---|---|---|
| Age (year), mean±s.d. | 26.8±5.2 | 25.5±4.6 | 27.7±10.2 |
| BMI (kg m−2), mean±s.d. | 23.3±2.3 | 23.3±2.7 | 23.3±7.6 |
| Semen donation interval (week), mean±s.d. | 8.67±3.66 | 9.52±3.87 | 8.10±3.63 |
| Smoker (n) | 1 | 0 | 1 |
| Caucasian (n) | 15 | 6 | 9 |
SSC: men with similar intra-individual sperm concentration in both samples; DSC: men with different intra-individual sperm concentration in both samples; BMI: body mass index; s.d.: standard deviation
Table 2.
Semen parameters of ejaculates donated on average 8 weeks apart
| Semen parameter | Ejaculate 1 | Ejaculate 2 | ||||||
|---|---|---|---|---|---|---|---|---|
|
|
|
|||||||
| All (n=15) | SSC (n=6) | DSC (n=9) | aP | All (n=15) | SSC (n=6) | DSC (n=9) | aP | |
| Sperm concentration (×106 ml-1) | 89.5±53.4 | 115.7±45.5 | 72.0±53.3 | 0.083 | 80.0±44.4 | 103.3±35.8 | 64.5±44.5 | 0.094 |
| Semen volume (ml) | 3.50±1.26 | 3.30±1.35 | 3.63±1.27 | 0.548 | 3.38±1.23 | 3.26±1.13 | 3.46±1.34 | 0.931 |
| Total sperm count (×106) | 294±182 | 371±201 | 244±160 | 0.181 | 273±171 | 335±164 | 231±171 | 0.328 |
| Progressive motility (%) | 61.6±17.4 | 65.8±17.2 | 58.7±18.0 | 0.372 | 53.7±13.9 | 64.6±15.0 | 68.0±12.9 | 0.979 |
aStatistical evaluation between SSC and DSC. Data are given as mean±s.d. SSC: men with similar intra-individual sperm concentration in both samples; DSC: men with different intra-individual sperm concentration in both samples; s.d.: standard deviation
Correlations of spermiogram parameters with the seminal plasma metabolome
The majority of metabolite concentrations (64.6%) correlated with sperm concentration. It is noteworthy that almost all lipids showed a positive correlation with sperm concentration, whereas this was only true for very few AA, BA, and AC. Progressive motility was negatively correlated with only four metabolites and the semen volume only with two metabolites (Supplementary Table 1).
Supplementary Table 1.
| Class of Metabolites | Progressive motility (%) | Concentration (10^6/ml) | Volume (ml) | Total count (10^6) | |
|---|---|---|---|---|---|
| Amino Acids | Ala | -0.321 | 0.273 | -0.163 | 0.128 |
| Arg | -0.200 | 0.070 | 0.092 | 0.168 | |
| Asn | -0.271 | 0.329 | -0.070 | 0.183 | |
| Asp | -0.347 | 0.439* | -0.132 | 0.311 | |
| Gln | -0.424* | -0.182 | 0.148 | -0.050 | |
| Glu | -0.190 | 0.417* | -0.135 | 0.302 | |
| Gly | -0.383 | 0.069 | -0.033 | 0.129 | |
| His | -0.353 | 0.052 | -0.046 | 0.080 | |
| Ile | -0.336 | -0.067 | 0.213 | 0.143 | |
| Leu | -0.402* | -0.009 | 0.187 | 0.167 | |
| Lys | -0.362 | 0.066 | -0.101 | 0.015 | |
| Met | -0.436* | 0.228 | -0.247 | 0.053 | |
| Orn | -0.024 | 0.356 | -0.176 | 0.203 | |
| Pro | -0.180 | 0.504** | -0.081 | 0.409* | |
| Thr | -0.357 | 0.127 | 0.041 | 0.164 | |
| Tyr | -0.344 | -0.132 | 0.212 | 0.074 | |
| Val | -0.374 | 0.071 | -0.009 | 0.089 | |
| Biogenic Amines | α-AAA | -0.607** | 0.193 | -0.207 | 0.015 |
| Carnosine | -0.228 | 0.004 | 0.169 | 0.169 | |
| Creatinine | -0.053 | 0.521** | -0.438* | 0.069 | |
| DOPA | 0.024 | -0.051 | 0.251 | 0.075 | |
| Taurine | 0.314 | -0.139 | 0.289 | 0.018 | |
| total DMA | -0.004 | 0.247 | -0.394* | -0.164 | |
| Acyl Carnitines | C0 | -0.235 | 0.385 | -0.214 | 0.222 |
| C2 | 0.180 | 0.188 | 0.097 | 0.330 | |
| C3 | 0.077 | 0.609** | -0.049 | 0.565** | |
| C3-DC (C4-OH) | 0.211 | 0.345 | 0.111 | 0.444* | |
| C4 | 0.224 | 0.322 | -0.097 | 0.296 | |
| C5 | 0.053 | 0.430* | -0.071 | 0.385* | |
| C5-DC (C6-OH) | 0.237 | 0.483** | 0.030 | 0.458* | |
| C5-OH (C3-DC-M | 0.081 | 0.541** | 0.055 | 0.520** | |
| C5:1-DC | -0.059 | 0.062 | -0.226 | -0.050 | |
| C6 (C4:1-DC) | 0.155 | 0.223 | 0.047 | 0.315 | |
| C6:1 | 0.303 | 0.187 | 0.162 | 0.246 | |
| C7-DC | 0.240 | 0.146 | 0.019 | 0.220 | |
| C16 | 0.079 | 0.350 | -0.111 | 0.252 | |
| Lyso-PC | lysoPC a C16:0 | -0.153 | 0.272 | -0.125 | 0.096 |
| lysoPC a C17:0 | 0.001 | 0.131 | -0.095 | -0.025 | |
| lysoPC a C18:0 | -0.137 | 0.405* | -0.174 | 0.184 | |
| lysoPC a C18:1 | -0.121 | 0.345 | -0.085 | 0.176 | |
| Acyl-acyl-PC | PC aa C30:0 | -0.302 | 0.521** | -0.092 | 0.354 |
| PC aa C30:2 | -0.135 | 0.262 | 0.099 | 0.234 | |
| PC aa C32:0 | -0.295 | 0.545** | 0.105 | 0.456* | |
| PC aa C32:1 | -0.149 | 0.540** | -0.070 | 0.416* | |
| PC aa C32:2 | -0.241 | 0.538** | 0.088 | 0.465** | |
| PC aa C34:1 | -0.154 | 0.506** | -0.084 | 0.326 | |
| PC aa C34:2 | -0.117 | 0.617** | 0.067 | 0.527** | |
| PC aa C36:1 | -0.154 | 0.558** | -0.133 | 0.345 | |
| PC aa C36:2 | -0.182 | 0.572** | 0.046 | 0.465** | |
| PC aa C36:3 | -0.251 | 0.696** | 0.034 | 0.562** | |
| PC aa C36:4 | -0.139 | 0.458* | -0.136 | 0.290 | |
| PC aa C36:6 | -0.128 | 0.536** | -0.140 | 0.367* | |
| PC aa C38:0 | -0.193 | 0.691** | -0.204 | 0.443* | |
| PC aa C38:3 | -0.166 | 0.680** | -0.061 | 0.500** | |
| PC aa C38:4 | -0.094 | 0.598** | -0.098 | 0.436* | |
| PC aa C38:5 | -0.040 | 0.429* | 0.113 | 0.370* | |
| PC aa C38:6 | -0.128 | 0.561** | -0.018 | 0.457* | |
| PC aa C40:3 | -0.082 | 0.606** | -0.051 | 0.472** | |
| PC aa C40:4 | -0.060 | 0.742** | -0.133 | 0.543** | |
| PC aa C40:5 | 0.051 | 0.540** | -0.083 | 0.363* | |
| PC aa C40:6 | -0.041 | 0.562** | -0.113 | 0.417* | |
| PC aa C42:1 | -0.239 | 0.561** | -0.306 | 0.263 | |
| PC aa C42:4 | -0.191 | 0.515** | 0.031 | 0.385* | |
| PC aa C42:5 | 0.048 | 0.607** | -0.052 | 0.536** | |
| Acyl-alkyl-PC | PC ae C30:2 | -0.285 | 0.296 | 0.053 | 0.258 |
| PC ae C32:1 | -0.203 | 0.683** | 0.023 | 0.624** | |
| PC ae C34:0 | -0.067 | 0.243 | -0.146 | 0.067 | |
| PC ae C34:1 | -0.134 | 0.684** | -0.155 | 0.470** | |
| PC ae C34:2 | -0.173 | 0.707** | 0.024 | .567** | |
| PC ae C34:3 | 0.035 | 0.594** | 0.026 | .468** | |
| PC ae C36:2 | -0.136 | 0.707** | -0.045 | .540** | |
| PC ae C36:3 | 0.061 | 0.640** | -0.051 | .487** | |
| PC ae C36:4 | -0.159 | 0.593** | -0.072 | 0.438* | |
| PC ae C38:2 | 0.083 | 0.398* | -0.063 | 0.281 | |
| PC ae C38:3 | -0.185 | 0.579** | -0.101 | 0.405* | |
| PC ae C38:4 | -0.125 | 0.646** | -0.228 | 0.422* | |
| PC ae C38:5 | -0.078 | 0.471** | -0.177 | 0.257 | |
| PC ae C38:6 | -0.158 | 0.585** | -0.147 | 0.388* | |
| PC ae C40:2 | -0.208 | 0.608** | -0.133 | 0.393* | |
| PC ae C40:3 | -0.055 | 0.686** | -0.122 | 0.484** | |
| PC ae C40:5 | 0.148 | 0.497** | -0.205 | 0.249 | |
| PC ae C40:6 | -0.052 | 0.686** | -0.229 | 0.441* | |
| PC ae C42:3 | -0.201 | 0.553** | -0.104 | 0.385* | |
| Spingomyelins | SM (OH) C14:1 | -0.129 | 0.490** | -0.248 | 0.216 |
| SM (OH) C16:1 | -0.230 | 0.525** | -0.229 | 0.261 | |
| SM (OH) C22:1 | -0.105 | 0.509** | -0.296 | 0.202 | |
| SM (OH) C22:2 | -0.221 | 0.659** | -0.247 | 0.364* | |
| SM (OH) C24:1 | -0.128 | 0.545** | -0.302 | 0.245 | |
| SM C16:0 | -0.199 | 0.459* | -0.159 | 0.232 | |
| SM C16:1 | -0.303 | 0.676** | -0.155 | 0.434* | |
| SM C18:0 | -0.205 | 0.538** | -0.100 | 0.329 | |
| SM C18:1 | -0.244 | 0.696** | -0.106 | 0.505** | |
| SM C24:0 | -0.146 | 0.567** | -0.260 | 0.281 | |
| SM C24:1 | -0.191 | 0.608** | -0.199 | 0.351 | |
| SM C26:0 | -0.013 | 0.691** | -0.067 | 0.544** | |
| SM C26:1 | -0.104 | 0.754** | -0.127 | 0.551** |
*The correlation is significant at the 0.05 level (two-tailed)
**The correlation is significant at the 0.01 level (two-tailed).
Visit-to-visit variability
The concentration between the initial and subsequent measurements differed for 18 metabolites (Supplementary Table 2). The majority of these (eight metabolites) belonged to the AC group. A comparison by subgroups revealed a change between measurements for only two metabolites in the SSC group, whereas this was true for 14 metabolites in the DSC group (Supplementary Table 3).
Supplementary Table 2.
| Class of Metabolites | Metabolite | 1st Measurement | 2nd Measurement | Reliability | ||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Mean conc. (µM) | SD | Mean conc. (µM) | SD | p value | CVW | CVB | ICC | CI 95% | ||
| Amino Acids | Ala | 730.35 | 421.37 | 891.37 | 427.58 | 0.127 | 0.28 | 0.48 | 0.58 | 0.09-0.85 |
| Arg | 805.90 | 82.85 | 838.83 | 80.71 | 0.305 | 0.08 | 0.08 | 0.18 | 0-0.65 | |
| Asn | 1747.57 | 755.98 | 2367.62 | 911.66 | 0.074 | 0.29 | 0.36 | 0.48 | 0-0.85 | |
| Asp | 1151.43 | 597.48 | 1175.28 | 377.18 | 0.735 | 0.28 | 0.33 | 0.39 | 0-0.76 | |
| Gln | 3197.31 | 1267.42 | 3185.74 | 657.24 | 0.946 | 0.19 | 0.27 | 0.46 | 0-0.8 | |
| Glu | 1972.25 | 611.45 | 1978.65 | 517.09 | 0.893 | 0.16 | 0.25 | 0.58 | 0.09-0.85 | |
| Gly | 2692.03 | 1104.92 | 2736.12 | 595.79 | 0.787 | 0.23 | 0.27 | 0.31 | 0-0.72 | |
| His | 1071.60 | 263.95 | 1130.37 | 206.61 | 0.414 | 0.14 | 0.18 | 0.36 | 0-0.75 | |
| Ile | 1336.92 | 272.05 | 1307.10 | 243.79 | 0.735 | 0.09 | 0.18 | 0.69 | 0.27-0.89 | |
| Leu | 2199.88 | 551.47 | 2316.57 | 517.92 | 0.455 | 0.15 | 0.20 | 0.51 | 0-0.82 | |
| Lys | 1874.12 | 618.07 | 1916.23 | 425.08 | 0.893 | 0.16 | 0.24 | 0.54 | 0.03-0.83 | |
| Met | 26.33 | 17.30 | 31.52 | 17.17 | 0.376 | 0.35 | 0.49 | 0.48 | 0-0.8 | |
| Orn | 49.33 | 34.25 | 36.78 | 25.09 | 0.013 | 0.29 | 0.68 | 0.74 | 0.36-0.91 | |
| Pro | 286.77 | 126.54 | 301.05 | 124.70 | 0.635 | 0.27 | 0.35 | 0.46 | 0-0.8 | |
| Thr | 1346.48 | 376.94 | 1432.69 | 397.97 | 0.340 | 0.15 | 0.25 | 0.64 | 0.18-0.87 | |
| Tyr | 1583.43 | 338.05 | 1571.46 | 234.84 | 0.839 | 0.10 | 0.16 | 0.65 | 0.2-0.88 | |
| Val | 1399.52 | 488.21 | 1463.38 | 350.23 | 0.635 | 0.17 | 0.26 | 0.55 | 0.04-0.84 | |
| Mean | 0.20 | 0.30 | 0.51 | |||||||
| Median | 0.17 | 0.26 | 0.51 | |||||||
| Biogenic Amines | α-AAA | 3.11 | 3.94 | 2.96 | 0.78 | 0.685 | 0.69 | 0.70 | 0.00 | 0-0.52 |
| Carnosine | 5.06 | 2.41 | 6.38 | 2.78 | 0.033 | 0.26 | 0.42 | 0.61 | 0.14-0.86 | |
| Creatinine | 230.78 | 128.35 | 223.82 | 128.13 | 0.376 | 0.12 | 0.56 | 0.86 | 0.62-0.95 | |
| DOPA | 0.54 | 0.26 | 0.43 | 0.13 | 0.376 | 0.30 | 0.31 | 0.00 | 0-0.48 | |
| Taurine | 363.16 | 60.43 | 341.45 | 75.89 | 0.244 | 0.07 | 0.18 | 0.69 | 0.26-0.89 | |
| total DMA | 3.84 | 1.66 | 6.52 | 7.37 | 0.455 | 0.72 | 0.75 | 0.00 | 0-0.53 | |
| Mean | 0.36 | 0.49 | 0.36 | |||||||
| Median | 0.28 | 0.49 | 0.31 | |||||||
| Acyl Carnitines | C0 | 259.93 | 139.30 | 221.25 | 92.82 | 0.519 | 0.25 | 0.38 | 0.38 | 0-0.75 |
| C2 | 202.83 | 97.11 | 169.62 | 62.57 | 0.277 | 0.26 | 0.38 | 0.46 | 0-0.78 | |
| C3 | 13.18 | 7.13 | 12.68 | 6.74 | 0.552 | 0.19 | 0.50 | 0.63 | 0.21-0.86 | |
| C3-DC (C4-OH) | 3.81 | 1.44 | 3.12 | 0.93 | 0.041 | 0.20 | 0.32 | 0.49 | 0-0.79 | |
| C4 | 53.04 | 25.51 | 42.19 | 21.81 | 0.095 | 0.35 | 0.44 | 0.44 | 0-0.77 | |
| C5 | 9.38 | 5.19 | 7.61 | 4.16 | 0.049 | 0.23 | 0.50 | 0.58 | 0.12-0.84 | |
| C5-DC (C6-OH) | 1.18 | 0.52 | 1.06 | 0.28 | 0.326 | 0.27 | 0.31 | 0.32 | 0-0.71 | |
| C5-OH (C3-DC-M) | 1.44 | 0.60 | 1.20 | 0.36 | 0.091 | 0.22 | 0.33 | 0.46 | 0-0.78 | |
| C5:1-DC | 1.91 | 3.22 | 0.58 | 0.36 | 0.002 | 0.43 | 1.42 | 0.46 | 0-0.78 | |
| C6 (C4:1-DC) | 2.10 | 0.51 | 1.73 | 0.47 | 0.015 | 0.21 | 0.22 | 0.25 | 0-0.66 | |
| C6:1 | 0.35 | 0.15 | 0.18 | 0.03 | 0.002 | 0.49 | 0.29 | 0.00 | 0-0.21 | |
| C7-DC | 0.54 | 0.21 | 0.40 | 0.13 | 0.007 | 0.27 | 0.32 | 0.40 | 0-0.75 | |
| C16 | 0.31 | 0.25 | 0.27 | 0.09 | 0.183 | 0.28 | 0.43 | 0.00 | 0-0.43 | |
| Mean | 0.28 | 0.45 | 0.38 | |||||||
| Median | 0.26 | 0.38 | 0.44 | |||||||
| Lyso-PC | lysoPC a C16:0 | 17.15 | 9.94 | 17.11 | 10.27 | 0.762 | 0.21 | 0.54 | 0.71 | 0.35-0.89 |
| lysoPC a C17:0 | 0.36 | 0.17 | 0.26 | 0.16 | 0.095 | 0.36 | 0.45 | 0.15 | 0-0.6 | |
| lysoPC a C18:0 | 3.68 | 1.75 | 3.58 | 1.69 | 0.720 | 0.18 | 0.43 | 0.68 | 0.28-0.88 | |
| lysoPC a C18:1 | 1.82 | 0.94 | 2.04 | 1.25 | 0.121 | 0.23 | 0.53 | 0.71 | 0.34-0.89 | |
| Mean | 23.01 | 22.99 | 0.25 | 0.49 | 0.56 | |||||
| Median | 0.22 | 0.49 | 0.69 | |||||||
| Acyl-acyl-PC | PC aa C30:0 | 1.08 | 0.55 | 0.93 | 0.37 | 0.303 | 0.21 | 0.41 | 0.61 | 0.18-0.85 |
| PC aa C30:2 | 0.57 | 0.44 | 0.40 | 0.19 | 0.169 | 0.39 | 0.56 | 0.25 | 0-0.66 | |
| PC aa C32:0 | 2.62 | 0.91 | 2.38 | 0.69 | 0.277 | 0.14 | 0.30 | 0.69 | 0.31-0.88 | |
| PC aa C32:1 | 0.48 | 0.15 | 0.43 | 0.15 | 0.359 | 0.24 | 0.26 | 0.17 | 0-0.61 | |
| PC aa C32:2 | 0.16 | 0.09 | 0.10 | 0.04 | 0.035 | 0.35 | 0.41 | 0.11 | 0-0.57 | |
| PC aa C34:1 | 16.53 | 8.21 | 15.29 | 7.36 | 0.804 | 0.20 | 0.45 | 0.67 | 0.28-0.88 | |
| PC aa C34:2 | 3.06 | 1.17 | 2.65 | 1.18 | 0.135 | 0.22 | 0.37 | 0.61 | 0.17-0.85 | |
| PC aa C36:1 | 7.32 | 3.91 | 6.79 | 3.73 | 0.639 | 0.20 | 0.50 | 0.66 | 0.26-0.87 | |
| PC aa C36:2 | 3.24 | 1.46 | 2.93 | 1.24 | 0.489 | 0.18 | 0.41 | 0.69 | 0.31-0.88 | |
| PC aa C36:3 | 1.75 | 0.50 | 1.71 | 0.77 | 0.762 | 0.18 | 0.34 | 0.65 | 0.24-0.87 | |
| PC aa C36:4 | 0.47 | 0.20 | 0.40 | 0.17 | 0.107 | 0.22 | 0.39 | 0.63 | 0.21-0.86 | |
| PC aa C36:6 | 0.05 | 0.02 | 0.04 | 0.02 | 0.095 | 0.30 | 0.39 | 0.45 | 0-0.77 | |
| PC aa C38:0 | 0.24 | 0.10 | 0.22 | 0.11 | 0.359 | 0.20 | 0.43 | 0.74 | 0.4-0.9 | |
| PC aa C38:3 | 1.24 | 0.49 | 1.14 | 0.53 | 0.303 | 0.16 | 0.40 | 0.78 | 0.47-0.92 | |
| PC aa C38:4 | 0.47 | 0.15 | 0.45 | 0.19 | 0.720 | 0.16 | 0.35 | 0.69 | 0.3-0.88 | |
| PC aa C38:5 | 0.25 | 0.08 | 0.18 | 0.06 | 0.005 | 0.25 | 0.29 | 0.27 | 0-0.68 | |
| PC aa C38:6 | 1.04 | 0.34 | 0.97 | 0.53 | 0.561 | 0.26 | 0.40 | 0.54 | 0.08-0.82 | |
| PC aa C40:3 | 0.10 | 0.04 | 0.09 | 0.03 | 0.252 | 0.24 | 0.34 | 0.53 | 0.06-0.81 | |
| PC aa C40:4 | 0.11 | 0.05 | 0.09 | 0.03 | 0.107 | 0.18 | 0.37 | 0.59 | 0.15-0.84 | |
| PC aa C40:5 | 0.12 | 0.05 | 0.09 | 0.04 | 0.002 | 0.19 | 0.37 | 0.65 | 0.24-0.87 | |
| PC aa C40:6 | 0.53 | 0.20 | 0.48 | 0.26 | 0.188 | 0.20 | 0.44 | 0.70 | 0.33-0.89 | |
| PC aa C42:1 | 0.05 | 0.01 | 0.06 | 0.01 | 0.489 | 0.09 | 0.22 | 0.88 | 0.69-0.96 | |
| PC aa C42:4 | 0.08 | 0.04 | 0.05 | 0.02 | 0.010 | 0.39 | 0.39 | 0.20 | 0-0.63 | |
| PC aa C42:5 | 0.06 | 0.02 | 0.05 | 0.02 | 0.081 | 0.20 | 0.32 | 0.46 | 0-0.78 | |
| Mean | 41.62 | 37.92 | 0.22 | 0.38 | 0.55 | |||||
| Median | 0.20 | 0.39 | 0.62 | |||||||
| Acyl-alkyl-PC | PC ae C30:2 | 0.04 | 0.02 | 0.03 | 0.01 | 0.095 | 0.45 | 0.36 | 0.00 | 0-0.39 |
| PC ae C32:1 | 0.24 | 0.09 | 0.23 | 0.12 | 0.330 | 0.19 | 0.41 | 0.75 | 0.42-0.91 | |
| PC ae C34:0 | 0.30 | 0.13 | 0.25 | 0.16 | 0.035 | 0.20 | 0.51 | 0.67 | 0.28-0.88 | |
| PC ae C34:1 | 0.57 | 0.23 | 0.53 | 0.25 | 0.524 | 0.15 | 0.41 | 0.73 | 0.38-0.9 | |
| PC ae C34:2 | 0.28 | 0.12 | 0.25 | 0.13 | 0.330 | 0.20 | 0.44 | 0.68 | 0.29-0.88 | |
| PC ae C34:3 | 0.06 | 0.03 | 0.05 | 0.02 | 0.135 | 0.17 | 0.39 | 0.76 | 0.43-0.91 | |
| PC ae C36:2 | 0.20 | 0.09 | 0.17 | 0.10 | 0.064 | 0.18 | 0.53 | 0.77 | 0.45-0.91 | |
| PC ae C36:3 | 0.11 | 0.07 | 0.09 | 0.05 | 0.055 | 0.26 | 0.57 | 0.64 | 0.22-0.86 | |
| PC ae C36:4 | 0.08 | 0.04 | 0.08 | 0.05 | 0.934 | 0.19 | 0.52 | 0.70 | 0.32-0.89 | |
| PC ae C38:2 | 0.14 | 0.07 | 0.12 | 0.07 | 0.073 | 0.22 | 0.53 | 0.59 | 0.14-0.84 | |
| PC ae C38:3 | 0.10 | 0.05 | 0.08 | 0.05 | 0.135 | 0.22 | 0.48 | 0.62 | 0.2-0.85 | |
| PC ae C38:4 | 0.09 | 0.05 | 0.07 | 0.04 | 0.252 | 0.24 | 0.55 | 0.64 | 0.23-0.86 | |
| PC ae C38:5 | 0.11 | 0.07 | 0.10 | 0.06 | 0.135 | 0.33 | 0.56 | 0.50 | 0.02-0.8 | |
| PC ae C38:6 | 0.84 | 0.49 | 0.88 | 0.73 | 0.890 | 0.20 | 0.68 | 0.84 | 0.6-0.94 | |
| PC ae C40:2 | 0.11 | 0.04 | 0.09 | 0.05 | 0.007 | 0.22 | 0.43 | 0.54 | 0.07-0.81 | |
| PC ae C40:3 | 0.08 | 0.05 | 0.05 | 0.02 | 0.012 | 0.34 | 0.47 | 0.30 | 0-0.69 | |
| PC ae C40:5 | 0.05 | 0.03 | 0.04 | 0.02 | 0.025 | 0.21 | 0.51 | 0.67 | 0.28-0.88 | |
| PC ae C40:6 | 0.11 | 0.08 | 0.09 | 0.05 | 0.135 | 0.19 | 0.61 | 0.80 | 0.52-0.93 | |
| PC ae C42:3 | 0.16 | 0.11 | 0.09 | 0.02 | 0.055 | 0.33 | 0.45 | 0.11 | 0-0.57 | |
| Mean | 3.68 | 3.30 | 0.24 | 0.49 | 0.60 | |||||
| Median | 0.21 | 0.51 | 0.67 | |||||||
| Spingomyelins | SM (OH) C14:1 | 1.37 | 0.53 | 1.28 | 0.63 | 0.489 | 0.18 | 0.41 | 0.69 | 0.3-0.88 |
| SM (OH) C16:1 | 0.86 | 0.34 | 0.78 | 0.40 | 0.330 | 0.18 | 0.43 | 0.69 | 0.3-0.88 | |
| SM (OH) C22:1 | 1.63 | 0.79 | 1.45 | 0.90 | 0.135 | 0.15 | 0.53 | 0.72 | 0.37-0.9 | |
| SM (OH) C22:2 | 0.47 | 0.20 | 0.41 | 0.20 | 0.135 | 0.14 | 0.44 | 0.79 | 0.49-0.92 | |
| SM (OH) C24:1 | 0.41 | 0.19 | 0.35 | 0.19 | 0.015 | 0.16 | 0.48 | 0.74 | 0.4-0.91 | |
| SM C16:0 | 68.26 | 30.73 | 62.91 | 25.29 | 0.561 | 0.21 | 0.38 | 0.58 | 0.14-0.84 | |
| SM C16:1 | 1.84 | 0.65 | 1.71 | 0.75 | 0.454 | 0.18 | 0.37 | 0.70 | 0.32-0.89 | |
| SM C18:0 | 8.42 | 3.55 | 7.74 | 3.08 | 0.489 | 0.18 | 0.38 | 0.65 | 0.24-0.87 | |
| SM C18:1 | 0.92 | 0.29 | 0.84 | 0.35 | 0.359 | 0.17 | 0.34 | 0.64 | 0.22-0.86 | |
| SM C24:0 | 13.83 | 7.13 | 13.12 | 7.04 | 0.978 | 0.18 | 0.50 | 0.70 | 0.32-0.89 | |
| SM C24:1 | 8.11 | 3.96 | 7.64 | 3.96 | 0.804 | 0.17 | 0.48 | 0.77 | 0.45-0.91 | |
| SM C26:0 | 0.52 | 0.20 | 0.48 | 0.19 | 0.303 | 0.18 | 0.37 | 0.73 | 0.37-0.9 | |
| SM C26:1 | 0.52 | 0.21 | 0.44 | 0.21 | 0.048 | 0.18 | 0.42 | 0.77 | 0.46-0.92 | |
| Mean | 107.144 | 99.160 | 0.17 | 0.42 | 0.71 | |||||
| Median | 0.18 | 0.42 | 0.70 | |||||||
| Mean total | 0.2321 | 0.4142 | 0.5376 | |||||||
| Median total | 0.2041 | 0.4061 | 0.6131 | |||||||
Supplementary Table 3.
| Class of Metabolites | Metabolite | SSC (similar sperm concentration) | DSC (differing sperm concentration) | |||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1st Measurement | 2nd Measurement | Reliability | 1 st Measurement | 2nd Measurement | Reliability | |||||||||||||
| Mean conc. (µM) | SD | Mean conc. (µM) | SD | p value | CVW | CVB | ICC | Mean conc. (µM) | SD | Mean conc. (µM) | SD | p value | CVW | CVB | ICC | |||
| Amino Acids | Ala | 828.61 | 495.00 | 890.61 | 260.11 | 0.563 | 0.26 | 0.42 | 0.62 | 646.12 | 364.67 | 892.02 | 556.12 | 0.109 | 0.30 | 0.56 | 0.59 | |
| Arg | 853.69 | 73.89 | 846.16 | 77.14 | 0.844 | 0.07 | 0.06 | 0.00 | 764.95 | 70.29 | 832.55 | 89.28 | 0.078 | 0.08 | 0.08 | 0.21 | ||
| Asn | 1968.26 | 862.16 | 2933.78 | 590.81 | 0.125 | 0.32 | 0.27 | 0.10 | 1471.71 | 591.84 | 1659.92 | 740.89 | 0.875 | 0.25 | 0.37 | 0.58 | ||
| Asp | 1342.47 | 703.76 | 1111.80 | 336.97 | 0.844 | 0.25 | 0.35 | 0.54 | 987.68 | 483.18 | 1229.70 | 427.10 | 0.297 | 0.31 | 0.33 | 0.29 | ||
| Gln | 3315.99 | 1176.34 | 2969.22 | 512.20 | 0.563 | 0.19 | 0.24 | 0.49 | 3095.59 | 1425.98 | 3371.33 | 747.15 | 0.469 | 0.20 | 0.31 | 0.49 | ||
| Glu | 2169.27 | 611.52 | 1971.72 | 417.05 | 0.688 | 0.16 | 0.21 | 0.47 | 1803.37 | 603.35 | 1984.60 | 624.29 | 0.469 | 0.15 | 0.30 | 0.66 | ||
| Gly | 2725.81 | 1127.83 | 2640.51 | 224.68 | 1.000 | 0.20 | 0.23 | 0.27 | 2663.07 | 1174.55 | 2818.07 | 806.78 | 0.688 | 0.25 | 0.32 | 0.38 | ||
| His | 1134.40 | 279.57 | 1094.41 | 66.15 | 0.844 | 0.14 | 0.13 | 0.04 | 1017.78 | 258.62 | 1161.19 | 281.65 | 0.219 | 0.15 | 0.22 | 0.52 | ||
| Ile | 1382.46 | 302.64 | 1297.02 | 192.00 | 0.563 | 0.08 | 0.17 | 0.77 | 1297.89 | 260.47 | 1315.75 | 296.59 | 0.938 | 0.09 | 0.19 | 0.67 | ||
| Leu | 2290.49 | 609.20 | 2288.05 | 304.49 | 1.000 | 0.12 | 0.18 | 0.55 | 2122.21 | 532.66 | 2341.02 | 676.54 | 0.375 | 0.17 | 0.24 | 0.52 | ||
| Lys | 2030.30 | 629.51 | 1825.09 | 132.05 | 0.563 | 0.13 | 0.19 | 0.42 | 1740.25 | 623.29 | 1994.35 | 575.70 | 0.219 | 0.19 | 0.29 | 0.61 | ||
| Met | 27.46 | 18.19 | 31.22 | 12.74 | 0.688 | 0.34 | 0.43 | 0.45 | 25.37 | 17.90 | 31.78 | 21.30 | 0.578 | 0.36 | 0.58 | 0.54 | ||
| Orn | 47.95 | 30.07 | 31.48 | 22.58 | 0.031 | 0.33 | 0.66 | 0.67 | 50.51 | 39.86 | 41.32 | 27.95 | 0.297 | 0.26 | 0.72 | 0.83 | ||
| Pro | 336.03 | 159.31 | 302.00 | 137.71 | 0.844 | 0.28 | 0.38 | 0.61 | 244.55 | 79.81 | 300.23 | 123.67 | 0.297 | 0.27 | 0.32 | 0.27 | ||
| Thr | 1468.83 | 408.67 | 1477.85 | 194.77 | 0.844 | 0.13 | 0.17 | 0.39 | 1241.60 | 342.33 | 1393.98 | 530.44 | 0.297 | 0.16 | 0.32 | 0.72 | ||
| Tyr | 1632.81 | 383.05 | 1545.19 | 229.81 | 0.438 | 0.11 | 0.18 | 0.72 | 1541.12 | 318.99 | 1593.97 | 254.97 | 0.578 | 0.09 | 0.16 | 0.62 | ||
| Val | 1498.42 | 495.84 | 1445.44 | 172.18 | 0.688 | 0.17 | 0.19 | 0.27 | 1314.75 | 503.65 | 1478.76 | 469.06 | 0.219 | 0.16 | 0.33 | 0.67 | ||
| Mean | 0.19 | 0.26 | 0.43 | 0.20 | 0.33 | 0.54 | ||||||||||||
| Median | 0.17 | 0.21 | 0.47 | 0.19 | 0.32 | 0.58 | ||||||||||||
| Biogenic Amines | a-AAA | 3.56 | 5.01 | 3.20 | 0.61 | 0.688 | 0.66 | 0.79 | 0.02 | 2.73 | 3.13 | 2.76 | 0.91 | 0.938 | 0.72 | 0.63 | 0.03 | |
| Carnosine | 5.70 | 2.73 | 6.86 | 2.71 | 0.313 | 0.22 | 0.40 | 0.71 | 4.52 | 2.15 | 5.98 | 2.98 | 0.109 | 0.30 | 0.46 | 0.56 | ||
| Creatinine | 265.44 | 165.34 | 256.11 | 158.89 | 0.438 | 0.14 | 0.61 | 0.85 | 201.06 | 89.07 | 196.13 | 99.28 | 0.938 | 0.11 | 0.47 | 0.87 | ||
| DOPA | 0.59 | 0.37 | 0.41 | 0.12 | 0.438 | 0.36 | 0.41 | 0.07 | 0.49 | 0.14 | 0.44 | 0.14 | 0.813 | 0.26 | 0.20 | 0.00 | ||
| Taurine | 401.71 | 40.49 | 382.14 | 75.36 | 0.688 | 0.07 | 0.14 | 0.54 | 330.12 | 56.38 | 306.57 | 60.92 | 0.375 | 0.08 | 0.17 | 0.61 | ||
| total DMA | 4.58 | 1.75 | 6.85 | 7.95 | 0.563 | 0.61 | 0.81 | 0.37 | 3.21 | 1.39 | 6.23 | 7.48 | 0.688 | 0.82 | 0.72 | 0.00 | ||
| Mean | 0.34 | 0.53 | 0.43 | 0.38 | 0.44 | 0.35 | ||||||||||||
| Median | 0.29 | 0.51 | 0.46 | 0.28 | 0.46 | 0.30 | ||||||||||||
| Acyl Carnitines | C0 | 257.83 | 127.23 | 244.60 | 29.57 | 0.813 | 0.24 | 0.30 | 0.43 | 261.44 | 157.41 | 204.57 | 120.14 | 0.375 | 0.26 | 0.45 | 0.38 | |
| C2 | 211.60 | 105.06 | 188.50 | 65.71 | 0.844 | 0.24 | 0.38 | 0.60 | 196.99 | 97.51 | 157.03 | 60.88 | 0.250 | 0.28 | 0.40 | 0.40 | ||
| C3 | 15.53 | 10.49 | 17.28 | 8.46 | 0.563 | 0.23 | 0.55 | 0.73 | 11.62 | 3.66 | 9.61 | 2.88 | 0.094 | 0.16 | 0.26 | 0.32 | ||
| C3-DC (C4-OH) | 4.20 | 1.79 | 3.28 | 0.85 | 0.313 | 0.22 | 0.33 | 0.48 | 3.55 | 1.20 | 3.02 | 1.02 | 0.164 | 0.18 | 0.31 | 0.51 | ||
| C4 | 55.52 | 24.09 | 46.63 | 18.94 | 0.438 | 0.34 | 0.33 | 0.21 | 51.38 | 27.71 | 39.22 | 24.16 | 0.129 | 0.36 | 0.53 | 0.53 | ||
| C5 | 11.94 | 7.50 | 10.48 | 5.87 | 0.625 | 0.30 | 0.53 | 0.50 | 7.96 | 3.05 | 6.01 | 1.67 | 0.027 | 0.19 | 0.31 | 0.57 | ||
| C5-DC (C6-OH) | 1.54 | 0.40 | 1.21 | 0.25 | 0.063 | 0.16 | 0.23 | 0.60 | 0.99 | 0.49 | 0.98 | 0.27 | 1.000 | 0.33 | 0.29 | 0.06 | ||
| C5-OH (C3-DC-M) | 1.82 | 0.87 | 1.52 | 0.30 | 0.438 | 0.26 | 0.33 | 0.51 | 1.23 | 0.26 | 1.02 | 0.25 | 0.129 | 0.20 | 0.17 | 0.08 | ||
| C5:1-DC | 2.25 | 2.81 | 0.76 | 0.22 | 0.125 | 0.39 | 0.97 | 0.23 | 1.73 | 3.58 | 0.47 | 0.40 | 0.020 | 0.45 | 1.80 | 0.49 | ||
| C6 (C4:1-DC) | 2.29 | 0.43 | 1.95 | 0.48 | 0.156 | 0.20 | 0.17 | 0.15 | 1.96 | 0.53 | 1.58 | 0.42 | 0.074 | 0.22 | 0.23 | 0.19 | ||
| C6:1 | 0.34 | 0.12 | 0.19 | 0.03 | 0.094 | 0.40 | 0.19 | 0.00 | 0.36 | 0.17 | 0.17 | 0.03 | 0.020 | 0.55 | 0.35 | 0.00 | ||
| C7-DC | 0.54 | 0.15 | 0.47 | 0.15 | 0.219 | 0.19 | 0.26 | 0.48 | 0.54 | 0.25 | 0.36 | 0.10 | 0.027 | 0.33 | 0.37 | 0.38 | ||
| C16 | 0.34 | 0.24 | 0.29 | 0.12 | 0.563 | 0.27 | 0.36 | 0.00 | 0.28 | 0.27 | 0.26 | 0.06 | 0.250 | 0.29 | 0.50 | 0.03 | ||
| Mean | 0.26 | 0.38 | 0.38 | 0.29 | 0.46 | 0.30 | ||||||||||||
| Median | 0.24 | 0.33 | 0.48 | 0.28 | 0.35 | 0.38 | ||||||||||||
| Lyso-PC | lysoPC a C16:0 | 20.25 | 10.93 | 23.87 | 10.90 | 0.563 | 0.19 | 0.46 | 0.80 | 15.09 | 9.29 | 12.61 | 7.30 | 0.910 | 0.23 | 0.52 | 0.58 | |
| lysoPC a C17:0 | 0.39 | 0.22 | 0.34 | 0.20 | 0.844 | 0.32 | 0.48 | 0.44 | 0.33 | 0.14 | 0.21 | 0.11 | 0.039 | 0.38 | 0.36 | 0.00 | ||
| lysoPC a C18:0 | 4.12 | 1.98 | 4.53 | 1.73 | 1.000 | 0.17 | 0.40 | 0.70 | 3.39 | 1.63 | 2.94 | 1.40 | 0.652 | 0.20 | 0.43 | 0.62 | ||
| lysoPC a C18:1 | 2.19 | 1.09 | 2.95 | 1.17 | 0.063 | 0.23 | 0.42 | 0.73 | 1.57 | 0.80 | 1.43 | 0.93 | 1.000 | 0.23 | 0.50 | 0.59 | ||
| Mean | 0.23 | 0.44 | 0.67 | 0.26 | 0.46 | 0.45 | ||||||||||||
| Median | 0.21 | 0.44 | 0.71 | 0.23 | 0.47 | 0.58 | ||||||||||||
| Acyl-acyl-PC | PC aa C30:0 | 1.13 | 0.52 | 1.21 | 0.38 | 0.688 | 0.17 | 0.37 | 0.80 | 1.05 | 0.59 | 0.75 | 0.22 | 0.074 | 0.24 | 0.42 | 0.45 | |
| PC aa C30:2 | 0.56 | 0.45 | 0.52 | 0.19 | 1.000 | 0.43 | 0.48 | 0.23 | 0.57 | 0.46 | 0.32 | 0.14 | 0.020 | 0.37 | 0.64 | 0.28 | ||
| PC aa C32:0 | 2.74 | 1.04 | 2.83 | 0.78 | 0.844 | 0.11 | 0.32 | 0.88 | 2.54 | 0.87 | 2.08 | 0.44 | 0.129 | 0.16 | 0.26 | 0.49 | ||
| PC aa C32:1 | 0.50 | 0.18 | 0.56 | 0.09 | 0.688 | 0.16 | 0.23 | 0.47 | 0.47 | 0.15 | 0.35 | 0.13 | 0.129 | 0.29 | 0.24 | 0.00 | ||
| PC aa C32:2 | 0.15 | 0.09 | 0.13 | 0.04 | 1.000 | 0.30 | 0.41 | 0.45 | 0.17 | 0.09 | 0.09 | 0.03 | 0.004 | 0.39 | 0.44 | 0.00 | ||
| PC aa C34:1 | 18.64 | 8.88 | 20.70 | 7.72 | 0.438 | 0.16 | 0.41 | 0.84 | 15.13 | 7.94 | 11.68 | 4.56 | 0.203 | 0.22 | 0.42 | 0.46 | ||
| PC aa C34:2 | 3.09 | 1.43 | 3.17 | 1.35 | 1.000 | 0.17 | 0.42 | 0.84 | 3.05 | 1.06 | 2.31 | 0.99 | 0.074 | 0.25 | 0.33 | 0.42 | ||
| PC aa C36:1 | 8.29 | 4.64 | 9.27 | 4.53 | 0.438 | 0.20 | 0.51 | 0.80 | 6.67 | 3.48 | 5.14 | 1.94 | 0.164 | 0.21 | 0.42 | 0.46 | ||
| PC aa C36:2 | 3.43 | 1.66 | 3.71 | 1.42 | 0.438 | 0.18 | 0.42 | 0.82 | 3.11 | 1.41 | 2.41 | 0.83 | 0.074 | 0.18 | 0.38 | 0.57 | ||
| PC aa C36:3 | 1.91 | 0.59 | 2.18 | 0.92 | 0.313 | 0.15 | 0.36 | 0.85 | 1.65 | 0.43 | 1.40 | 0.49 | 0.164 | 0.21 | 0.24 | 0.33 | ||
| PC aa C36:4 | 0.47 | 0.25 | 0.47 | 0.21 | 1.000 | 0.21 | 0.46 | 0.76 | 0.47 | 0.18 | 0.35 | 0.13 | 0.055 | 0.23 | 0.34 | 0.53 | ||
| PC aa C36:6 | 0.05 | 0.02 | 0.05 | 0.02 | 1.000 | 0.22 | 0.38 | 0.63 | 0.05 | 0.02 | 0.03 | 0.01 | 0.098 | 0.36 | 0.38 | 0.33 | ||
| PC aa C38:0 | 0.29 | 0.12 | 0.29 | 0.13 | 1.000 | 0.17 | 0.41 | 0.82 | 0.21 | 0.08 | 0.18 | 0.07 | 0.301 | 0.22 | 0.34 | 0.60 | ||
| PC aa C38:3 | 1.37 | 0.61 | 1.45 | 0.67 | 0.563 | 0.10 | 0.45 | 0.94 | 1.16 | 0.41 | 0.93 | 0.31 | 0.098 | 0.20 | 0.30 | 0.52 | ||
| PC aa C38:4 | 0.50 | 0.20 | 0.53 | 0.24 | 0.563 | 0.13 | 0.41 | 0.89 | 0.45 | 0.12 | 0.39 | 0.14 | 0.496 | 0.18 | 0.26 | 0.41 | ||
| PC aa C38:5 | 0.24 | 0.09 | 0.21 | 0.08 | 0.438 | 0.19 | 0.34 | 0.73 | 0.25 | 0.07 | 0.17 | 0.05 | 0.004 | 0.30 | 0.25 | 0.00 | ||
| PC aa C38:6 | 1.04 | 0.48 | 1.15 | 0.55 | 0.438 | 0.18 | 0.45 | 0.82 | 1.03 | 0.23 | 0.85 | 0.51 | 0.426 | 0.31 | 0.35 | 0.30 | ||
| PC aa C40:3 | 0.10 | 0.04 | 0.10 | 0.03 | 0.563 | 0.17 | 0.37 | 0.74 | 0.10 | 0.04 | 0.08 | 0.03 | 0.098 | 0.29 | 0.31 | 0.39 | ||
| PC aa C40:4 | 0.13 | 0.06 | 0.11 | 0.04 | 0.563 | 0.13 | 0.39 | 0.82 | 0.10 | 0.05 | 0.08 | 0.02 | 0.164 | 0.20 | 0.31 | 0.30 | ||
| PC aa C40:5 | 0.12 | 0.05 | 0.10 | 0.05 | 0.031 | 0.14 | 0.45 | 0.89 | 0.11 | 0.04 | 0.08 | 0.02 | 0.055 | 0.22 | 0.30 | 0.44 | ||
| PC aa C40:6 | 0.57 | 0.29 | 0.59 | 0.35 | 0.844 | 0.16 | 0.54 | 0.89 | 0.50 | 0.12 | 0.40 | 0.16 | 0.074 | 0.22 | 0.28 | 0.37 | ||
| PC aa C42:1 | 0.06 | 0.01 | 0.06 | 0.01 | 0.844 | 0.08 | 0.18 | 0.75 | 0.05 | 0.01 | 0.05 | 0.01 | 0.301 | 0.10 | 0.25 | 0.91 | ||
| PC aa C42:4 | 0.08 | 0.04 | 0.06 | 0.03 | 0.219 | 0.26 | 0.44 | 0.60 | 0.08 | 0.04 | 0.05 | 0.01 | 0.055 | 0.47 | 0.34 | 0.00 | ||
| PC aa C42:5 | 0.06 | 0.02 | 0.06 | 0.02 | 1.000 | 0.15 | 0.33 | 0.73 | 0.06 | 0.02 | 0.05 | 0.02 | 0.070 | 0.23 | 0.34 | 0.35 | ||
| Mean | 0.18 | 0.40 | 0.75 | 0.25 | 0.34 | 0.37 | ||||||||||||
| Median | 0.17 | 0.41 | 0.81 | 0.23 | 0.34 | 0.40 | ||||||||||||
| Acyl-alkyl-PC | PC ae C30:2 | 0.04 | 0.03 | 0.04 | 0.01 | 0.844 | 0.40 | 0.41 | 0.25 | 0.04 | 0.02 | 0.03 | 0.01 | 0.098 | 0.49 | 0.34 | 0.00 | |
| PC ae C32:1 | 0.27 | 0.08 | 0.28 | 0.14 | 1.000 | 0.16 | 0.38 | 0.80 | 0.23 | 0.09 | 0.19 | 0.10 | 0.164 | 0.21 | 0.41 | 0.71 | ||
| PC ae C34:0 | 0.36 | 0.17 | 0.33 | 0.21 | 1.000 | 0.17 | 0.52 | 0.67 | 0.26 | 0.10 | 0.20 | 0.09 | 0.008 | 0.21 | 0.40 | 0.59 | ||
| PC ae C34:1 | 0.63 | 0.25 | 0.66 | 0.28 | 0.688 | 0.09 | 0.41 | 0.94 | 0.53 | 0.21 | 0.44 | 0.20 | 0.301 | 0.18 | 0.39 | 0.55 | ||
| PC ae C34:2 | 0.32 | 0.14 | 0.34 | 0.13 | 0.688 | 0.14 | 0.39 | 0.90 | 0.26 | 0.11 | 0.19 | 0.11 | 0.098 | 0.23 | 0.43 | 0.47 | ||
| PC ae C34:3 | 0.07 | 0.03 | 0.06 | 0.02 | 0.438 | 0.17 | 0.38 | 0.76 | 0.05 | 0.02 | 0.05 | 0.02 | 0.250 | 0.17 | 0.38 | 0.73 | ||
| PC ae C36:2 | 0.22 | 0.11 | 0.21 | 0.12 | 1.000 | 0.12 | 0.54 | 0.90 | 0.18 | 0.08 | 0.14 | 0.08 | 0.039 | 0.22 | 0.49 | 0.64 | ||
| PC ae C36:3 | 0.13 | 0.06 | 0.11 | 0.06 | 0.563 | 0.26 | 0.45 | 0.59 | 0.10 | 0.08 | 0.07 | 0.04 | 0.055 | 0.26 | 0.66 | 0.61 | ||
| PC ae C36:4 | 0.09 | 0.05 | 0.10 | 0.06 | 0.844 | 0.23 | 0.52 | 0.66 | 0.08 | 0.04 | 0.07 | 0.04 | 0.734 | 0.17 | 0.53 | 0.73 | ||
| PC ae C38:2 | 0.17 | 0.09 | 0.16 | 0.10 | 0.438 | 0.12 | 0.58 | 0.90 | 0.12 | 0.04 | 0.09 | 0.04 | 0.164 | 0.28 | 0.29 | 0.19 | ||
| PC ae C38:3 | 0.11 | 0.05 | 0.11 | 0.07 | 1.000 | 0.21 | 0.51 | 0.78 | 0.09 | 0.05 | 0.07 | 0.03 | 0.027 | 0.22 | 0.41 | 0.42 | ||
| PC ae C38:4 | 0.10 | 0.05 | 0.09 | 0.05 | 0.563 | 0.22 | 0.49 | 0.69 | 0.08 | 0.06 | 0.06 | 0.03 | 0.359 | 0.25 | 0.61 | 0.60 | ||
| PC ae C38:5 | 0.10 | 0.09 | 0.11 | 0.05 | 1.000 | 0.39 | 0.60 | 0.49 | 0.12 | 0.06 | 0.09 | 0.07 | 0.020 | 0.28 | 0.57 | 0.57 | ||
| PC ae C38:6 | 0.92 | 0.65 | 1.15 | 1.03 | 0.438 | 0.17 | 0.80 | 0.94 | 0.79 | 0.38 | 0.70 | 0.41 | 0.359 | 0.23 | 0.50 | 0.69 | ||
| PC ae C40:2 | 0.13 | 0.06 | 0.11 | 0.06 | 0.156 | 0.14 | 0.51 | 0.87 | 0.10 | 0.02 | 0.08 | 0.03 | 0.039 | 0.28 | 0.24 | 0.03 | ||
| PC ae C40:3 | 0.09 | 0.05 | 0.05 | 0.02 | 0.094 | 0.39 | 0.45 | 0.36 | 0.07 | 0.04 | 0.04 | 0.01 | 0.129 | 0.31 | 0.47 | 0.23 | ||
| PC ae C40:5 | 0.05 | 0.02 | 0.04 | 0.02 | 0.438 | 0.19 | 0.36 | 0.63 | 0.05 | 0.03 | 0.04 | 0.02 | 0.055 | 0.23 | 0.62 | 0.71 | ||
| PC ae C40:6 | 0.12 | 0.04 | 0.12 | 0.06 | 1.000 | 0.16 | 0.42 | 0.80 | 0.11 | 0.10 | 0.07 | 0.04 | 0.020 | 0.21 | 0.76 | 0.77 | ||
| PC ae C42:3 | 0.18 | 0.12 | 0.10 | 0.02 | 0.156 | 0.32 | 0.44 | 0.07 | 0.15 | 0.10 | 0.09 | 0.02 | 0.359 | 0.33 | 0.47 | 0.13 | ||
| Mean | 0.21 | 0.48 | 0.69 | 0.25 | 0.47 | 0.49 | ||||||||||||
| Median | 0.17 | 0.45 | 0.76 | 0.23 | 0.47 | 0.59 | ||||||||||||
| Spingomyelins | SM (OH) C14:1 | 1.51 | 0.64 | 1.70 | 0.82 | 0.313 | 0.17 | 0.44 | 0.82 | 1.27 | 0.47 | 1.00 | 0.25 | 0.055 | 0.19 | 0.29 | 0.45 | |
| SM (OH) C16:1 | 0.97 | 0.42 | 1.01 | 0.56 | 1.000 | 0.11 | 0.48 | 0.92 | 0.78 | 0.28 | 0.63 | 0.16 | 0.164 | 0.22 | 0.26 | 0.32 | ||
| SM (OH) C22:1 | 1.99 | 1.07 | 1.95 | 1.27 | 0.844 | 0.10 | 0.59 | 0.95 | 1.38 | 0.47 | 1.11 | 0.32 | 0.129 | 0.18 | 0.25 | 0.24 | ||
| SM (OH) C22:2 | 0.52 | 0.28 | 0.50 | 0.28 | 0.688 | 0.12 | 0.54 | 0.93 | 0.43 | 0.14 | 0.36 | 0.11 | 0.164 | 0.16 | 0.30 | 0.59 | ||
| SM (OH) C24:1 | 0.50 | 0.26 | 0.44 | 0.26 | 0.156 | 0.13 | 0.55 | 0.87 | 0.36 | 0.10 | 0.30 | 0.09 | 0.098 | 0.17 | 0.26 | 0.53 | ||
| SM C16:0 | 72.42 | 29.89 | 80.87 | 22.42 | 0.313 | 0.22 | 0.31 | 0.50 | 65.48 | 32.75 | 50.94 | 20.05 | 0.129 | 0.20 | 0.43 | 0.56 | ||
| SM C16:1 | 1.92 | 0.87 | 2.04 | 0.81 | 0.438 | 0.10 | 0.42 | 0.90 | 1.78 | 0.51 | 1.49 | 0.66 | 0.250 | 0.23 | 0.32 | 0.50 | ||
| SM C18:0 | 8.58 | 3.58 | 9.41 | 3.21 | 0.313 | 0.16 | 0.36 | 0.80 | 8.32 | 3.75 | 6.62 | 2.58 | 0.129 | 0.20 | 0.39 | 0.55 | ||
| SM C18:1 | 0.91 | 0.40 | 0.93 | 0.41 | 0.563 | 0.06 | 0.44 | 0.97 | 0.92 | 0.21 | 0.78 | 0.32 | 0.301 | 0.24 | 0.26 | 0.34 | ||
| SM C24:0 | 16.33 | 8.86 | 17.52 | 8.65 | 0.438 | 0.13 | 0.51 | 0.89 | 12.17 | 5.67 | 10.19 | 3.96 | 0.496 | 0.21 | 0.37 | 0.47 | ||
| SM C24:1 | 9.14 | 4.79 | 9.82 | 5.01 | 0.438 | 0.13 | 0.51 | 0.92 | 7.42 | 3.42 | 6.19 | 2.40 | 0.301 | 0.19 | 0.38 | 0.59 | ||
| SM C26:0 | 0.62 | 0.19 | 0.55 | 0.21 | 0.313 | 0.17 | 0.32 | 0.73 | 0.45 | 0.19 | 0.43 | 0.17 | 0.820 | 0.19 | 0.38 | 0.69 | ||
| SM C26:1 | 0.59 | 0.24 | 0.53 | 0.21 | 0.219 | 0.09 | 0.40 | 0.96 | 0.48 | 0.20 | 0.38 | 0.20 | 0.164 | 0.24 | 0.42 | 0.65 | ||
| Mean | 0.13 | 0.45 | 0.86 | 0.20 | 0.33 | 0.50 | ||||||||||||
| Median | 0.13 | 0.44 | 0.90 | 0.20 | 0.32 | 0.53 | ||||||||||||
| Mean total | 0.2055 | 0.4051 | 0.6220 | 0.2501 | 0.3911 | 0.4347 | ||||||||||||
| Median total | 0.1740 | 0.4085 | 0.7050 | 0.2249 | 0.3515 | 0.4899 | ||||||||||||
Reliability of seminal plasma metabolites
The majority of seminal plasma metabolites showed a good reproducibility over time with a median ICC of 0.61 (Supplementary Table 2). However, the reliability varied among the different metabolite subclasses. The metabolite class exhibiting the least reliability was BA (median ICC: 0.31). The highest stability over time was observed for SM (median ICC: 0.70) and LPC (median ICC: 0.69). PC ae showed a median ICC of 0.67 and PC aa showed a median ICC of 0.62. AA exhibited a median ICC of 0.51 and AC had a median ICC of 0.44 (Figure 1).
Figure 1.
Intraclass correlation coefficients (ICCs) of seminal plasma metabolite measurements, classified according to metabolite class. Boxes represent the interquartile range (IQR). Whiskers are 1.5 times the IQR. Mean values are indicated by plus signs and circles represent outliers. AA: amino acids; AC: acylcarnitines; BA: biogenic amines; LPC: lysophosphatidylcholines; PC aa: diacyl-phosphatidylcholines; PC ae: acyl-alkyl-phosphatidylcholines; SM: sphingomyelins.
Twenty-one of the 96 metabolites (21.9%) showed low reliability (ICC < 0.40; Supplementary Table 2). Of these, four metabolites belonged to the AA group, three to BA, five to AC and PC aa each, three to PC ae, and one to LPC. None of the SM had an ICC below 0.40. Excellent reliability (ICC > 0.74) was found for 13 (13.5%) metabolites. Among these were five PC ae, four SM, three PC aa, and one BA. None of the AC or LPC reached this level of reliability.
As ICC alone is insufficient for a comprehensive reflection of metabolome stability, CVW and CVB were also determined (Supplementary Table 2). The interindividual variability was high (median CVB: 0.41) and exceeded 0.20 for 91 metabolites (94.8%). The median variability within individuals was 0.20 and the CVW was below 0.20 for 46 metabolites (47.9%). There were notable differences in CVB and CVW between various metabolite subclasses. The SM group exhibited the lowest CVW (median: 0.18), while the highest CVW was observed for BA (median: 0.28). The CVB was the highest for PC ae (median: 0.51) and the lowest for AA (median: 0.26; Figure 2).
Figure 2.
Within-person coefficients of variation (CVW) of metabolite concentrations in seminal plasma according to metabolite class compared to between-person coefficients of variation (CVB). Boxes represent the interquartile range (IQR). Whiskers are 1.5 times the IQR. Mean values are indicated by plus signs and circles represent outliers. AA: amino acids; AC: acylcarnitines; BA: biogenic amines; LPC: lysophosphatidylcholines; PC aa: diacyl-phosphatidylcholines; PC ae: acyl-alkyl-phosphatidylcholines; SM: sphingomyelins.
It is important to reiterate that the ICC alone may not fully reflect the stability of the metabolic phenotype, potentially leading to incorrect assumptions about metabolic variability. The ICC is elevated when the variability within an individual is minimal and/or the variability between individuals is pronounced. For example, both SM C18:1 and PC ae C36:3 exhibited an ICC of 0.64. However, SM C18:1 had much lower variability, with a CVW of 0.17 compared to 0.26 for PC ae C36:3.
Comparison of subgroups
It has long been known that many metabolites show different correlations with sperm concentration. Therefore, we also investigated whether the intra-individual variability of sperm concentration affects the reliability of metabolite measurements. Donors with similar intra-individual sperm concentration were classified as SSC and those with different intra-individual sperm concentration as DSC. ICC, CVW, and CVB per subgroup are included in Supplementary Table 3.
The SSC group showed a higher median ICC than the DSC group (0.70 and 0.49, respectively). The SSC samples showed an excellent reliability for 40 metabolites (41.7%), whereas only four metabolites (4.2%) of DSC samples, namely Orn, creatinine, PC aa C42:1, and PC ae C40:6, had an ICC > 0.74. The number of metabolites with poor reproducibility (ICC < 0.40) was 18 (18.8%) in the SSC group and 37 (38.5%) in the DSC group, respectively (Figure 3). Metabolites that showed a poor reliability in both groups were Arg, Gly, alpha-aminoadipic acid (α-AAA), 3,4-dihydroxyphenylalanine (DOPA), total dimethylarginine (DMA), C6 (C4:1-DC), C6:1, C16, PC aa C30:2, PC ae C30:2, PC ae C40:3, and PC ae C42:3. Furthermore, PC aa C42:1 and PC ae C40:6 both showed an excellent reliability in both groups. Interestingly, the largest difference regarding the reliability was found within the lipid classes, where metabolites that were highly reliable in the SSC were the least reliable candidates in the DSC subgroup. This was true for PC aa C36:3 (0.85 vs 0.33), PC aa C38:6 and PC aa C40:4 (both 0.82 vs 0.30) and PC aa C40:6 (0.89 vs 0.37), PC ae C38:2 (0.90 vs 0.19) and PC ae C40:2 (0.87 vs 0.03) as well as for SM (OH) C16:1 (0.92 vs 0.32), SM (OH) C22:1 (0.95 vs 0.24) and SM C18:1 (0.97 vs 0.34). On the contrary, excellent ICC values of a metabolite in the DSC group were never poor in SSC.
Figure 3.
Intraclass correlation coefficients (ICCs) of seminal plasma metabolite measurements according to metabolite class in men with similar intra-individual sperm concentration in both samples (SSC), as compared to men with different intra-individual sperm concentration (DSC). Boxes represent the interquartile range (IQR). Whiskers are 1.5 times the IQR. Mean values are indicated by plus signs and circles represent outliers. AA: amino acids; AC: acylcarnitines; BA: biogenic amines; LPC: lysophosphatidylcholines; PC aa: diacyl-phosphatidylcholines; PC ae: acyl-alkyl-phosphatidylcholines; SM: sphingomyelins.
The higher ICC in the SSC group was mainly caused by a lower within person variability (Figure 4). Approximately 77.0% of all metabolites showed a lower CVW (difference > 1%) in SSC compared to DSC. This was especially the case for SM, PC aa, and PC ae.
Figure 4.
Within-person coefficients of variation (CVW) of seminal plasma metabolites according to metabolite class in men with similar intra-individual sperm concentration in both samples (SSC), as compared to men with different intra-individual sperm concentration (DSC). Boxes represent the interquartile range (IQR). Whiskers are 1.5 times the IQR. Mean values are indicated by plus signs and circles represent outliers. AA: amino acids; AC: acylcarnitines; BA: biogenic amines; LPC: lysophosphatidylcholines; PC aa: diacyl-phosphatidylcholines; PC ae: acyl-alkyl-phosphatidylcholines; SM: sphingomyelins.
DISCUSSION
The present work shows the fundamental reproducibility of metabolite measurements in seminal plasma over a period of 3–12 weeks. Almost two-thirds of the 96 metabolites investigated in this study showed good or excellent reliability. From these, only 18 metabolites exhibited different concentrations, six of which belonged to the category of carnitines. It can be concluded that metabolite levels in seminal plasma are relatively stable over the observed period, with some exceptions. Therefore, a single measurement can be used to adequately determine the metabolic signature for the majority of study purposes. However, the utility of a single measurement is limited for highly variable metabolites.
As this work is, to the best of our knowledge, the first investigation of the reliability of metabolite measurements in seminal plasma, a comparison with ICC values from previous research is not feasible. However, Alipour et al.14 investigated the impact of abstinence duration on the intra-individual variability of the seminal plasma metabolome. In this NMR-based approach, which according to the method can only comprise highly abundant metabolites, the concentrations of the majority of 30 metabolites, including AA, did not differ between a few days and 2 h of abstinence. This is consistent with our results, which demonstrated a low CVW and at least a fair reliability for the majority of AA. The cited study found a change in the levels of pyruvate, fructose, acetate, choline, N-acetylglucosamine, O-acetylcarnitine, uridine, and GPC after a period of 2 h compared to a few days of abstinence. Furthermore, progressive motility of sperm was observed to be higher in the second ejaculate, while sperm concentration and semen volume were found to be lower after only 2 h. Thus, this very short abstinence period of only 2 h may have exceeded the secretory capacity of the accessory glands. It is also conceivable that the lower sperm concentration in the second ejaculate may have contributed to the inconsistency observed in some metabolites.
The results of several reliability studies of metabolite measurements in human blood serum or blood plasma are largely consistent with our findings. For example, Yin et al.26 reported a median ICC of 0.62 for metabolites in blood plasma samples collected 4 months apart, while Breier et al.24 found a similar ICC (0.63) for plasma samples collected at a 14-day interval. Other studies examining the short-term variability of the serum metabolome showed a median ICC of 0.65, 0.57, and 0.64, respectively.23,25,27
The present study revealed that the reliability of measurements varied among metabolite groups and was the lowest for BA and AC and the highest for lipids. BA, as the least reliable metabolite group, showed the widest ICC range (from 0 for total DMA, α-AAA, and DOPA to 0.86 for Crea). It should be noted that the median ICC of BA is not fully comparable with the results of studies in human blood plasma, as different BA were analyzed. It is noteworthy, however, that Crea demonstrated excellent reliability in these studies. This is in full agreement with our results. While some recent studies on blood plasma have also identified AC as the metabolite group with the lowest reproducibility (median ICC: 0.45 and 0.44, respectively),23,27 another study has described AC as the most reliable class of metabolites (median ICC: 0.69).26 The authors mentioned that a high interplate variability of AC measurements may have led to a higher CVB and, thus, a higher ICC. Regardless of this, it must be emphasized that most AC are present at much higher concentrations in seminal plasma than in blood plasma.28 They are transported from blood plasma to the epididymis via active transporters in Sertoli cells and seminiferous tubules, where they play a role in the maturation of spermatozoa and in energy production (beta oxidation in mitochondria).29 In this context, correlations with the sperm count have been reported.30 The high CVW of AC levels in seminal plasma is likely due to a broad physiological range that allows flexible adaptation to different energetic requirements.
SM is the most abundant lipid species in human seminal plasma.1 Also, in the present study, the total amount of SM was found to be higher than the total amounts of PC and LPC. Lipids were the most reliable metabolites over time, which is consistent with the results of other authors.23,25,27 These findings support the suitability of lipids as potential biomarkers for the diagnosis and treatment of male fertility disorders. However, it is essential to consider the individual susceptibilities of the lipids toward hydrolysis and oxidation, as these processes could lead to wrong interpretations of oxylipins or lysolipids as promising molecules.
As the ICC considers both within- and between-person variance, some authors have also reported that the CVW and CVB contribute to the ICC in human blood.23,24 The CVW covered a similar range as in our work. However, the CVB was only slightly higher than the CVW in these studies, whereas the median CVB in our work was twice the median CVW.
The considerable variability observed in the measurements between individuals may be due to the wide range of sperm concentrations of the ejaculates or the varying duration of abstinence. Moreover, the lifestyle factors and dietary habits of the participants were not taken into account, except for smoking. Since there is strong evidence that various factors, including smoking habits,18 BMI,8 past illnesses,31 or age32,33 can affect the composition of the seminal plasma metabolome, controlling for these effects could reduce the CVB. Only recently, the influence of the lifestyle on the activity of phospholipases, enzymes that cleave phospholipids, on the brain was reviewed.34 Given the high abundance of phospholipases in seminal plasma, it is plausible that lifestyle factors may have a direct impact on the lipidome, a part of the metabolome of this body fluid. In this context, it is important to note that establishing meaningful reference values is challenging when the CVB is much higher than the CVW. Even if a metabolite concentration falls within the reference range of the population, an altered concentration may be pathological for an individual.12,13
Comparison of reliability between SSC and DSC
A number of studies have indicated a correlation between sperm concentration and metabolite levels in seminal plasma.7,9,10 Consequently, we asked whether intra-individual variability affects the reliability of the measurements.
The reliability of SSC samples was markedly higher compared to DSC (median ICC: 0.71 vs 0.49). The proportion of metabolites exhibiting excellent reliability was 41.7% in SSC and included mainly PC and SM, while only 4.2% of metabolites demonstrated excellent reliability in DSC. Consistently, the intra-individual variability was observed to be lower in SSC for more than two-thirds of the metabolites. In the light of these findings, we agree with the assumption of several authors35,36,37,38 that there is an interaction between spermatozoa and seminal plasma metabolites. This phenomenon is exemplified by the altered seminal plasma metabolome observed in cases of oligozoo- and azoospermia.3,39 The assembly and disassembly of biological membranes may be a contributing factor, which could explain the pronounced impact on the higher reliability of lipid measurements in SSC.
The participants in the present study were not selected based on specific characteristics but were normozoospermic according to the 5th edition of WHO guidelines17 and did not suffer from any metabolic or cardiovascular disease. Accordingly, the ICC estimated should be interpreted as minimum values that can likely be improved by tight study protocols. Furthermore, our study was conducted as a single-center study, whereby the seminal plasma samples were processed on the same equipment. The metabolome analysis was also conducted in a single laboratory on the same mass spectrometer, which avoided center-specific effects.
There are some limitations. The number of participants is only small and larger numbers would be needed to approve the results of this study. There were no duplicates measured, so that a statement about the analytical variance is not possible. Lifestyle factors or dietary habits, except for smoking, were not reported. Furthermore, the samples from the two semen donation time points were measured on different 96-well plates. Although the measured values were normalized plate wise, possible batch effects cannot be excluded.
This study demonstrated the reproducibility of metabolite measurements in human seminal plasma. In conclusion, 78.1% of the seminal plasma metabolites measured by targeted LC-MS/MS were reliable in two donations over 8 weeks. Therefore, the measurement of a single sample is sufficient for most study purposes. However, metabolites with high variability should be interpreted with caution. To address the potential impact of sperm concentration on metabolite levels, a stratified analysis by this parameter is recommended.
Further studies are required to investigate the impact of lifestyle factors and semen parameters on inter- and intra-individual variability. Additionally, research is needed to assess the long-term stability of seminal plasma metabolites, as the findings on the stability of the seminal plasma metabolome may not be applicable to an extended period of time and, thus, to the aging male.
AUTHOR CONTRIBUTIONS
JB recruited the donors, performed semen analysis, and analyzed the data. JB and KME drafted the manuscript. KME and SG conceived the study. All authors read and approved the final manuscript.
COMPETING INTERESTS
All authors declare no competing interests.
ACKNOWLEDGMENTS
The authors are grateful to Dr. Sven Baumann (Forensic Toxicology, Leipzig University Leipzig, Germany) and Dr. Sergei Chetyrkin (Mass Spectrometry Core Lab, Vanderbilt University, Nashville, TN, USA) for statistics advices. The study was financially supported by the German Andrological Society (DGA; research grant to KME), the German Research Foundation (EN 1279/3-1), and the Federal Ministry of Education and Research (BMBF; 01GR2304A).
Supplementary Information is linked to the online version of the paper on the Asian Journal of Andrology website.
REFERENCES
- 1.Jakop U, Müller K, Müller P, Neuhauser S, Callealta Rodríguez I, et al. Seminal lipid profiling and antioxidant capacity:a species comparison. PLoS One. 2022;17:e0264675. doi: 10.1371/journal.pone.0264675. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Drabovich AP, Saraon P, Jarvi K, Diamandis EP. Seminal plasma as a diagnostic fluid for male reproductive system disorders. Nat Rev Urol. 2014;11:278–88. doi: 10.1038/nrurol.2014.74. [DOI] [PubMed] [Google Scholar]
- 3.Boguenet M, Bocca C, Bouet PE, Serri O, Chupin S, et al. Metabolomic signature of the seminal plasma in men with severe oligoasthenospermia. Andrology. 2020;8:1859–66. doi: 10.1111/andr.12882. [DOI] [PubMed] [Google Scholar]
- 4.Mehrparavar B, Minai-Tehrani A, Arjmand B, Gilany K. Metabolomics of male infertility:a new tool for diagnostic tests. J Reprod Infertil. 2019;20:64–9. [PMC free article] [PubMed] [Google Scholar]
- 5.Blaurock J, Baumann S, Grunewald S, Schiller J, Engel KM. Metabolomics of human semen:a review of different analytical methods to unravel biomarkers for male fertility disorders. Int J Mol Sci. 2022;23:9031. doi: 10.3390/ijms23169031. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Nicholson JK, Lindon JC. Systems biology:metabonomics. Nature. 2008;455:1054–6. doi: 10.1038/4551054a. [DOI] [PubMed] [Google Scholar]
- 7.Engel KM, Baumann S, Rolle-Kampczyk U, Schiller J, von Bergen M, et al. Metabolomic profiling reveals correlations between spermiogram parameters and the metabolites present in human spermatozoa and seminal plasma. PLoS One. 2019;14:e0211679. doi: 10.1371/journal.pone.0211679. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Zhou YF, Hou YY, Ban Q, Zhang ML, Huang T, et al. Metabolomics profiling of seminal plasma in obesity-induced asthenozoospermia. Andrology. 2023;11:1303–19. doi: 10.1111/andr.13412. [DOI] [PubMed] [Google Scholar]
- 9.Xu Y, Lu H, Wang Y, Zhang Z, Wu Q. Comprehensive metabolic profiles of seminal plasma with different forms of male infertility and their correlation with sperm parameters. J Pharm Biomed Anal. 2020;177:112888. doi: 10.1016/j.jpba.2019.112888. [DOI] [PubMed] [Google Scholar]
- 10.Qiao S, Wu W, Chen M, Tang Q, Xia Y, et al. Seminal plasma metabolomics approach for the diagnosis of unexplained male infertility. PLoS One. 2017;12:e0181115. doi: 10.1371/journal.pone.0181115. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Fleiss JL. The Design and Analysis of Clinical Experiments. Hoboken: Wiley; 1999. pp. 1–32. [Google Scholar]
- 12.Harris EK. Effects of intra-and interindividual variation on the appropriate use of normal ranges. Clin Chem. 1974;20:1535–42. [PubMed] [Google Scholar]
- 13.Petersen PH, Fraser CG, Sandberg S, Goldschmidt H. The index of individuality is often a misinterpreted quantity characteristic. Clin Chem Lab Med. 1999;37:655–61. doi: 10.1515/CCLM.1999.102. [DOI] [PubMed] [Google Scholar]
- 14.Alipour H, Duus RK, Wimmer R, Dardmeh F, Du Plessis SS, et al. Seminal plasma metabolomics profiles following long (4-7 days) and short (2 h) sexual abstinence periods. Eur J Obstet Gynecol Reprod Biol. 2021;264:178–83. doi: 10.1016/j.ejogrb.2021.07.024. [DOI] [PubMed] [Google Scholar]
- 15.Gupta A, Mahdi AA, Ahmad MK, Shukla KK, Bansal N, et al. A proton NMR study of the effect of Mucuna pruriens on seminal plasma metabolites of infertile males. J Pharm Biomed Anal. 2011;55:1060–6. doi: 10.1016/j.jpba.2011.03.010. [DOI] [PubMed] [Google Scholar]
- 16.Gupta A, Mahdi AA, Shukla KK, Ahmad MK, Bansal N, et al. Efficacy of Withania somnifera on seminal plasma metabolites of infertile males:a proton NMR study at 800 MHz. J Ethnopharmacol. 2013;149:208–14. doi: 10.1016/j.jep.2013.06.024. [DOI] [PubMed] [Google Scholar]
- 17.World Health Organization. WHO Laboratory Manual for the Examination and Processing of Human Semen. 5th ed. Geneva: World Health Organization; 2010. [Google Scholar]
- 18.Engel KM, Baumann S, Blaurock J, Rolle-Kampczyk U, Schiller J, et al. Differences in the sperm metabolomes of smoking and nonsmoking men. Biol Reprod. 2021;105:1484–93. doi: 10.1093/biolre/ioab179. [DOI] [PubMed] [Google Scholar]
- 19.Kenéz Á, Dänicke S, Rolle-Kampczyk U, von Bergen M, Huber K. A metabolomics approach to characterize phenotypes of metabolic transition from late pregnancy to early lactation in dairy cows. Metabolomics. 2016;12:165. [Google Scholar]
- 20.Mangat CS, Bharat A, Gehrke SS, Brown ED. Rank ordering plate data facilitates data visualization and normalization in high-throughput screening. J Biomol Screen. 2014;19:1314–20. doi: 10.1177/1087057114534298. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Rosner B. Fundamentals of Biostatistics. 7th ed. Boston: Cengage Learning; 2010. p. 675. [Google Scholar]
- 22.Pleil JD, Wallace MA, Stiegel MA, Funk WE. Human biomarker interpretation:the importance of intra-class correlation coefficients (ICC) and their calculations based on mixed models, ANOVA, and variance estimates. J Toxicol Environ Health B Crit Rev. 2018;21:161–80. doi: 10.1080/10937404.2018.1490128. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Floegel A, Drogan D, Wang-Sattler R, Prehn C, Illig T, et al. Reliability of serum metabolite concentrations over a 4-month period using a targeted metabolomic approach. PLoS One. 2011;6:e21103. doi: 10.1371/journal.pone.0021103. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Breier M, Wahl S, Prehn C, Fugmann M, Ferrari U, et al. Targeted metabolomics identifies reliable and stable metabolites in human serum and plasma samples. PLoS One. 2014;9:e89728. doi: 10.1371/journal.pone.0089728. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Agueusop I, Musholt PB, Klaus B, Hightower K, Kannt A. Short-term variability of the human serum metabolome depending on nutritional and metabolic health status. Sci Rep. 2020;10:16310. doi: 10.1038/s41598-020-72914-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Yin X, Prendiville O, McNamara AE, Brennan L. Targeted metabolomic approach to assess the reproducibility of plasma metabolites over a four month period in a free-living population. J Proteome Res. 2022;21:683–90. doi: 10.1021/acs.jproteome.1c00440. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Oosterwegel MJ, Ibi D, Portengen L, Probst-Hensch N, Tarallo S, et al. Variability of the human serum metabolome over 3 months in the EXPOsOMICS personal exposure monitoring study. Environ Sci Technol. 2023;57:12752–9. doi: 10.1021/acs.est.3c03233. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Kohengkul S, Tanphaichitr V, Muangmun V, Tanphaichitr N. Levels of L-carnitine and L-O-acetylcarnitine in normal and infertile human semen:a lower level of L-O-acetycarnitine in infertile semen. Fertil Steril. 1977;28:1333–6. doi: 10.1016/s0015-0282(16)42979-1. [DOI] [PubMed] [Google Scholar]
- 29.Mongioi L, Calogero AE, Vicari E, Condorelli RA, Russo GI, et al. The role of carnitine in male infertility. Andrology. 2016;4:800–7. doi: 10.1111/andr.12191. [DOI] [PubMed] [Google Scholar]
- 30.Olesti E, Boccard J, Rahban R, Girel S, Moskaleva NE, et al. Low-polarity untargeted metabolomic profiling as a tool to gain insight into seminal fluid. Metabolomics. 2023;19:53. doi: 10.1007/s11306-023-02020-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Matzkin ME, Beguerie C, de Zuñiga I, Martinez G, Frungieri MB. Impact of COVID-19 on sperm quality and the prostaglandin and polyamine systems in the seminal fluid. Andrology. 2024;12:1078–95. doi: 10.1111/andr.13548. [DOI] [PubMed] [Google Scholar]
- 32.Guo Y, Li J, Hao F, Yang Y, Yang H, et al. A new perspective on semen quality of aged male:the characteristics of metabolomics and proteomics. Front Endocrinol. 2022;13:1058250. doi: 10.3389/fendo.2022.1058250. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Lu JC, Jing J, Yao Q, Fan K, Wang GH, et al. Relationship between lipids levels of serum and seminal plasma and semen parameters in 631 Chinese subfertile men. PLoS One. 2016;11:e0146304. doi: 10.1371/journal.pone.0146304. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Pethe A, Joshi S, Ali Dar T, Poddar NK. Revisiting the role of phospholipases in Alzheimer's:crosstalk with processed food. Crit Rev Food Sci Nutr. 2024 doi: 10.1080/10408398.2024.2377290. doi:10.1080/10408398.2024.2377290. [Online ahead of print] [DOI] [PubMed] [Google Scholar]
- 35.Agarwal A, Mulgund A, Hamada A, Chyatte MR. A unique view on male infertility around the globe. Reprod Biol Endocrinol. 2015;13:37. doi: 10.1186/s12958-015-0032-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Apostoli P, Porru S, Morandi C, Menditto A. Multiple determination of elements in human seminal plasma and spermatozoa. J Trace Elem Med Biol. 1997;11:182–4. doi: 10.1016/S0946-672X(97)80052-1. [DOI] [PubMed] [Google Scholar]
- 37.Komarek RJ, Pickett BW, Lanz RN, Jensen RG. Lipid composition of bovine spermatozoa and seminal plasma. J Dairy Sci. 1964;47:531–4. [Google Scholar]
- 38.Huacuja L, Delgado NM, Calzada L, Wens A, Reyes R, et al. Exchange of lipids between spermatozoa and seminal plasma in normal and pathological human semen. Arch Androl. 1981;7:343–9. doi: 10.3109/01485018108999329. [DOI] [PubMed] [Google Scholar]
- 39.Murgia F, Corda V, Serrenti M, Usai V, Santoru ML, et al. Seminal fluid metabolomic markers of oligozoospermic infertility in humans. Metabolites. 2020;10:64. doi: 10.3390/metabo10020064. [DOI] [PMC free article] [PubMed] [Google Scholar]