Sperm Mitochondrial DNA Copy Number Exhibits Higher Longitudinal Consistency than Conventional Semen Parameters in a Clinical Cohort

Abstract

Research Question:

What is the temporal stability of sperm mitochondrial DNA copy number (mtDNAcn) relative to conventional semen parameters and sperm DNA integrity measures?

Design:

Longitudinal observational study with repeated semen samples in which conventional semen parameters, sperm mitochondrial DNA copy number, and sperm DNA integrity measures were assessed. Ninety-five men participating in the Folic Acid and Zinc Supplementation Trial (FAZST) who provided paired semen samples at baseline and at 2, 4, or 6 months. Temporal stability of sperm mtDNAcn, conventional semen parameters, high DNA stainability (HDS), and DNA fragmentation index (DFI), was assessed using intraclass correlation coefficients (ICCs).

Results:

Sperm mtDNAcn and HDS exhibited the highest temporal stability (ICCs > 0.90), exceeding that of all conventional semen parameters and DFI (ICC range, 0.37–0.76). The temporal stability of sperm mtDNAcn was consistent across sample intervals, with no significant differences in ICCs between timepoints. Sperm mtDNAcn was inversely correlated with sperm concentration, total sperm count, motility, and morphology, and positively correlated with DFI and HDS.

Conclusion:

Sperm mtDNAcn exhibits greater temporal stability than conventional semen parameters, suggesting it may be less susceptible to short-term ejaculate variability and supporting further investigation of its potential utility as a biomarker of male reproductive health.

Introduction

Difficulties achieving pregnancy affect a sizeable share of reproductive-age couples, and male-factor infertility plays a central role in nearly half of these cases (Agarwal et al., 2021; WHO, 2023). Clinical evaluation of male reproductive potential relies predominantly on conventional semen analysis, which assesses sperm concentration, motility, morphology, semen volume, and total sperm count (Jungwirth et al., 2012; Sawant et al., 2024). Although these parameters form the basis of male fertility assessment, several exhibit marked intra-individual fluctuation that reflects biological, environmental, and technical influences (Chiu et al., 2017). Numerous studies in sub-fertile populations have documented limited temporal stability of standard semen parameters (Keel, 2006; Leushuis et al., 2010). This instability reduces the interpretability of single timepoint measurements and complicates both diagnosis and clinical monitoring (Pleil et al., 2018).

In parallel, there is increasing interest in molecular biomarkers that reflect more fundamental aspects of sperm physiology. Among these is sperm mitochondrial DNA copy number (mtDNAcn; reviewed by Sawant and colleagues (Sawant et al., 2026a)). Elevated mtDNAcn has been associated with poor semen parameters (May-Panloup et al., 2003; Song and Lewis, 2008; Wu et al., 2019a), diminished embryo quality in an IVF cycle (Wu et al., 2019b), and a lower likelihood of achieving pregnancy (Rosati et al., 2020; Sawant et al., 2025). As a result, sperm mtDNAcn represents a potential predictor of reproductive success and a promising candidate biomarker.

Within-person variability may help to explain limitations of conventional semen parameter analysis to predict clinical outcomes (Guzick et al., 2001; Jungwirth et al., 2012; Lewis, 2007). Accordingly, classification of infertility status based on a semen sample at one time point may not match that of subsequent assessments. ‘Transient’ infertility based on a single sample and ‘persistent’ infertility in repeated samples may have different underlying explanations and may require different clinical care. In prior work in the Sperm Environmental Epigenetics and Development Study (SEEDS) cohort (n=126), we evaluated variation in fertility status based on semen parameters at repeated clinic visits. Sperm mtDNAcn was associated with fertility status across multiple visits and may better capture decrements in fertility and persistent infertility status; however, mtDNAcn was measured at only one time point (Wu et al., 2019a). To date, the temporal stability of sperm mtDNAcn has not been well described. While prior studies have characterized the reproducibility of conventional semen parameters and established that single-sample assessments are of limited reliability (Barroso et al., 1999; Francavilla et al., 2007), no study, to our knowledge, has evaluated within-person reproducibility of sperm mtDNAcn. Establishing temporal stability is a prerequisite for biomarker validation (Fraser, 2001, 2017; Tworoger and Hankinson, 2006), and addressing this gap is the primary contribution of the present work. In this study, we evaluated the intra-individual reproducibility of sperm mtDNAcn relative to conventional semen and sperm DNA integrity parameters within a clinical cohort. We hypothesized that the temporal stability of sperm mtDNAcn is higher than that of conventional semen parameters.

Materials and Methods

Study Sample

For the current study, deidentified banked whole semen samples and associated data from 100 men were obtained from participants in the Folic Acid and Zinc Supplementation Trial (FAZST). This was a multi-center, double-blind, block-randomized, placebo-controlled clinical trial from 2013–2017 evaluating the effects of daily folic acid and zinc supplementation in men attempting conception with their partners (Schisterman et al., 2020). Heterosexual couples seeking infertility treatment were recruited across four U.S. fertility centres in Utah (n=88), Iowa (n=8), Illinois (n=1), and Minnesota (n=3), with men randomized 1:1 to supplements or placebo and followed for six months after the first baseline visit. The study was approved by institutional review boards at all sites, registered at clinicaltrials.gov (NCT01857310), and overseen by an independent data safety board. Eligible couples included women aged 18–45 years and men aged ≥18 years who were actively trying to conceive and willing to delay fertility treatments long enough to complete at least three weeks of study supplementation (Schisterman et al., 2020). In this study, selection was restricted to those in the placebo arm of the trial with a baseline sample available and paired sample collected at the 2, 4, or 6 months clinic visit. Of the 100 men, five semen samples did not contain sufficient sperm for DNA extraction; therefore, the final analytic sample size included 95 men and 190 semen samples. Specifically, all participants had an initial sample from visit 1 (v1) at baseline (n=95); and a paired second sample at clinic visit 2 (v2) at baseline plus 2 months (n=30); or, visit 3 (v3) at baseline plus 4 months (n=31); or, visit 4 (v4) at baseline plus 6 months (n=34). Wayne State University’s Institutional Review Board (IRB) reviewed the current research and determined that, because it involved only deidentified semen samples and associated data, the study did not constitute human participant research (WSU IRB Non-Human Participant Research Number: 2021–126).

Sample collection and analysis

Semen samples were obtained after 1–7 days of sexual abstinence (recommended, 2–5 days). Immediately following collection, specimens were aliquoted using standardized protocols at each site, snap-frozen in liquid nitrogen, and stored at −70°C before being shipped on dry ice to the National Institutes of Health Biorepository (Rockville, Maryland) for centralized storage. Semen quality was evaluated at each study visit using WHO 5th edition criteria (Cooper et al., 2010), including measurements of semen volume and assessments of sperm concentration, motility, and morphology (Schisterman et al., 2020).

Semen processing, Sperm DNA isolation, and mtDNAcn measurement

Semen samples were received as crude frozen aliquots and were processed using a single step 50% density gradient centrifugation to exclude somatic cells and debris. Sperm pellets were then microscopically inspected for somatic cell contamination using an inverted phase-contrast microscope (Nikon) at 40x magnification. Briefly, semen samples thawed in a heating block at 37°C were transferred to 15-mL conical tubes (Falcon) and centrifuged at 500 × g for 10 min at room temperature using a swing-bucket rotor. Following removal of the seminal plasma, the sperm pellet was gently resuspended in 1.5 mL of ORIGIO Sperm Wash medium (Cooper Surgical) and layered over 1 mL of 50% PureCeption density-gradient medium (Cooper Surgical). The entire available thawed sample was processed; consequently, the volume loaded onto the density gradient varied according to the initial sample volume. Samples were centrifuged at 500 × g for 25 min at room temperature. The sperm-enriched pellet was collected with approximately 500 μL of the overlying medium and transferred to a clean 15-mL conical tube. The recovered fraction was washed with 2 mL of ORIGIO Sperm Wash medium and centrifuged at 1,000 × g for 5 min at room temperature. The resulting pellet was resuspended in approximately 1 mL of Quinn’s Sperm Washing Medium (Cooper Surigcal) for sperm counting and aliquoting. Aliquots containing the required number of sperm were subsequently centrifuged at 1,300 × g for 5 min at room temperature, and the supernatant was removed before downstream processing or storage at −80°C.

Sperm DNA was extracted using the All-Prep DNA/RNA kit (QIAGEN). DNA concentration was first determined using a NanoDrop (Thermo Scientific^™ NanoDrop^™ One^C Spectrophotometer), then diluted in Nuclease Free Water (Invitrogen) to 0.5 ng/μL. Sperm mtDNAcn was quantified using a triplex, probe-based digital polymerase chain reaction (dPCR), adapted from previously described methods (Sawant et al., 2026b; Sawant et al., 2025). Each 15 μL dPCR reaction contained 2 ng of DNA, with final primer concentrations of 250 μM for both the minor and major arc targets and a 1× concentration for RNase P (Applied Biosystems #A30064). Water was included instead of DNA as a negative control. The cycling conditions were as follows: initial activation at 37°C for 45 min, followed by 1 cycle at 95°C for 2 min, then 40 cycles of denaturation at 95°C for 15 s, and annealing/extension at 60°C for 1 min. All reactions were run in duplicate on the Qiacuity dPCR platform (QIAGEN) in 96-well dPCR plates (QIAGEN). Sperm mtDNAcn values were normalized to nuclear DNA copy number using RNase P. Sperm mtDNAcn was calculated using the following formula: mtDNAcn = copy number of minor arc per copy number of RNase P. Details of primer sequences have been described in a prior publication (Sawant et al., 2025).

DFI and HDS measurements

Sperm DNA integrity was assessed using the Acridine Orange Flow Cytometry protocol, a validated flow cytometric method that simultaneously quantifies the extent of sperm nuclear DNA fragmentation and chromatin condensation status in thousands of sperm per sample (Bjorndahl et al., 2022; Evenson, 2022; Evenson et al., 1999; Evenson et al., 2002). For sample preparation, frozen semen aliquots were thawed by immersion in a 37°C water bath for 30 seconds. A volume of semen was then diluted in TNE buffer (150 mM NaCl, 10 mM Tris-HCL, 1 mM Na₂-EDTA, pH=7.4) to a final volume of 100uL at a final concentration of approximately 1–2 × 10⁶ sperm/ml. Immediately after, 200 μl of acid detergent solution (0.08 N HCl, 150 mM NaCl, 0.1% (v:v) Triton X-100, pH=1.20) was added to the sperm suspension, mixed by pipetting, and incubated at room temperature for 30 seconds. Immediately after, 400 μl of staining solution (3.7 mM Citric Acid monohydrate, 12.6 mM Na₂HPO₄, 1 mM Na₂-EDTA, 150 mM NaCl, pH=6.0, containing 6 ug/mL Acridine Orange (Polysciences Inc, #04539) hydrochloride hydrate) was added. Samples were then loaded onto a flow cytometer (Beckman Coulter, CytoFLEX S) equipped with a 488 nm blue laser and configured with green (FITC, 515–530 nm bandpass) and red (630 nm long-pass) filters. Data acquisition was initiated at 2.5 minutes following Acridine Orange staining solution addition, at a slow rate setting in the flow cytometer, with a maximum of 200 events/second. A total of 6,000 flow cytometry events were collected per sample. Duplicates were run for each sample. Raw flow cytometric data were exported as Flow Cytometry Standard (FCS) files and processed using a proprietary algorithm (https://sdfindex.com - SDF-Index^™) to calculate DFI and HDS (Evenson, 2022; SDF-Index).

Calculating the temporal stability of the measured variables

To assess the temporal intra-individual stability of semen parameters, DFI, HDS, and mtDNAcn, we analyzed repeated samples collected from participants across multiple study visits. To address the right-skewed distribution of these variables (as checked with the Shapiro-wilk test), they were transformed using log₁₀. Temporal stability was quantified using the intraclass correlation coefficient (ICC) (Koo and Li, 2016) across the baseline visit and repeated measurements. In the primary analysis, all paired follow-up samples were considered regardless of duration between sample collections. ICCs and their 95% confidence intervals were estimated using a two-way mixed-effects ICC model with absolute agreement and single-measure units implemented in the irr package in R (R Core Team, 2024). For the secondary analysis, stability was assessed separately by duration between sample collection, with paired samples grouped by visit at sample collection: v2 at 2 months, v3 at 4 months, or v4 at 6 months. ICCs were calculated with the two-way mixed-effects model ICC for each paired time-point, namely, v1-v2, v1-v3, and v1-v4. Both analyses used absolute agreement as the ICC formulation. Formal comparisons of ICCs were additionally evaluated using the dep_ci function in the compicc R package. This function calculates the difference between each pair of ICCs and its 95% confidence interval. Two ICCs are considered significantly different if the 95% confidence interval for their difference does not include zero. Spearman correlations were carried out between all measured variables. All data processing, statistical analyses, and visualizations were performed in R (version 4.2.2) using the dplyr, tidyr, purrr, irr, and ggplot2 packages.

Results

Descriptive Statistics

Table 1 presents demographic and reproductive characteristics of the analytic cohort, consisting of 190 semen samples from 95 men. Men had a median (IQR) age at visit 1 of 32.0 (29.0–35.0) years and median (IQR) BMI at visit 1 of 27.0 (23.4–30.7) kg/m². The median (IQR) duration of sexual abstinence at visit 1 prior to sample collection was 3.0 (2.0–4.0) days. Sperm mtDNAcn displayed substantial inter-individual variability, with a median of 3.7 copies and IQR of 2.3–6.7 copies. Using the WHO 2010 (5th edition) reference ranges for semen analysis, a minority (< 20%) of samples fell below thresholds for concentration, volume, total sperm count, and motility, while a larger proportion (29.5%) had morphology below the 4% cutoff.

Table 1:

Cohort descriptions for the subset of men from the FAZST.

Variable (n=95 men)	Median (IQR)	Range
Age at visit 1 (years)	32.0 (29.0–35.0)	22–46
BMI at visit 1 (kg/m2)	27.0 (23.4–30.7)	17.7–49
Abstinence at visit 1 (days)	3.0 (2.0–4.0)	1–7
Semen Parameters (n= 190 samples)	Median (IQR) or mean (SD)	% of samples below WHO reference rangea (n)
Volume (mL)	3.6 (2.7–5.1)	1.6 (3)
Total Sperm Count (millions/ejaculate)	247.8 (125.9–475.6)	7.9 (15)
Concentration (millions/mL)	71.5 (36.5–124.0)	11.6 (22)
Motility (%)	54.5 (18.2)	16.8 (32)
WHO Normal Morphology (%)a	6.0 (3.0–9.0)	29.5 (56)
Sperm DNA Integrity measures (n=178 samples)	Median (IQR)	% of samples below recommended cut-offsc (n)
DFI (%)b	14.5 (9.9–21.2)	16.9 (30)
HDS (%)b	8.3 (5.9–13.0)	16.3 (29)
Sperm mitochondrial marker (n=190 samples)	Median (IQR)	Range
mtDNAcnd (copies)	3.7 (2.3–6.7)	0.45–42

Normally distributed continuous variables are presented as mean (standard deviation), and non-normally distributed continuous variables are presented as median (interquartile range).

^aRefers to WHO 2010 Manual for Semen Analysis

^bFor DFI and HDS, n=178 (89 pairs) due to some samples not having enough sperm for measurement

^cRefers to thresholds in Sperm Chromatin Structure Assay protocol as set by Evenson in 2022 (Evenson, 2022)

^dmtDNAcn does not have a reference range in the WHO 2010 Manual for Semen Analysis. mtDNAcn was calculated as copy number of minor arc per copy number of RNase P.

DFI, DNA Fragmentation Index; FAZST, Folic Acid and Zinc Supplementation Trial; HDS, High DNA Stainability; mtDNAcn, sperm mitochondrial DNA copy number; SD, standard deviation.

Correlations between mtDNAcn and semen characteristics

Spearman correlations between sperm mtDNAcn and conventional semen parameters demonstrated several notable associations (Figure 1). Sperm mtDNAcn was significantly negatively correlated with sperm concentration (ρ = −0.51, p = 4.35×10⁻¹⁴), total sperm count (ρ = −0.48, p = 7.70×10⁻¹³), motility (ρ = −0.31, p = 1.53×10⁻⁵), and morphology (ρ = −0.21, p = 3.05×10⁻³). mtDNAcn was significantly positively correlated with HDS (ρ = 0.27, p = 3.42×10⁻⁴) and DFI (ρ = 0.31, p = 5.52×10⁻⁵), reflecting its relationship with markers of chromatin immaturity and DNA fragmentation.

Figure 1:

Spearman correlation matrix of mtDNAcn, conventional semen parameters, and sperm DNA integrity measures.

The Spearman rank correlation coefficients (ρ, rho) between mtDNAcn, conventional semen parameters (morphology, sperm volume, motility, sperm concentration and total sperm count) and sperm DNA integrity measures (HDS and DFI) are shown. DFI, DNA fragmentation index; HDS, high DNA stainability; mtDNAcn, sperm mitochondrial DNA copy number. Colours represent the magnitude of Spearman rank correlation coefficient values, with the corresponding values indicated by the colour bar on the right. N=190 samples from 95 men.

ICCs for measured variables and intra-parameter associations

Figure 2 displays ICC estimates with 95% confidence intervals for all measured parameters comparing baseline (v1) with the repeated sample (v2, v3, or v4). Sperm mtDNAcn and HDS exhibited the highest temporal stability, with ICC values of 0.91 and 0.92, respectively. Among conventional semen parameters, sperm concentration demonstrated the strongest reproducibility (ICC = 0.76), followed by total sperm count (ICC = 0.69) and motility (ICC = 0.68). DFI showed similar stability (ICC = 0.74), while sperm volume displayed lower reliability (ICC = 0.52). Morphology had the lowest reliability (ICC = 0.37), consistent with well-established variability arising from subjective laboratory assessment (Barroso et al., 1999). In this analysis, mtDNAcn demonstrated significantly higher temporal stability than sperm volume, total sperm count, concentration, motility, morphology and DFI (95% CI of the difference in ICCs of 0.258–0.563, 0.123–0.349, 0.069–0.258, 0.130–0.363, 0.352–0.727 and 0.089–0.251, respectively), while its ICC was comparable with that of HDS (Supplementary Table 1).

Figure 2:

Intraclass correlation coefficients (ICC) for baseline (v1) and repeat samples collected at all timepoints (v2, v3 and v4) for mtDNAcn, conventional semen parameters (morphology, sperm volume, motility, sperm concentration and total sperm count), and sperm DNA integrity measures (HDS and DFI) across visits. Black dot represents the ICC, and the whiskers denote 95% confidence intervals. Numbers above the whiskers indicate ICCs. N=190 samples from 95 men. Statistical analysis of these data is presented in Supplementary Table 1.

Note: DFI = DNA Fragmentation Index, HDS = High DNA Stainability.

In analyses of stability based on timing of the follow-up paired sample, estimates for v1–v2, v1–v3, and v1–v4 comparisons were similar to those of the primary analysis (Figure 3). For all parameters, there were no significant differences between ICCs of the same parameter at different intervals (Supplementary Table 2). Sperm mtDNAcn had the highest ICC measured in a single visit pair (0.95 for v1-v2) and this was similar for v1-v3 and v1-v4, with ICC values near ~ 0.90. HDS also had high ICCs that were similar for all three pairings. DFI, sperm concentration, total sperm count, and motility again demonstrated moderate reproducibility across intervals, whereas sperm volume and morphology remained the least stable.

Figure 3:

Intraclass correlation coefficients (ICCs) for baseline (v1) and repeat samples at each individual timepoint (v2, v3 or v4). Dot represents the ICC, and the whiskers denote 95% confidence intervals. Numbers above the whiskers indicate ICCs. v2, v3 and v4 is baseline plus 2, 4 and 6 months, respectively. N= 95 men at v1; 30, 31, and 34 men at v2, v3, and v4 respectively. Statistical analysis of these data is presented in Supplementary Table 2.

Note: DFI = DNA Fragmentation Index, HDS = High DNA Stainability.

Discussion

To our knowledge, this is the first study to evaluate the temporal stability of sperm mtDNAcn relative to conventional semen parameters and sperm DNA integrity measures. Based on the observed ICC of 0.91, mtDNAcn demonstrated higher temporal stability than conventional semen parameters, which had ICCs ranging from 0.37 to 0.76.

Several studies have documented substantial within-person variability in semen analysis, which complicates interpretation of a single semen analysis (Barroso et al., 1999; Francavilla et al., 2007; Keel, 2006). Our findings align with this body of work. Morphology demonstrated the lowest stability in our cohort, and semen volume also showed only modest reproducibility. These characteristics are known to fluctuate in response to abstinence patterns, illness, lifestyle factors, and short-term biological variation (Keel, 2006). Sperm concentration and total sperm count exhibited moderate to good reproducibility, indicating reasonable temporal reliability. However, these measures are also known to be influenced by both biological and analytical factors, and therefore ICC estimates below 1.0 are expected rather than indicative of poor measurement performance (Aznavour et al., 2022). The overall pattern of instability observed for most semen parameters confirms that many of these measures have limited reliability as indicators of long-term reproductive potential when assessed from a single sample.

The measures with the highest temporal stability were mtDNAcn and HDS, with ICCs of 0.91 and 0.92, respectively. Elevated sperm mtDNAcn has been interpreted as a marker of impaired spermatogenesis, overcompensation of mitochondrial function, and abnormalities in chromatin remodeling during sperm maturation (Oluwayiose et al., 2020). These processes occur during the approximately ten-week spermatogenic cycle and progress relatively slowly over time. As a result, molecular features that capture testicular or epididymal dysfunction may exhibit less short-term variability than standard semen parameters, which are more sensitive to transient fluctuations in ejaculatory physiology and overall health. The high ICC observed for mtDNAcn supports this interpretation. The strong negative correlations between mtDNAcn and sperm concentration, total sperm count, and motility are consistent with prior studies linking elevated mtDNAcn to impaired semen quality (May-Panloup et al., 2003; Song and Lewis, 2008; Wu et al., 2019a). HDS had a similar ICC to mtDNAcn, but while prior studies indicate that HDS is associated with time to pregnancy (TTP) (Buck Louis et al., 2014; Evenson et al., 1999; Spano et al., 2000), the utility of HDS as a predictor of male fertility and ART outcomes remains controversial (Blachman-Braun et al., 2020; Lin et al., 2008; Shen et al., 2025). It is well established that abstinence duration can influence DFI and HDS, and ideally paired samples would be collected under identical abstinence conditions, as done in a study by Esteves et al. on DFI reproducibility using the sperm chromatin dispersion assay (Esteves et al., 2022). In our cohort the recommended abstinence window was 2–5 days but actual intervals ranged from 1–7 days. Of 95 paired samples, 59 (62%) had abstinence durations within 1 day of each other. A sensitivity analysis restricted to these 59 pairs yielded ICC estimates comparable to those from the full sample size (95 pairs) (Supplementary Figure 1), indicating that the abstinence variability in our cohort did not meaningfully affect our reproducibility findings.

Our study design allowed for the assessment of temporal stability across varying sample intervals as part of this study. The v1–v2, v1–v3, and v1–v4 comparisons correspond to 2-, 4-, and 6-month intervals, respectively. In humans, a complete cycle of spermatogenesis requires approximately 64–74 days, with an additional 10–14 days required for epididymal maturation and transport (Heller and Clermont, 1963). Thus, these sampling intervals encompass approximately one, two, and three spermatogenic cycles. Across these biologically distinct intervals, ICC estimates for mtDNAcn remained consistently high, possibly indicating that mtDNAcn stability is not substantially influenced by cycle-to-cycle variation in spermatogenesis within this timeframe. Although the analysis was limited to a follow-up period of approximately six months, had a small sample size, and measured the ICCs between only two timepoints, this time window captures successive rounds of sperm production and maturation and merits additional consideration.

The study has several strengths and limitations. Approximately 88% (84/95 men) of the study cohort was based in Utah, and all study centers followed the same semen processing protocol, thereby enhancing methodological consistency. However, the cohort consisted exclusively of men seeking infertility care, which may limit the generalizability of these findings to the general population. The conventional semen parameters for the cohort were largely within normal limits, with 44% (42/95) of men meeting the WHO 2010 (5th edition) reference values for all 5 basic semen parameters (volume, total sperm count, sperm concentration, motility and morphology). Because a diagnosis of male infertility depends on more than conventional semen parameters alone, our findings should be interpreted as representing men from infertile couples who were predominantly normozoospermic, rather than a population with generally impaired semen quality. Although our data suggest similar stability across 2-, 4-, and 6-month intervals between sample collection, additional studies with repeated measures will be needed to evaluate the stability of sperm mtDNAcn beyond the six-month interval assessed in the current study.

This is the largest study to date on the temporal stability of sperm mtDNAcn, however, the relatively small number of total participants (n=95), as well as in each subgroup analysis by time interval (n=30, 31, and 34 for 2-, 4-, and 6-month intervals, respectively) reduces the precision of ICC estimates and limits interpretation of comparisons of temporal stability across intervals. In addition, each participant contributed only two semen samples. Characterization of within-person variability is strengthened by more repeated measurements per individual, particularly for semen analysis (Keel, 2006) and would allow for more robust ICC estimation. That said, paired replicates are a well-established design for ICC estimation, and the subgroup analyses, while limited in sample size, serve a useful purpose in distinguishing between repeatability (reliability between any two samples) and temporal stability (reliability as a function of time between samples). The precision of subgroup estimates is reflected in their confidence interval widths; for example, the ICC for mtDNAcn at v1-v2 was 0.95 (95% CI: 0.91–0.97) and at v1-v3 was 0.86 (95% CI: 0.75–0.93). The interval-specific ICC estimates appeared broadly similar across timepoints; however, these comparisons were limited by small subgroup sample sizes and lower statistical power. The absence of statistically significant differences should not be interpreted as evidence of equivalence. Because the analytic subset was randomly selected from placebo participants with available paired specimens, there is a small potential for selection bias. Future studies with larger samples and longer observation periods would better characterize how mtDNAcn stability changes over time.

Conclusion

Our findings show that sperm mtDNAcn demonstrates lower intra-individual variation than conventional semen parameters and DFI, indicating high within-person reproducibility across repeated ejaculates in this cohort. These results suggest mtDNAcn is a candidate biomarker warranting validation in larger prospective studies. Future studies of the relationship of sperm mtDNAcn with ART outcomes are warranted to determine mtDNAcn as a measure of male reproductive health and fertility outcomes.

Key Message.

Sperm mitochondrial DNA copy number (mtDNAcn) shows high temporal stability across repeated semen samples, outperforming conventional semen parameters. This temporal stability remains high even with differing time intervals, supporting mtDNAcn as a robust biomarker for male reproductive assessment, and strengthening its utility for longitudinal studies, clinical interpretation, and predictive modelling.

Biography

Savni Sawant completed her Ph.D. at Wayne State University, where she studied sperm molecular biomarkers for predicting fertility and fecundity. Her work integrates mitochondrial DNA copy number, sperm DNA methylation, and semen parameters using bioinformatics across general population and IVF cohorts, with translational relevance to male reproductive health.