Shorter abstinence decreases sperm deoxyribonucleic

acid fragmentation in ejaculate

Jaime Gos
alvez, Ph.D.,a Mercedes Gonz
alez-Mart
ınez, Ph.D.,a Carmen L
andez, Ph.D.,a
opez-Fern

c b
Jos
e L. Fern
andez, M.D., Ph.D., and Pascual S

anchez-Mart
ın, M.D., Ph.D.
a
Departamento de Biolog
ıa, Unidad de Gen
etica, Universidad Aut
onoma de Madrid, Madrid; b Cl
ınica Ginemed, Seville; and
c
Instituto de Investigaci
on Biom
edica de A Coru~na, Complejo Hospitalario Universitario, A Coru~na, Spain

Objective: To evaluate the relationship between duration of sexual abstinence and sperm selection on sperm DNA
fragmentation (SDF).
Design: Prospective study based on normozoospermic individuals.
Setting: Fertility and IVF unit and university unit for research.
Patient(s): Two cohorts of normozoospermic individuals: 21 men (aged 25–35 years) attending a clinic and with
clearly adverse female factors; and a group of 12 selected donors (aged 20–25 years).
Intervention(s): SDF assessment using the sperm chromatin the dispersi
on test (Halosperm) in two cohorts of
normozoospermic men.
Main Outcome Measure(s): SDF assessment after 24 hours of abstinence with recurrent ejaculation (one every 24
hours) using neat sperm samples; and SDF assessment before and after sperm selection with abstinence of 3 hours.
Result(s): Lower baseline levels of SDF were observed after shorter periods of abstinence between ejaculations (24
hours and 3 hours) than those recommended. This effect is much more marked after quick repetitive ejaculation
(3 hours of abstinence) and sperm selection.
Conclusion(s): The present results challenge the role of abstinence in current male infertility treatments and sug-
gest that SDF can be efficiently reduced by a biological practice consisting of short-term recurrent ejaculation
coupled with effective sperm selection. (Fertil Steril
2011;96:1083–6. 2011 by American Society for
Reproductive Medicine.)
Key Words: Male factor, sexual abstinence, sperm DNA fragmentation, sperm quality, sperm chromatin
dispersion test

Low sperm DNA damage is considered an important factor in
achieving a healthy pregnancy. Consequently, different strategies
have been used to reduce the impact of a high frequency of sperm
with fragmented DNA in the ejaculate used for assisted reproductive
techniques (ART): [1] ex vivo direct selection of sperm containing
undamaged DNA; [2] intake of certain drugs that may decrease
the rate of sperm DNA fragmentation (SDF); and [3] testicular
sperm retrieval. Swim-up techniques or centrifugation in density
gradients (1) by themselves or coupled with more sophisticated
approaches, such as intracytoplasmic injection (ICSI) of high-
resolution morphologically selected spermatozoa (2), use of hyalur-
onic acid for sperm selection before ICSI (physiologic ICSI) (3), and
magnetic cell sorting using annexin V–conjugated microbeads (4),
all reduce the possibility of selecting sperm containing fragmented
DNA. Additionally, the administration of antioxidant compounds
(5) or even antibiotics could be effective in improving the outcome
in certain situations, such as Chlamydia infection (6). Finally, testic-
ular sperm obtained after biopsy may contain a reduced SDF propor-
tion in comparison with ejaculated sperm (7).
Received May 19, 2011; revised July 22, 2011; accepted August 17, 2011;
published online September 15, 2011.
J.G. has nothing to disclose. M.G.-M. has nothing to disclose. C.L.-F. has
nothing to disclose. J.L.F. has nothing to disclose. P.S.-M. has nothing
to disclose.
This work was partially supported by a grant (BFU2010-16738) from
Ministerio de Ciencia y Tecnolog
ıa, Spain (MCYT).
Reprint requests: Jaime Gosa
lvez Berenguer, Ph.D., University Auton-
oma, de Madrid, Departamento de Biolog
ıa, Darwin sn, Madrid
28049, Spain (E-mail: jaime.gosalvez@uam.es).
In ART, as a rule a 3- to 4-day period of abstinence before insem-
ination is recommended. Although this practice can be used as
a strategy for interlaboratory homogenization of the results to assess
sperm parameters, there is no clear scientific evidence showing its
utility as a practice to achieve greater reproductive success. Thus,
its application in ICSI is not scientifically justified for this latter
point. Therefore, the question of whether a prolonged period of ab-
stinence from ejaculation results in an improvement of the final
sperm quality should be assessed. Indeed, some previous studies
have showed the negative effects of prolonged abstinence in some
semen parameters (8, 9). The aim of this prospective study was to
analyze the effect of recurrent ejaculation on SDF, be it positive,
negative, or none.
MATERIALS AND METHODS
Two different experiments were conducted; a diagram of the whole experi-
mental design is shown in Figure 1. In experiment 1 a cohort of 21 normozoo-
spermic men (aged 25–35 years) attending a clinic and with clearly adverse
female factors to achieving pregnancy were included in the analysis. Each in-
dividual abstained for 4 days, and after that the level of SDF was quickly as-
sessed (neat 96 hours of abstinence [Nt96h]; Fig. 1, top). After this period,
each individual continued with a recurrent ejaculation (one every 24 hours)
for 4 consecutive days, and the SDF was reassessed using the last ejaculation
(Nt24h; Fig. 1, top). In experiment 2, a group of 12 selected donors (aged 20–
25 years) abstained for 96 hours, and the baseline levels of SDF were as-
sessed (Nt96h; Fig. 1, bottom). An additional semen sample was obtained
3 hours after the first ejaculation (neat 3 hours of abstinence [Nt3h];
Fig. 1, bottom). In this second experiment, each sample (Nt96h and Nt3h;
Fig. 1, bottom) was divided to produce two aliquots. One was considered

0015-0282/$36.00 Fertility and Sterility Vol. 96, No. 5, November 2011 1083
doi:10.1016/j.fertnstert.2011.08.027 Copyright ª2011 American Society for Reproductive Medicine, Published by Elsevier Inc.
 FIGURE 1
Timeline diagram according to the experimental design. Times of ejaculation are highlighted with gray circles, intervals and times of
abstinence are highlighted with white squares, and times of DNA fragmentation testing (SCD) are highlighted in blue circles. Green circles
represent the whole ejaculate used to produce two aliquots to assess DNA damage in neat and selected samples. Experiment 1, Nt96h: SDF
in neat semen samples after 96 hour of abstinence; Nt24h: SDF in neat semen samples after 24 hours of abstinence within a series of recurrent
ejaculations each 24 hours. Experiment 2, Nt96h: SDF in neat semen samples after 96 hours of abstinence; St96h: SDF in selected sperm
samples using equivalent aliquots; Nt3h: SDF in neat semen samples after 3 hours of abstinence; St3h: SDF in selected sperm samples using
equivalent aliquots.

Gos
alvez. Sexual abstinence and sperm DNA damage. Fertil Steril 2011.

as neat semen (Nt3h), and the rest was processed using standard density gra-
dient centrifugation for sperm selection (selected 3 hours of abstinence
[St3h]; Fig. 1, bottom). The SDF was assessed in the four new subpopulations
(Nt96h/St96h and Nt3h/St3h; Fig. 1, bottom).
For sperm selection using density gradient centrifugation, a 1-mL aliquot
of neat sperm was gently overlaid on a column consisting of 1 mL of 45% and
1 mL of 90% density gradient material (Sperm grad-TM; Vitrolife) in a 15-
mL conical tube and centrifuged at 300
g for 15 minutes at room temper-
ature. The pellets were washed once with 5 mL of Sperm Prep Medium
(Medi-Cult), centrifuged at 300
g for 8 minutes, and then resuspended
in 0.5 mL Sperm Prep Medium as in the washing step. Sperm DNA fragmen-
tation was analyzed with Halosperm (Halotech). The SDF was determined
from chromatin dispersion patterns using the Halosperms kit (Halotech
DNA), a commercial variant of the sperm chromatin dispersion (SCD) test.
For each experiment, 25 mL of diluted spermatozoa (10
106 spermato-
zoa/mL) was added to a vial containing low melting point agarose and gently
mixed. A small aliquot of the agarose–sperm mixture (10 mL) was then
spread on pretreated slides (provided in the kit), covered with a glass cover-
slip, and placed in a refrigerator on a cold metallic plate for 5 minutes. After
solidification, a first step including a controlled denaturation of the DNA us-
ing an acid solution for 7 minutes at room temperature was performed. Once
the DNA was denatured in the putative DNA breaks, a controlled extraction
of the nuclear proteins was produced. The DNA halos from the SCD test were
visualized by fluorescence microscopy after being stained with the DNA in-
tercalating fluorochrome 40 ,6-diamidino-2-phenylindole, dihydrochloride
(Invitrogen) in antifading Vectashield (Vector Laboratories). The double-
staining protocol used Gel Red for DNA and 2,7-dibrom-4-hydroxy-mer-
cury-fluorescein (BMF) for specific protein staining (Sigma-Aldrich). A
charge-coupled device camera (Leica DFC350 FX; Leica Microsystems)
with two independent ultraviolet and blue excitation filters was used for vi-
sualization and capture. Files were merged using Adobe Photoshop CS3
(Adobe Systems). This gives rise to partially deproteinized nucleoids in
which the DNA loops expand, forming halos of chromatin dispersion in those
spermatozoa free of DNA damage. However, the spermatozoa nucleoids
whose DNA is fragmented either do not develop a dispersion halo, or the
halo is minimal.
Statistical comparison was performed by nonparametric tests: Mann-
Whitney U test to compare values obtained between two groups and the
Kruskal-Wallis test to compare three or more groups. The significance level
was 0.05. Statistical analyses were performed with SPSS v.15.0.
RESULTS
6-Diamino-2-phenylindole (DAPI) for selective DNA staining,
counterstained with BMF for proteins, provided high-resolution im-
ages after processing sperm samples with the SCD test. Sperm heads
exhibiting the chromatin dispersion halo formed by expanding DNA
loops from a central core formed by DNA and protein remnants con-
tain a putative normal DNA molecule (Fig. 2A). Sperm heads with
small or absence of haloes around the core indicate spermatozoa
with fragmented DNA. Figure 2B shows a sample analyzed after 3
hours of abstinence and sperm selection, in which the frequency
of sperm displaying sperm containing fragmented DNA was reduced
notably. It is interesting to note that the combination of short absti-
nence and sperm selection tends to produce a homogeneous sperm
subpopulation with large and homogenous haloes of chromatin dis-
persion (compare Fig. 2A and B).
Descriptive statistics for both experiments are shown in Figure 3.
In the first experiment the SDF obtained after the first ejaculation
(Nt96h) was 26.9  11 (mean  SD), whereas this value drops to
19.6  8.4 after the established period of recurrent ejaculation/absti-
nence (Nt24h). A reduction in the level of SDF close to 25% was
achieved (Fig. 3A), and significant differences were obtained
when both samples were compared (W ¼ 134.5; P¼.031). In the

1084 Gos
alvez et al. Sexual abstinence and sperm DNA damage Vol. 96, No. 5, November 2011
 FIGURE 2
Sperm DNA fragmentation as visualized after application of the
SCD test and staining using DAPI for DNA and counterstaining
with 2,7-dibrom-4-hydroxy-mercury-fluorescein (BMF) for protein
visualization. (A) Fresh neat sperm sample after 3 hours of
abstinence and (B) equivalent semen sample after sperm
selection. Note the reduction of sperm DNA fragmentation and the
tendency to produce homogenous haloes of chromatin
dispersion.
Gos
alvez. Sexual abstinence and sperm DNA damage. Fertil Steril 2011.
case of experiment 2, for the first ejaculation and 3 hours later, neat
or selected, the Kruskal-Wallis test showed that significant differ-
ences exist when all the samples are compared (H ¼ 17.1;
P¼.0006). Comparing the results for SDF obtained from different
samples in this experiment by pairs, we found 22.2  7.4 in
Nt96h, whereas in St96h the value observed was 17.0  5.5 (Z ¼
3.06; P¼.002); this means that sperm selection results in a signif-
icant decrease (nearly 22%; Fig. 3B) in the level of SDF after 96
hours of abstinence and sperm selection. The second ejaculation af-
ter 3 hours of abstinence produced the following results: 20.8  6.7
for Nt3h and 10.8  6.3 for St3h (Fig. 3B); and these differences
were also highly significant (Z ¼ 2.78; P¼.005). This shows
that this decrease is much more marked when sperm selection is con-
ducted on sperm samples collected after only 3 hours of abstinence
(compare whole results in Fig. 3B). There was also some decrease in
the level of SDF in neat semen samples after 3 hours of abstinence,
although the differences were not statistically different (Nt96h vs.
Nt3h; W ¼ 62.5; P¼.60). Furthermore, significant differences
were found when comparing the selected fractions (St96h vs.
Fertility and Sterility
FIGURE 3
Box-and-whisker plots to compare SDF levels assessed at
different durations of abstinence (A, experiment 1; B, experiment
2). Black, white, and red boxes for experiment 1 represent neat
semen samples. Red boxes for experiment 2 represents neat
semen samples, whereas yellow boxes represent selected sperm
after centrifugation in density gradient. The intensity in SDF
reduction is shown next to green arrows. Experiment 1, Nt96h:
SDF value distribution in neat semen samples after 96 hours of
abstinence; Nt24h: SDF value distribution in neat semen samples
after 24 hours of abstinence within a series of recurrent
ejaculations each 24 hours. Experiment 2, Nt96h: SDF value
distribution in neat semen samples after 96 hours of abstinence;
St96h: SDF value distribution in selected sperm samples using
equivalent aliquots; Nt3h: SDF value distribution in neat semen
samples after 3 hours of abstinence; St3h: SDF value distribution
in selected sperm samples using equivalent aliquots. Sperm
selection was performed using centrifugation in density gradient.
Gos
alvez. Sexual abstinence and sperm DNA damage. Fertil Steril 2011.
St3h; W ¼ 32.9; P¼.02). This means that after only 3 hours of ab-
stinence sperm selection is much more efficient in reducing SDF
(see descriptive statistics in Fig. 3B). Finally, the whole effect on
SDF reduction after recurrent ejaculation and only 3 hours of absti-
nence coupled with sperm selection is seen when Nt96h (22.2  7.4)
was compared with St3h (10.8  6.3). This produces an effective re-
duction of SDF of nearly 48% (Fig. 3B), and significant differences
were also found (W ¼ 21.5; P¼.003).
DISCUSSION
Although SDF can be efficiently reduced by a biological practice
consisting of short-term recurrent ejaculation coupled with effective
sperm selection, there are subtle differences between the different
strategies tested here. After 24 hours of abstinence without sperm se-
lection (experiment 1) a reduction in the level of SDF close to 25% is
achieved (Fig. 3). This reduction is similar after 24 hours of absti-
nence and sperm selection (22%; experiment 2, Fig. 3) but greater
after 3 hours of abstinence and sperm selection (44%; experiment
2, Fig. 3). The effective reduction comparing 3 hours of abstinence
1085
and sperm selection performed on both samples is on the order of
31% (Fig. 3). Over both experiments, if we compare the level of
SDF in neat semen samples obtained after a relatively long period
of abstinence (24 hours) and the level of SDF obtained in the second
sample derived after 3 hours of abstinence coupled with sperm se-
lection, a reduction in SDF close to 48% is produced (Fig. 3).
These results are preliminary, and this is probably not a methodo-
logic practice that produces equivalent results in patients with
a severe male factor. However, the noninvasive character of the prac-
tice and the clear positive response of the semen sample for this
practice suggest its utilization in couples in which a normozoosper-
mic male partner is present, the female factor is limiting the out-
come, and the couple must be assisted with ICSI. We do not
discount similar beneficial effects on SDF in the setting of male fac-
tor infertility. Indeed, it has been suggested that patients with male
factor infertility should collect the semen after just 1 day of absti-
nence, because they produce peaks of best mean sperm motility
and normal morphology in moderate oligozoospermic samples after
1 to 2 days of abstinence (9).
The proportion of SDF seems to be significantly increased during
transit from the seminiferous tubules to the epididymis, possibly re-
lated to oxidative stress, so DNA damage must accumulate in the
sperm population owing to the nonfunction of DNA repair pathways
(10). In fact, testicular sperm show lower SDF than that found in
ejaculated sperm; thus, testicular biopsy is being proposed as
a way to obtain sperm cells with better DNA integrity (7). This
should be true in normozoospermic men but seems not to be the
case in azoospermia: samples from men with azoospermia due to
testicular failure show higher SDF than those from men with ob-
structive azoospermia (11).
Overall, if oxidative stress in the epididymis increases SDF, then
the more prolonged the storage time in epididymis the higher will be
the increase in SDF. Thus, reducing the storage time should help to
prevent the extratesticular increase in SDF. However, we have also
found in the literature that these beneficial effects may not be so ev-
ident; De Jonge et al. (12) reported that abstinence had no statisti-
cally significant influence in parameters such as viability, total and
grade A motility, morphology, or SDF. Interestingly, these authors
reported something that seems quite logical, because it is a fact
that the percentage of sperm with immature chromatin was statisti-
cally significantly increased with 1 day of abstinence. One of the hy-
potheses we are addressing now is that immature sperm produced
after recurrent ejaculation accumulates higher levels of sperm con-
taining fragmented DNA, and immature sperm are not positively se-
lected after centrifugation in density gradient, thus producing large
differences in SDF between neat and selected sperm samples. Al-
though the beneficial effect of this noninvasive practice for reducing
SDF seems evident, especially after sperm selection, and probably
may compensate for any possible negative effects from increasing
levels of reactive oxygen species (ROS) or other free radicals pro-
duced after semen processing, this issue needs to be analyzed in
the context of recurrent ejaculation in more detail. In this respect,
two points need to be considered: [1] separating spermatozoa
from seminal plasma during semen processing is associated with in-
creased ROS production, which may impair sperm function dramat-
ically (13, 14); and [2] increased sperm immaturity after recurrent
ejaculation produces increased oxidative stress due to the
cytoplasmic content in the proximal droplets. This also may
impair sperm DNA quality because these sperm alterations may
also increase ROS production (15).
Of course, short-term ejaculations and sperm selection can be in-
dicated in those cases in which ICSI is necessary. However, we must
bear in mind that most men experience a refractory period immedi-
ately after ejaculation, and these periods could be longer under
stressing environments, such as assisted reproduction. Short-term
and recurrent ejaculations, which unlike testis biopsy are noninva-
sive methodologies for obtaining sperm, could be a reasonable strat-
egy to improve the DNA quality for sperm used in ICSI. For IUI, this
strategy could be limited by the amount of sperm finally recovered,
although it cannot be discarded without investigation.
Acknowledgments: The authors thank A. Gos
albez and B. Mart
ın for techni-
cal assistance and Prof. Godfrey M. Hewitt for critical reading of the
manuscript.

REFERENCES
1. Jackson RE, Bormann CL, Hassun PA, Rocha AM,
Motta EL, Serafini PC, et al. Effects of semen storage
and separation techniques on sperm DNA fragmenta-
tion. Fertil Steril 2010;94:2626–30.
2. Wilding M, Coppola G, di Matteo L, Palagiano A,
Fusco E, Dale B. Intracytoplasmic injection of mor-
phologically selected spermatozoa (IMSI) improves
outcome after assisted reproduction by deselecting
physiologically poor quality spermatozoa. J Assist
Reprod Genet 2011;28:253–62.
3. Parmegiani L, Cognigni GE, Bernardi S, Troilo E,
Ciampaglia W, Filicori M. ‘‘Physiologic ICSI’’: hya-
luronic acid (HA) favors selection of spermatozoa
without DNA fragmentation and with normal nu-
cleus, resulting in improvement of embryo quality.
Fertil Steril 2010;93:598–604.
4. Said T, Agarwal A, Grunewald S, Rasch M,
Baumann T, Kriegel C, et al. Selection of nonapop-
totic spermatozoa as a new tool for enhancing assis-
ted reproduction outcomes: an in vitro model. Biol
Reprod 2006;74:530–7.
5. Zini A, Al-Hathal N. Antioxidant therapy in male infer-
tility: fact or fiction? Asian J Androl 2011;13:374–81.
6. Gallegos G, Ramos B, Santiso R, Goyanes V,
Gos
alvez J, Fern
andez JL. Sperm DNA fragmenta-
tion in infertile men with genitourinary infection by
Chlamydia trachomatis and Mycoplasma sp. Fertil
Steril 2008;90:328–34.
7. Greco E, Scarselli F, Iacobelli M, Rienzi L, Ubaldi F,
Ferrero S, et al. Efficient treatment of infertility due
to sperm DNA damage by ICSI with testicular sper-
matozoa. Hum Reprod 2005;20:226–30.
8. Carlsen E, Petersen JH, Andersson AM,
Skakkebaek NE. Effects of ejaculatory frequency
and season on variations in semen quality. Fertil Steril
2004;82:358–66.
9. Levitas E, Lunenfeld E, Weiss N, Friger M, Har-
Vardi I, Koifman A, et al. Relationship between the
duration of sexual abstinence and semen quality:
analysis of 9,489 semen samples. Fertil Steril
2005;83:1680–6.
10. Gil-Guzm
an E, Ollero M, L
opez MC, Sharma RK,
Alvarez JG, Thomas AJ, et al. Differential production
of reactive oxygen species by subsets of human sper-
matozoa at different stages of maturation. Hum Re-
prod 2001;16:1922–30.
11. Meseguer M, Santiso R, Garrido N, Gil-Salom M,
Remoh
ı J, Fern
andez JL. Sperm DNA fragmentation
levels in testicular sperm samples from azoospermic
males, assessed by the sperm chromatin dispersion
(SCD) test. Fertil Steril 2009;92:1638–45.
12. De Jonge C, LaFromboise M, Bosmans E,
Ombelet W, Cox A, Nijs M. Influence of the absti-
nence period on human sperm quality. Fertil Steril
2004;82:57–65.
13. Aitken RJ, Clarkson JS. Significance of reactive oxy-
gen species and antioxidants in defining the efficacy
of sperm preparation techniques. J Androl
1988;9:367–76.
14. Zini A, Mak V, Phang D, Jarvi K. Potential adverse
effect of semen processing on human sperm deoxyri-
bonucleic acid integrity. Fertil Steril 1999;72:496–9.
15. Said TM, Agarwal A, Sharma RK, Thomas AJ Jr,
Sikka SC. Impact of sperm morphology on DNA
damage caused by oxidative stress induced by b-nic-
otinamide adenine dinucleotide phosphate. Fertil
Steril 2005;83:95–103.

1086 Gos
alvez et al. Sexual abstinence and sperm DNA damage Vol. 96, No. 5, November 2011