Human Reproduction Vol.21, No.11 pp. 2868–2875, 2006
doi:10.1093/humrep/del250
Advance Access publication June 22, 2006.
Quantitative effects of male age on sperm motion
E.Sloter1,2,3,4, T.E.Schmid1,2,5, F.Marchetti1,5, B.Eskenazi2, J.Nath3 and A.J.Wyrobek1,5,6
1
Biosciences Directorate, Lawrence Livermore National Laboratory, Livermore, 2School of Public Health, University of California,
Berkeley, CA and 3Genetics and Developmental Biology Program, West Virginia University, Morgantown, WV, USA
4
Present address: Developmental and Reproductive Toxicology, WIL Research Laboratories, Ashland, OH, USA
5
Present address: Life Sciences Division, Lawrence Berkeley National Laboratory, Berkeley, CA, USA
6
To whom correspondence should be addressed at: Life Sciences Division, Lawrence Berkeley National Laboratory, Berkeley, CA, USA.
E-mail: [email protected]
BACKGROUND: Semen quality is associated with fertility status, but there is little quantitative information on
risk factors that affect semen quality, especially in non-clinical populations. Advancing male age has been associated with a decline in semen quality, with the largest effect being on sperm motility. However, there is little quantitative data on the specific components of sperm motion that are affected by male age. METHODS: We
performed linear regression analyses of 14 aspects of semen quality measured by computer-assisted semen analysis (CASA) in a non-clinical cohort of 90 non-smoking men, aged 22–80 years, who had no history of infertility
or reproductive problems. RESULTS: We found age-associated declines in CASA-determined motility (% motile,
0.8% per year; % progressively motile, 0.9% per year; % rapidly motile, 0.4% per year, P £ 0.001) and three
quantitative aspects of sperm motion [linearity (LIN), 0.2% per year; straight line velocity (VSL), 0.2% per year,
and average path velocity (VAP), 0.3% per year, P < 0.05], with no evidence for age thresholds and no significant
association with abstinence duration. Age was not significantly associated with amplitude of lateral head (ALH)
displacement, beat cross frequency (BCF) and nuclear elongation or size. CONCLUSIONS: Quantitative analysis
of sperm motion indicates that as men age, they produce fewer motile sperm, which are able to travel less along a
linear path, thus covering less forward distance per unit time. These findings may have fertility implications for
men who choose to delay fatherhood.
Key words: CASA/human/male age/motility/sperm
Introduction
One in six couples of childbearing age is unable to conceive,
and male factors contribute to >25% of cases (Templeton,
1995). About 60% of male factor infertility may be due to
genetic causes, but the contributions from environmental and
host factors such as age and diet are poorly understood. As
more couples are choosing to delay childbearing (Ventura
et al., 1997), it is becoming increasingly important to understand how male age affects the risk of infertility, especially in
non-clinical populations.
Prior rodent and human studies have reported that certain
aspects of semen quality decline with age (Kidd et al., 2001;
Sloter et al., 2004). However, most human studies have been
conducted in clinical settings with sperm donors or men with
fertility concerns. Thus, it is unclear whether these findings are
relevant to the general population. Also, few of the clinical studies included men over the age of 50 years and seldom considered
potential confounders that might explain changes with age, such
as smoking history or occupational exposures (Kidd et al.,
2001). We recently reported age-associated reductions in semen
volume, sperm concentration and subjective sperm motility
using conventional semen analyses in a group of ∼100 healthy
workers and retirees (Eskenazi et al., 2003). In that report, the
number of motile sperm decreased by 0.7% per year of age.
The purpose of this study was to investigate the effects of
age on sperm, quantitatively, using computer-assisted semen
analysis (CASA) to measure 14 aspects of sperm numbers,
motility and sperm head morphometry and to provide direct
insight into the specific kinematic components of sperm
motion that are affected by male age.
Materials and methods
Study population
Our study population consisted of 97 healthy male volunteers, aged
22–80 years, who were recruited for the Age and Genetic Effects on
Sperm (AGES) study between October 1997 and July 1998 among active employees/retirees at the Lawrence Livermore National Laboratory
(LLNL, Livermore, CA, USA). This workforce was relatively homogeneous and healthy, with all participants enrolled in comprehensive
2868 © The Author 2006. Published by Oxford University Press on behalf of the European Society of Human Reproduction and Embryology. All rights reserved.
For Permissions, please email: [email protected]
Sperm motility parameters and male age
health care programmes (Eskenazi et al., 2003). Recruitment was
conducted by the staff at the University of California at Berkeley
(UCB, Berkeley, CA, USA) to maintain donor anonymity, and the
LLNL housed the sperm analysis laboratory. The AGES study was
approved by the Institutional Review Board of each participating
institution, and all participants provided written consent. Potential
participants were screened by telephone using the following exclusion criteria: current or prior fertility or reproductive problems;
smoked cigarettes in the last 6 months; vasectomy; reported history
of undescended testicle or prostate cancer; chemotherapy or radiation treatments for cancer; or a previous semen analysis with zero
sperm count. If men at screening had had a fever over 101°F in the
prior 3 months, their appointment date was moved to 3 months from
the date of the fever. At least 15 men were enrolled for each age decade between 20 and 70 years; additional men aged >70 years were
also enrolled. Eligible men were mailed a questionnaire on medical
and reproductive history, sociodemographic characteristics (age,
race and education), occupation, possible exposures, diet and lifestyle habits. Completed questionnaires were mailed to UCB and
reviewed with the participant over the telephone. Men were
requested to provide a semen sample by masturbation into the collection container after abstinence for 2–7 days and to deliver it
inside an insulated container to a drop box at the LLNL, noting the
actual duration of abstinence.
CASA analyses
Coded specimens were delivered within 2 h of collection to the semen
analysis laboratory at the LLNL. CASA was performed using the
HTM-Ceros semen analyzer (Hamilton Thorne Research, USA)
according to the manufacturer’s operation guidelines. Semen samples
were maintained at room temperature until analysed. Fifty microlitres
of each sample were diluted 1:1 using Dulbecco’s phosphate-buffered
saline solution (DPBS) with 1 g/L of glucose and 0.3 g/l of bovine
serum albumin (BSA) (WHO, 1992). All samples with >70 × 106/mL
sperm were diluted to a standardized sperm concentration of ∼35 ×
106/mL sperm. About 3–4 μL of diluted semen was pipetted into one
side of a 2X-CEL 20-μm-depth chambered microscope slide (Hamilton
Thorne Research) maintained at 37°C by a MiniTherm slide warmer
(Hamilton Thorne Research). After 1 min, multiple microscope fields
spanning the entire slide preparation were analysed on an Olympus
CH30 microscope equipped with a 10× negative phase objective for
0.5 s per field using a video frame rate of 60 Hz. When possible, at
least 150 motile sperm were evaluated per drop of semen. Between 2
and 4 drops of diluted semen were evaluated per donor, and the
median value was used for data analyses. Each CASA field was
archived onto videotape as backup using a Sony Hi-Fi Stereo VCR
(model SLV-975HF) and a Videonics TitleMaker 2000.
The 14 CASA measurements used in this study are listed in Tables I
and II. Two measures of sperm number (concentration and total
Table I. Computer-assisted semen analysis measurements of sperm motility and numbers by age decade among 90 healthy non-smoking donors (the California
Age and Genetic Effects on Sperm study)
Age group (years)
Number of donors
MOTa (%)
RAPb (%)
PROc (%)
Concentration (106/ml)
Total count (106)
20–29
30–39
40–49
50–59
60–69
70–80
All donors
P value for aged
19
20
16
17
13
5
90
56.0 (32–64)
42.8 (28–53)
38.5 (14–47)
37.5 (7–49)
20.8 (12–39)
3.0 (2–5)
38.0 (12–53)
<0.0001
31.5 (17–41)
27.3 (22–36)
25.8 (10–33)
20.5 (4–33)
12.8 (5–25)
1.5 (2–3)
25.0 (6–34)
<0.0001
22.5 (12–30)
21.0 (14–26)
16.5 (6–23)
14.5 (2–23)
9.3 (4–18)
1.0 (0.4–2)
17.0 (5–34)
<0.0001
96.5 (55–146)
81.2 (68–104)
109.8 (70–226)
100.0 (70–185)
107.9 (43–198)
55.5 (30–270)
92.2 (58–172)
0.8
289.8 (118–489)
289.6 (207–445)
398.5 (219–915)
232.3 (126–286)
179.3 (153–315)
55.5 (43–987)
255.0 (151–454)
0.1
Median values are presented (interquartile ranges 25th–75th percentile shown in parentheses). MOT, per cent motile; PRO, per cent progressive motility; RAP, per
cent rapid motility.
a
Slow sperm (<5 μm/s) were counted as motile. Thus, % motile sperm equals the number of sperm with any head movement divided by total sperm analysed × 100.
b
Per cent rapid sperm equals the number of sperm with VAP ≥ 25 μm/s divided by the total number of sperm analysed × 100.
c
Per cent progressive sperm equals the number of sperm with VAP ≥ 25 μm/s and STR ≥ 80% divided by the total number of sperm analysed × 100.
d
Based on linear regression analyses, after adjusting for a history of urinary tract infections, time from collection to analysis, past smoking history and body mass
index (BMI).
Table II. Computer-assisted semen analysis measurements of sperm kinematics, size and shape by age decade among 90 healthy non-smoking men (the California
Age and Genetic Effects on Sperm study)
Age group
(years)
20–29
30–39
40–49
50–59
60–69
70–80
All donors
P value for agec
Number Sperm motion parametersa
of donors
LIN
VSL
(VSL/VCL) (μm/s)
19
20
16
17
13
5
90
57.5 (55–63)
58.0 (56–62)
54.3 (50–58)
53.0 (50–61)
54.8 (51–60)
50.0 (45–57)
56.0 (51–60)
0.005
49.1 (47–54)
51.0 (44–57)
45.9 (42–56)
41.3 (36–51)
52.0 (44–56)
32.7 (27–48)
47.9 (40–55)
0.03
Morphometry parametersb
VAP
(μm/s)
STR
(VSL/VAP)
VCL
(μm/s)
BCF
(Hz)
ALH
(μm)
SIZ
ELO
55.8 (53–61)
60.4 (51–67)
55.2 (50–64)
48.8 (45–58)
59.9 (50–63)
38.6 (35–55)
55.5 (49–63)
0.04
83.5 (83–86)
83.3 (82–86)
83.0 (78–85)
82.0 (77–86)
84.4 (80–86)
81.5 (73–85)
83.5 (80–86)
0.1
80.7 (77–87)
86.3 (74–96)
86.7 (78–94)
77.4 (71–87)
88.8 (76–100)
68.7 (53–87)
82.8 (75–92)
0.1
25.2 (23–28)
24.5 (23–27)
24.8 (23–26)
23.9 (21–26)
25.4 (24–27)
21.6 (19–26)
24.8 (23–27)
0.4
3.7 (3.4–4.3)
3.7 (3.4–4.1)
4.1 (3.9–4.7)
3.6 (3.4–3.7)
4.1 (3.6–4.6)
3.9 (3.2–4.4)
3.8 (3.4–4.3)
0.9
7.2 (6.6–7.3)
7.4 (7.0–7.6)
7.4 (6.9–7.7)
7.1 (6.4–7.6)
7.1 (6.6–7.3)
7.2 (6.9–7.3)
7.2 (6.7–7.6)
0.4
67.0 (65–70)
66.3 (65–67)
67.3 (66–70)
68.0 (65–69)
66.3 (65–69)
64.0 (60–67)
67.0 (65–69)
0.4
ALH, amplitude of lateral head; BCF, beat cross frequency; LIN, linearity; STR, straightness of trajectory; VCL, curvilinear velocity; VSL, straight line velocity.
See Figure 1 for definitions of kinematic parameters.
b
See Materials and methods for the definition of SIZ and ELO parameters. Median values (interquartile ranges 25th–75th percentile shown in parentheses).
c
Based on linear regression analyses, after adjusting for a history of urinary tract infections, time from collection to analysis, past smoking history and body mass
index (BMI).
a
2869
E.Sloter et al.
VAP*
BCF
ALH
VCL
VSL*
Calculated parameters:
LIN* = VSL*/VCL x 100
STR = VSL*/VAP* x 100
Figure 1. Sperm kinematic parameters measured by computerassisted semen analysis (CASA). See Methods section for additional
description. Asterisk denotes the parameters affected by age in the
current study. ALH, amplitude of lateral head; BCF, beat cross frequency; LIN, linearity; STR, straightness of trajectory; VCL, curvilinear velocity; VSL straight line velocity; VAP, time average
velocity.
count) and three measures of gross motility [per cent motile (MOT),
per cent progressive motility (PRO) and per cent sperm with rapid
motility (RAP)] were used. MOT includes RAP plus PRO. Per cent
rapid sperm equals the number of sperm with time-average velocity
(VAP) ≥ 25 μm/s divided by the total number of sperm analysed ×
100. Per cent progressive sperm equals the number of sperm with
time-average velocity (VAP) ≥ 25 μm/s and straightness of trajectory
(STR) ≥ 80% divided by the total number of sperm analysed × 100.
The seven kinematic measurements used in our study were described
by Rijsselaere et al. (2003) and are shown in Figure 1. Straight line
velocity (VSL) is the time-average velocity of a sperm head along the
straight line between its first and last detected positions. Curvilinear
velocity (VCL) is the average velocity measured over the actual
point-to-point track followed by the cell. The time-average velocity
(VAP) measures the sperm head along its spatial average trajectory
(i.e. smoothed version of VCL). Linearity (LIN) is the linearity of the
curvilinear trajectory calculated as VSL/VCL × 100. STR is the LIN
of the sperm average path calculated as VSL/VAP × 100. Amplitude
of lateral head displacement (ALH) is the maximum lateral displacement of a sperm head about its spatial average trajectory (i.e. track
width). Beat cross frequency (BCF) is the time-average rate at which
the curvilinear sperm track crosses its average path trajectory. The
elongation ratio of minor to major axis of each sperm nucleus (ELO)
and the size of each sperm nucleus (SIZ) in square microns were also
measured.
Statistical analyses
Regression analyses were conducted using StatView 5.0 (SAS Institute, Cary, NC, USA). The CASA motility parameters were approximately normally distributed. Sperm counts and concentrations by
CASA were log-normally distributed. Separate multiple linear regression models were used to examine the relationships of age with each
of the CASA parameters after controlling for potential confounding
factors. The following variables and classes of variables were considered as potential covariates: time from sample collection to sample
processing; duration of sexual abstinence before semen collection;
exposure to occupational chemicals and radiation; history of working
with radioisotopes; history of tobacco use; alcohol and caffeine
intake; prescription and non-prescription medication use; history of
2870
chronic diseases such as high blood pressure, heart problems or diabetes; history of genitourinary disease including urinary tract or other
genitourinary infection, sexually transmitted disease, varicocele or a
history of infertility; fatherhood history; BMI; ethnicity; and dietary
supplement usage. Covariates were included in the multivariate models if they changed the regression coefficient by >10% or if they were
statistically significant at P < 0.1. Comparisons between CASA and
conventional results were performed by linear regression analysis.
Linear regression analysis was also used to test for possible dilution
effects on the sperm motility parameters.
Results
Characteristics of study population
The 97 male volunteers recruited for the AGES study were on
average 46.4 years of age (SD = 15.8 years, range = 22–80
years) and well distributed among the age decades (Table I).
Visual microscopic analyses identified seven men, all of whom
were >60 years, who were either azoospermic or had no motile
sperm, thereby reducing to 90 the number of men evaluated by
CASA. Two of the four azoospermic men and all of the men
with zero sperm motility had fathered children earlier in life,
and none had been diagnosed with fertility problems in their
past. The aetiology of these defects was not evident even from
their questionnaires but is possibly age related.
By CASA, the median sperm concentration was 92 × 106/ml
sperm, and total sperm count was 255 × 106 sperm; 38% of sperm
were motile (MOT), 25% were rapidly motile (>25 μm/s, RAP)
and 17% were progressively motile (i.e. rapid and linear, PRO).
These CASA measures were highly correlated with the conventional semen quality data among the same subjects (Eskenazi
et al., 2003): per cent motile sperm, R2 = 0.94, Figure 2A; per
cent progressively motile sperm, R2 = 0.88, Figure 2B; and
sperm concentration, R2 = 0.85, Figure 2C. There was a 1:1
within-specimen correspondence between the conventional
and CASA measurements for per cent motile and sperm concentration, but conventional analysis detected ∼30% more progressively motile sperm than did CASA.
Effects of age
CASA analyses did not detect an age effect on sperm concentration or total count per specimen but did detect an effect on
MOT, RAP and PRO (Table I). Regression analyses showed
age-related decreases, with no indication of a threshold (Figure
3A–C). The variation among individuals was high (Tables I and
II) but not related to age. Age remained strongly associated with
MOT (P < 0.0001), RAP (P < 0.0001) and PRO (P < 0.0001),
after adjusting for a history of urinary tract infections, time
from collection to analysis, smoking history and BMI. Regression analyses showed that MOT, RAP and PRO decreased 0.8,
0.4 and 0.9% per year of age, P < 0.001 (Figure 3A–C). Abstinence was not a confounder of the effects of age on CASAdetermined motility (P >0.1). Men in their 20s averaged 31.5%
RAP sperm compared with 20.5% for men in their 50s and
1.5% among men in their 70s.
Table II and Figure 3D–F summarize the effects of age on the
seven CASA kinematic measurements described in Figure 1.
Linear regression modelling showed significant age-related
Sperm motility parameters and male age
A. % motile sperm
100
90
80
70
CASA
60
50
40
30
20
y = 1.03x - 4.48
10
R2 = 0.94
0
0
10
20
30
40
50
60
70
80
90 100
Conventional
B. % progressive sperm
70
60
CASA
50
40
decreases in average path velocities (VAP, 0.3% per year, P =
0.04), VSL (0.2% per year, P = 0.03) and LIN (0.2% per year,
P = 0.005). There was no evidence of an age-related threshold,
illustrated visually in Figure 3 and confirmed using non-linear
modelling (data not shown). Although variation was high
among individuals, it was not related to age (Table II). Omitting the five donors aged >69 years did not eliminate the statistical significance of these findings (VAP: P = 0.03; VSL: P =
0.02; LIN: P = 0.01). Table III provides the values for these
sperm parameters for 25- versus 55-year-old men, as predicted
from the linear regression analyses (Figure 3).
We found no evidence of an association between age and
lateral head amplitude (ALH), BCF, STR or VCL (P > 0.1).
There was also no evidence of an association between age and
nuclear size or elongation ratio (Table II).
The associations between age and CASA-determined motility parameters were not affected by the DPBS/glucose diluent
used. Sperm concentration, which dictated the amount of diluent used (see Methods), was not associated with CASAdetermined motility (Figure 4); i.e. younger men did not have
significantly higher sperm concentrations than older men (P =
0.3, Table I). Also, the proportion of diluent used (i.e. the dilution factor) was not associated with CASA-determined motility
measurements as shown in Figure 5 for %motility (r = 0.03,
P > 0.1) nor with changes in PRO, RAP, VAP or LIN (P > 0.1).
30
Correlations among CASA parameters
20
y = 0.71x + 0.75
10
R2 = 0.88
0
0
10
20
30
40
50
60
70
Conventional
C. Sperm concentration
CASA
1000
Table IV summarizes the correlations among 14 CASA parameters for the 90 men in this study. The three measurements of
motility, MOT, PRO and RAP, were highly inter-correlated (P
< 0.01). VSL and VAP were highly correlated with each other
(P < 0.01), highly correlated with VCL (P < 0.01), moderately
correlated with MOT and PRO (P < 0.05) and moderately correlated with STR (P < 0.05), but neither measure is significantly
correlated with LIN. However, STR and LIN are highly correlated (P < 0.01). As summarized in Table IV, there was no correlation between the nuclear elongation ratio and any of the
sperm motility (MOT, PRO and RAP) or kinematic parameters.
Discussion
100
y = 0.85x + 11.81
R2 = 0.85
10
10
100
1000
Conventional
Figure 2. Comparison of the computer-assisted semen analysis
(CASA) and conventional semen quality data for per cent motile
sperm (A), per cent progressive sperm (B) and sperm concentration
(C). Each data point represents analyses of a single sperm sample
from each of the 90 healthy male volunteers evaluated using the
HTM-Ceros analyzer and conventional semen assays (see Methods
section for details). Linear regression lines are shown along with the
coefficients of determination (R2) and β-coefficients. Sperm concentrations were log-normally distributed.
CASA provides quantitative measurements of three general
aspects of sperm motility as well as specific kinematic measures of sperm motion; our study is the first to use CASA to
assess the effects of age on sperm kinematics. We present
quantitative evidence of age-associated reductions in the proportion of motile sperm and their straight line motion. Regression analyses indicate that there are no age-related thresholds
for the sperm motility (MOT, RAP and PRO) and sperm kinematic (LIN, VSL and VAP) end-points, consistent with a gradual decline in sperm motion with advancing male age.
The major effects of advancing male age were on the highly
correlated CASA-determined sperm motility parameters
(MOT, RAP and PRO) and on three sperm kinematic parameters (LIN, VSL and VAP). MOT, RAP and PRO have been
associated with decreased pregnancy rates and longer time to
pregnancy in partners of older men (Abramsson, 1988). VSL,
LIN and VAP have been correlated with fertilization rates
2871
E.Sloter et al.
y = –0.76x + 69.2
R2 = 0.25
A
y = –0.16x + 61.9
R2 = 0.06
D
90
80
80
70
70
60
60
50
50
40
30
40
20
30
10
0
20
20
30
40
50
60
70
80
20
30
40
Age
50
60
70
80
Age
y = –0.39x + 34.0
R2 = 0.22
B
y = –0.17x + 54.7
R2 = 0.06
E
60
80
50
70
40
60
30
50
20
40
10
30
20
0
20
30
40
50
60
70
80
20
30
40
Age
50
60
70
80
Age
y = –0.52x + 46.2
R2 = 0.23
C
y = –0.1169x + 61.333
R2 = 0.09
F
70
70.0
60
65.0
50
60.0
40
55.0
30
50.0
20
45.0
10
40.0
0
20
30
40
50
60
70
80
20.0
30.0
40.0
50.0
60.0
70.0
80.0
Age
Age
Figure 3. The relationship between age in years and computer-assisted semen analysis (CASA) motility parameters for per cent motile (A), per
cent progressive motility (B) and per cent rapid motility (C) and for the CASA sperm kinematic parameters VAP (D), straight line velocity (E)
and linearity (F). Linear regression lines are shown along with the coefficients of determination (R2) and β coefficients. Each data point represents
the median value of repeated measurements from each donor sample.
Table III. Comparison of predicted values for sperm motion parameters
among 25- and 55-year-old men
Computer-assisted semen
analysis motion parameter
25 years of age
55 years of age
VAP**
VSL**
VCL*
ALH
BCF
STR*
LIN***
58.0
50.5
85.5
3.8
25.4
83.7
58.3
53.2
45.4
81.1
3.8
25.3
82.1
54.7
ALH, amplitude of lateral head; BCF, beat cross frequency; LIN, linearity;
STR, straightness of trajectory; VCL, curvilinear velocity; VSL straight line
velocity. Based on linear regression analyses, after adjusting for a history of
urinary tract infections, time from collection to analysis, past smoking history
and BMI. *P < 0.1, **P < 0.05, ***P < 0.01 for linear regression analyses (see
Table II).
2872
in vivo and may be bioindicators of the fertilizing ability of
human sperm (Hirano et al., 2001). These age-related reductions in sperm motility predict fertility problems for men who
choose to delay attempting fatherhood and may provide a
mechanistic explanation for previous findings of decreased
pregnancy rates and longer times to pregnancy in partners of
older men (Abramsson, 1988; Ducot et al., 1988; Brzechffa
et al., 1998; Spandorfer et al., 1998).
Inspection of the sperm kinematic parameters affected by age
versus those that were not (Table II, Figure 3) leads to a quantitative model of the effects of male ageing on motile sperm.
Using 25 versus 55 years of age as an example (Table III), the
age-related decline in VSL predicts that the sperm of older men
travel ∼10% shorter overall distance per unit time (50.5 versus
45.4 μm/s). The age-related declines for VAP and VCL suggest that sperm of older men also travel ∼9% slower along their
average path and ∼6% slower along the point–point path of
Sperm motility parameters and male age
Sperm concentration
1000
100
y = 0.91x + 104.05
2
R = 0.03
10
0
20
40
60
80
100
Motility (MOT)
Figure 4. Relationship between sperm concentration (×106/ml;
logarithmic scale) and computer-assisted semen analysis (CASA)determined per cent motile sperm. Individual data points as well as
the linear regression line (r = 0.02; P > 0.1) are shown.
30
y = 0.01x +
4.12
Dilution factor
25
2
R = 0.002
20
15
10
5
0
0
20
40
60
80
100
Motility (MOT)
Figure 5. Relationship between dilution factor and per cent motility.
Individual data points as well as the linear regression line (r = 0.02;
P > 0.1) are shown.
travel (Figure 3). The lack of a significant age effect on ALH
and BCF suggests that age does not alter the frequency at
which sperm cross the average path (∼25 times per second) nor
the amplitude of deviation around that average path (∼3.8 μm).
Taken together with the age-related decrease in LIN and a trend
towards a decrease in STR, our model suggests that sperm of
older men follow a less linear path (i.e. more curved) than sperm
of younger men, which is consistent with our visual assessments
of these donors using conventional semen analyses with phasecontrast light microscopy (Eskenazi et al., 2003).
Our study did not find an age effect on ALH, and ALH was
not correlated with any other motion or kinematic parameter
(Table IV). ALH is often used to estimate IVF outcome
(Barlow et al., 1991). ALH measures the vigour of flagellar
beating in conjunction with the frequency of cell rotation
(Verstegen et al., 2002), which has been associated with the
ability of sperm to penetrate cervical mucus and fuse with
oocytes. Taking together, our CASA findings suggest that
although advancing age may not directly diminish the ability of
sperm to penetrate and fuse with oocytes, older men will produce fewer sperm that reach the oocyte as a consequence of
reduced per cent of motile sperm and reduced LIN of motion
among the remaining motile sperm.
In our study, CASA measurements of overall sperm motility (MOT, PRO and RAP) compared well with the corresponding conventional measures. However, we had reported
a small age effect on both sperm concentration and counts by
conventional analysis (Eskenazi et al., 2003) but did not find
this by CASA analyses (Figure 2). This might be explained in
part by the observation of Verstegen et al. (2002) that conventional sperm counts are more reliable than those by CASA
analyses at low sperm concentration. This may be caused by
the high-background seminal debris particles of similar size
and luminosity present in human semen that would be counted
as immotile sperm heads (Verstegen et al., 2002).
Our study design has several notable strengths. Our study
has significantly larger number of older men than any previous
study (Kidd et al., 2001). Recruitment for our study was limited to men drawn from a single large occupational setting;
men with ill health or with known infertility-related problems
were excluded before analyses. The cohort was relatively
homogeneous, consisting of employed and retired workers of
middle to high socioeconomic class with employer-paid access
to medical care, unlike almost all previous studies of age
effects on conventional semen quality parameters (Homonnai
et al., 1982; Nieschlag et al., 1982; Schwartz et al., 1983;
Dondero et al., 1985; Abramsson, 1988; Check et al., 1989;
Singer et al., 1990; Carlsen et al., 1992; Gallardo et al., 1996;
Haidl et al., 1996; Irvine et al., 1996; Rolf et al., 1996; Berling
and Wolner-Hanssen, 1997; Lemcke et al., 1997; Spandorfer
et al., 1998; Kidd et al., 2001). Our study provided detailed
information on a wide range of potentially confounding factors
(abstinence, time between sample collection and sample processing, smoking habits, etc.), and regression analyses were
adjusted in accordance with identified covariates. The statistical power to detect an association between semen parameters
and age was improved by the addition of larger numbers of
older men, compared with previous studies (Abramsson, 1988;
Check et al., 1989; Gallardo et al., 1996; Irvine et al., 1996;
Spandorfer et al., 1998).
Our findings are consistent with those of Singh et al. (2003),
who reported a relation between age and reduced motility in
sperm of men aged 20–57 years by conventional semen analysis, but their study was of a mixed population of patients
from an infertility clinic and a non-clinical group. In contrast,
Chen et al. (2004) did not find a statistically significant relationship between age and semen parameters, but their study
was limited to men up to 54 years of age and they only reported
per cent motile sperm.
Age-related cellular or physiologic changes in the reproductive tract, hormonal changes or increased oxidative damage
2873
E.Sloter et al.
Table IV. Correlation among the computer-assisted semen analysis parameters of sperm motility, kinematics, size and shape among 90 healthy non-smoking men
MOT
PRO
RAP
VAP
VSL
VCL
ALH
BCF
STR
LIN
ELO
SIZ
MOT
PRO
RAP
VAP
VSL
VCL
ALH
BCF
STR
LIN
ELO
SIZ
1.00
0.72**
0.98**
0.54*
0.54*
0.44
0.14
0.20
0.39
0.39
−0.08
0.43
1.00
0.98**
0.60*
0.62*
0.48
0.09
0.25
0.46
0.48
−0.07
0.45
1.00
0.56*
0.56*
0.47
0.15
0.17
0.38
0.39
−0.05
0.44
1.00
0.98**
0.92**
0.28
0.37
0.57*
0.62
−0.19
0.37
1.00
0.86**
0.19
0.45
0.69*
0.71
−0.22
0.41
1.00
0.53
0.23
0.38
0.30
−0.17
0.38
1.00
−0.25
−0.16
−0.41
0.01
0.18
1.00
0.59
0.52
−0.38
0.51
1.00
0.79**
−0.31
0.46
1.00
−0.12
0.21
1.00
−0.39
1.00
ALH, amplitude of lateral head; BCF, beat cross frequency; LIN, linearity; MOT, per cent motile; PRO, per cent progressive motility; RAP, per cent rapid motility;
STR, straightness of trajectory; VCL, curvilinear velocity; VSL straight line velocity. **P ≤ 0.01, *P ≤ 0.05.
may be primary events leading to decreased motility in older
men. Age effects on the prostate may contribute to changes in
seminal plasma which may affect sperm motility (Schneider
and Monticone, 1978). There may also be age-related changes
in the epididymis, where sperm acquire the capacity for vigorous forward motility during transit (Hamilton and Naftolin,
1981). Also, the epididymis is a hormonally sensitive tissue
that plays an important role in sperm maturation. Reactive oxygen species (ROS) play a role in male infertility, where excessive amounts impair spermatozoal motility. Ageing results in
histological and biochemical changes in the epididymis that are
suggestive of oxidative damage (Robaire and Hales, 2003). A
recent article showed that higher antioxidant intake over the
normal dietary and supplement use range was associated with
higher motility in the same set of men used in our study
(Eskenazi et al., 2005). Also, antioxidant intake attenuates
the adverse effect of age on sperm motility. Further studies
are warranted to determine the cellular damage pathways that
lead to detriment in some kinematic parameters and not in
others.
Our findings predict that the average man may become progressively less fertile as he ages due to reduced proportions of
motile sperm and decreased abilities of motile sperm to
maintain forward motion along a linear path. However, the
numerical relationships between the per cent changes in CASAdetermined motility parameters and the magnitude of decline of
male fertility remain to be established. Our data also suggest
that, unlike fertility in women, there appears to be no evidence of
an age threshold for these sperm parameters for men, but rather a
gradual decline with advancing age throughout normal reproductive years and into senescence. Thus, our findings may have
important implications for men who choose to delay fatherhood,
because age may reduce their chance for success.
Acknowledgements
This work was performed under the auspices of the US Department of
Energy by the LLNL contract W-7405-ENG-48, with funding from
Superfund P42 ES04705 from the National Institute of Environmental
Health Sciences (Eskenazi and Wyrobek, study directors).
2874
References
Abramsson L (1988) On the investigation of men from infertile relations. A
clinical study with special regard to anamnesis, physical examination,
semen-, hormone- and chromosome analyses, from men with non-“normal”
semen. Scand J Urol Nephrol Suppl 113,1–47.
Barlow P, Delvigne A, Van Dromme J, Van Hoeck J, Vandenbosch K and
Leroy F (1991) Predictive value of classical and automated sperm analysis
for in-vitro fertilization. Hum Reprod 6(8),1119–1124.
Berling S and Wolner-Hanssen P (1997) No evidence of deteriorating semen
quality among men in infertile relationships during the last decade: a study
of males from Southern Sweden. Hum Reprod 12(5),1002–1005.
Brzechffa PR, Daneshmand S and Buyalos RP (1998) Sequential clomiphene
citrate and human menopausal gonadotrophin with intrauterine insemination:
the effect of patient age on clinical outcome. Hum Reprod 13(8),2110–2114.
Carlsen E, Giwercman A, Keiding N and Skakkebaek NE (1992) Evidence for
decreasing quality of semen during past 50 years. BMJ 305(6854),609–613.
Check JH, Shanis B, Bollendorf A, Adelson H and Breen E (1989) Semen
characteristics and infertility in aging. Arch Androl 23(3),275–277.
Chen Z, Godfrey-Bailey L, Schiff I and Hauser R (2004) Impact of seasonal
variation, age and smoking status on human semen parameters: The Massachusetts General Hospital experience. J Exp Clin Assist Reprod 1(1),2.
Dondero F, Mazzilli F, Giovenco P, Lenzi A and Cerasaro M (1985) Fertility
in elderly men. J Endocrinol Invest 8(Suppl 2),87–91.
Ducot B, Spira A, Feneux D and Jouannet P (1988) Male factors and the likelihood of pregnancy in infertile couples. II. Study of clinical characteristics—
practical consequences. Int J Androl 11(5),395–404.
Eskenazi B, Wyrobek AJ, Sloter E, Kidd SA, Moore L, Young S and Moore D
(2003) The association of age and semen quality in healthy men. Hum
Reprod 18(2),447–454.
Eskenazi B, Kidd SA, Marks AR, Sloter E, Block G and Wyrobek AJ (2005)
Antioxidant intake is associated with semen quality in healthy men. Hum
Reprod 20(4),1006–1012.
Gallardo E, Simon C, Levy M, Guanes PP, Remohi J and Pellicer A (1996)
Effect of age on sperm fertility potential: oocyte donation as a model. Fertil
Steril 66(2),260–264.
Haidl G, Jung A and Schill WB (1996) Ageing and sperm function. Hum
Reprod 11(3),558–560.
Hamilton D and Naftolin F (1981) Basic Reproductive Medicine. MIT Press,
Cambridge, MA.
Hirano Y, Shibahara H, Obara H, Suzuki T, Takamizawa S, Yamaguchi C,
Tsunoda H and Sato I (2001) Relationships between sperm motility characteristics assessed by the computer-aided sperm analysis (CASA) and fertilization rates in vitro. J Assist Reprod Genet 18(4),213–218.
Homonnai ZT, Fainman N, David MP and Paz GF (1982) Semen quality and
sex hormone pattern of 29 middle aged men. Andrologia 14(2),164–170.
Irvine S, Cawood E, Richardson D, MacDonald E and Aitken J (1996) Evidence of deteriorating semen quality in the United Kingdom: birth cohort
study in 577 men in Scotland over 11 years. BMJ 312(7029),467–471.
Kidd SA, Eskenazi B and Wyrobek AJ (2001) Effects of male age on semen
quality and fertility: a review of the literature. Fertil Steril 75(2),237–248.
Sperm motility parameters and male age
Lemcke B, Behre HM and Nieschlag E (1997) Frequently subnormal semen
profiles of normal volunteers recruited over 17 years. Int J Androl
20(3),144–152.
Nieschlag E, Lammers U, Freischem CW, Langer K and Wickings EJ (1982)
Reproductive functions in young fathers and grandfathers. J Clin Endocrinol
Metab 55(4),676–681.
Rijsselaere T, Van Soom A, Maes D and de Kruif A (2003) Effect of technical
settings on canine semen motility parameters measured by the HamiltonThorne analyzer. Theriogenology 60(8),1553–1568.
Robaire B and Hales BF (2003) Mechanisms of action of cyclophosphamide
as a male-mediated developmental toxicant. Adv Exp Med Biol 518,
169–180.
Rolf C, Behre HM and Nieschlag E (1996) Reproductive parameters of older
compared to younger men of infertile couples. Int J Androl 19(3),135–142.
Schneider EL and Monticone RE (1978) Aging and sister chromatid exchange.
II. The effect of the in vitro passage level of human fetal lung fibroblasts on
baseline and mutagen-induced sister chromatid exchange frequencies. Exp
Cell Res 115(2),269–276.
Schwartz D, Mayaux MJ, Spira A, Moscato ML, Jouannet P, Czyglik F and
David G (1983) Semen characteristics as a function of age in 833 fertile
men. Fertil Steril 39(4),530–535.
Singer R, Sagiv M, Levinsky H and Allalouf D (1990) Andrological parameters in men with high sperm counts and possible correlation with age. Arch
Androl 24(2),107–111.
Singh NP, Muller CH and Berger RE (2003) Effects of age on DNA
double-strand breaks and apoptosis in human sperm. Fertil Steril
80(6),1420–1430.
Sloter E, Nath J, Eskenazi B and Wyrobek AJ (2004) Effects of male age on
the frequencies of germinal and heritable chromosomal abnormalities in
humans and rodents. Fertil Steril 81(4),925–943.
Spandorfer SD, Avrech OM, Colombero LT, Palermo GD and Rosenwaks Z
(1998) Effect of parental age on fertilization and pregnancy characteristics
in couples treated by intracytoplasmic sperm injection. Hum Reprod
13(2),334–338.
Templeton A (1995) Infertility-epidemiology, aetiology and effective management. Health Bull (Edinb) 53(5),294–298.
Ventura SJ, Peters KD, Martin JA and Maurer JD (1997) Births and deaths:
United States, 1996. Mon Vital Stat Rep 46(1 Suppl 2),1–40.
Verstegen J, Iguer-Ouada M and Onclin K (2002) Computer assisted semen
analyzers in andrology research and veterinary practice. Theriogenology
57(1),149–179.
WHO (1992) WHO Laboratory Manual for the Examination of Human Semen
and Sperm-Cervical Mucus Interaction. Cambridge University Press, New
York, pp. 1–107.
Submitted on September 20, 2005; resubmitted on January 20, 2006, May 5,
2006; accepted on May 19, 2006
2875