Intelligence and Semen Quality are Positively Correlated

Abstract

Human cognitive abilities inter-correlate to form a positive matrix, from which a large first factor, called ‘Spearman's g’ or general intelligence, can be extracted. General intelligence itself is correlated with many important health outcomes including cardio-vascular function and longevity. However, the important evolutionary question of whether intelligence is a fitness-related trait has not been tested directly, let alone answered. If the correlations among cognitive abilities are part of a larger matrix of positive associations among fitness-related traits, then intelligence ought to correlate with seemingly unrelated traits that affect fitness—such as semen quality. We found significant positive correlations between intelligence and 3 key indices of semen quality: log sperm concentration (r = .15, p = .002), log sperm count (r = .19, p < .001), and sperm motility (r = .14, p = .002) in a large sample of US Army Veterans. None was mediated by age, body mass index, days of sexual abstinence, service in Vietnam, or use of alcohol, tobacco, marijuana, or hard drugs. These results suggest that a phenotype-wide fitness factor may contribute to the association between intelligence and health. Clarifying whether a fitness factor exists is important theoretically for understanding the genomic architecture of fitness-related traits, and practically for understanding patterns of human physical and psychological health.

Full text

Intelligence and semen quality are positively correlated
Rosalind Arden
a,
⁎, Linda S. Gottfredson
b
, Geoffrey Miller
c
, Arand Pierce
d
a
Social, Genetic, Developmental & Psychiatry Centre, Institute of Psychiatry, King's College London, London SE5 8AF, United Kingdom
b
School of Education, University of Delaware, Newark, DE 19716 USA
c
Psychology Department, Logan Hall 160, University of New Mexico, MSC03 2220 Albuquerque, NM 87131 USA
d
Department of Pathology, School of Medicine, University of New Mexico, MSCO8 4640, Albuquerque, NM 87131 USA
article info abstract
Article history:
Received 3 July 2008
Revised 10 September 2008
Accepted 1 November 2008
Available online xxxx
Human cognitive abilities inter-correlate to form a positive matrix, from which a large first
factor, called ‘Spearman's g’or general intelligence, can be extracted. General intelligence itself
is correlated with many important health outcomes including cardio-vascular function and
longevity. However, the important evolutionary question of whether intelligence is a fitness-
related trait has not been tested directly, let alone answered. If the correlations among cognitive
abilities are part of a larger matrix of positive associations among fitness-related traits, then
intelligence ought to correlate with seemingly unrelated traits that affect fitness—such as
semen quality. We found significant positive correlations between intelligence and 3 key
indices of semen quality: log sperm concentration (r=.15, p=.002), log sperm count (r=.19,
pb.001), and sperm motility (r=.14, p=.002) in a large sample of US Army Veterans. None
was mediated by age, body mass index, days of sexual abstinence, service in Vietnam, or use of
alcohol, tobacco, marijuana, or hard drugs. These results suggest that a phenotype-wide fitness
factor may contribute to the association between intelligence and health. Clarifying whether a
fitness factor exists is important theoretically for understanding the genomic architecture of
fitness-related traits, and practically for understanding patterns of human physical and
psychological health.
© 2008 Published by Elsevier Inc.
Keywords:
Intelligence
Fitness
g
Semen quality
Sperm
Fertility
1. Introduction
The new field of ‘cognitive epidemiology’is emerging from
the surprising discovery that intelligence correlates with many
important health outcomes, even longevity (Batty, Deary, &
Gottfredson, 2007). These correlations may be mediated by
lifestyle factors such as eating well, exercising, avoiding
cigarettes, working in safer, less stressful jobs, and having
better access to health care. However, there is a further
possibility, suggested by the nature of intelligence itself: all
cognitive abilities, no matter how diverse, inter-correlate
positively, forming a matrix (or ‘manifold’). This commonality
depends primarily on a single underlying factor called
Spearman's g, named for its discoverer—the Britishpsychologist
Charles Spearman (1863–1945). Spearman's gusually accounts
for around half the total variance in any broad battery of tests
(Carroll, 1993). The term ‘intelligence’is often used inter-
changeably with ‘g’(as here) because the behavioural manifes-
tations of gfit well with popular conceptions of intelligence:
“the faculty of understanding, quickness of mental apprehen-
sion”(Little, Fowler, & Coulson,1984).
Could the positive manifold among cognitive abilities be
part of a larger manifold among all fitness-related traits? (By
‘fitness’we mean the statistical propensity to survival and
reproductive success, given ancestrally typical conditions:
fitness-related traits help or harm fitness). If so, the gfactor
may be a special case of a more general ‘fitness factor’or f
factor that captures individual differences in general pheno-
typic quality (Houle, 2000). To investigate this possibility, we
analysed correlations between intelligence and a fitness-
related trait that has little face-value association: semen
qualit y (Bonde et al.,1998; World Health Organization,1999).
These two traits arise from distinct organ systems (brain
Intelligence xxx (2008) xxx–xxx
⁎Corresponding author.
E-mail address: arden.rosalind@gmail.com (R. Arden).
INTELL-00490; No of Pages 6
0160-2896/$ –see front matter © 2008 Published by Elsevier Inc.
doi:10.1016/j.intell.2008.11.001
Contents lists available at ScienceDirect
Intelligence
journal homepage:
ARTICLE IN PRESS
Please cite this article as: Arden, R., et al., Intelligence and semen quality are positively correlated, Intelligence (2008),
doi:10.1016/j.intell.2008.11.001

versus testicles), composed of distinct cells types: neurons
and glia in brains, spermatagonia and Sertoli cells in sperm
(for a cytological comparison of neurons and sperm see
Meizel, 2004).
2. Methods
2.1. Participants
Our sample consists of US Vietnam-era Army veterans
enrolled in the Vietnam Experience Study (VES) by the US
Centers for Disease Control. In 1985 4462 veterans underwent
extensive medical and mental examinations; of these, 571
provided semen samples. Our analyses focused on the 425
men who had not been vasectomised, who collected their
entire ejaculate without spillage, and for whom we had
complete cognitive data for the measures listed below. These
425 men were aged 31 to 44 (mean 37.6, SD 2.5) at time of
testing. 355 were White, 48 Black, 16 Hispanic, 4 Asian or
Pacific Islanders and 2 Native American or Native Alaskan. The
characteristics of the men who participated in the semen
analysis were nearly identical to those of the larger group of
4462 medical examinees (The Centers for Disease Control,
1989 p.216). Comprehensive details of the sampling design
are found in VES project reports (The Centers for Disease
Control, 1989 pp.11–14 and also on the web Centers for
Disease Control, 2007).
We ascertained representativeness further by examining
the General Technical Test (GT) and Armed Forces Qualifica-
tion Test scores obtained at induction (~age 18) for (a) the full
sample of 18,313 veterans in the VES study, (b) the subsample
of 4462 men who were given comprehensive examinations in
1985 (~age 37), and (c), of those, the 425 men in our analyses
who provided semen samples. GT means (and SDs) at the
three levels of sampling were very similar, respectively,104.4
(20.1), 106.0 (20.3), and 106.5 (20.5).
2.2. Semen collection and measures
The men were asked to abstain from ejaculating for at least
48 h before semen collection. They were given plastic
containers for collection and insulating cups to keep the
semen warm. Following masturbation in their hotel rooms
(without using lubricants or condoms), the men delivered
their samples to the VES study receiving desk within 30 min
of ejaculation. They noted the number of days since their most
recent ejaculation. In the mid 1980s, objective, standardized
computerised imaging had become available for measuring
sperm concentration and motility. The Cellsoft computer-
assisted semen analyzer system was used to video-record
semen samples within 140 min of collection. For methodo-
logical details of the semen analysis protocols see (Centers for
Disease Control, 1989 pp.197–199 and Centers for Disease
Control, 2007, Supplement B pp.3–5).
We analysed three semen measures: sperm concentration
(millions of sperm per ml of semen, log
10
transformed to
minimize skew), sperm count (millions of sperm in the total
ejaculate, log
10
transformed to minimize skew), and sperm
motility (percentage of motile sperm). We selected these
three measures because they are associated with fertility
(World Health Organization, 1999). Sperm concentration and
sperm motility are associated with the likelihood of fertiliza-
tion (Guzick et al., 2001; van der Merwe, Kruger, Oehninger, &
Lombard, 2005). Sperm morphology which is also associated
in the andrology literature with the likelihood of conception
was excluded from our analyses because there is ongoing
debate among evolutionary researchers over whether differ-
ent adaptive sperm morphs exist in humans as a response to
sperm competition (see for review Shackelford & Pound,
2006). If there are sperm morphs that are adaptive under
some conditions (such as sperm competition) then knowing
the quantity of morphologically-typical sperm will be only
partially useful, whereas even under sperm competition,
concentration, count and motility would be useful indicators
of fertility.
2.3. Cognitive tests and extraction of the g factor
The 4462 veterans studied in 1985 took 16 neuropsycho-
logical tests. Of these, we selected the 5 tests that provided
the most psychometrically sound measures of the broad
spatial, quantitative, and verbal abilities typically tapped by
IQ test batteries. These were: the Verbal and Arithmetic tests
of the Army Classification Battery (Montague, Williams,
Gieseking, & Lubin, 1957), the Information and Block Design
subtests of the Wechsler Adult Intelligence Scale—Revised
(Wechsler, 1981); and the Reading subtest of the Wide Range
Achievement Test (Jastak & Jastak, 1965). We used principal
axis factoring on these five tests to extract a gfactor (which
explained N60% of the total variance among the test scores).
We calculated gscores for all the men in the sample of 4462,
which included the 425 men in this report, and used those g
scores as the index of each man's intelligence.
2.4. Covariates
We assessed five key covariates that might confound any
observed statistical relation between intelligence and semen
quality. We focused on covariates that might confound our
results by influencing both semen quality and intelligence
since factors that influence only one trait would not explain
the correlation between the two traits.
These covariates were: age (at time of testing in 1985),
reported days of sexual abstinence before ejaculation (‘days
abstinence’), body mass index (BMI), alcohol use (self-
reported alcoholic drinks per month currently), and smoking
(self-reported cigarettes per day smoked currently). Intelli-
gence and semen quality may both decline with age; days
abstinence increases sperm concentration and count (Levitas
et al., 2005); intelligence may predict compliance with VES
instructions to abstain for at least 48 h before semen
collection. Finally, less intelligent men might take poorer
care of their health, resulting in more obesity, alcohol
consumption, and smoking. Cigarette smoking is known to
harm semen quality (Soares & Melo, 2008) as does higher BMI
(Nguyen, Wilcox, Skjaerven, & Baird, 2007). Cigarette smok-
ing, alcohol use and obesity have not been shown to harm
intelligence, but since they are pervasively harmful to health
in many epidemiological studies we included them as
potential confounds.
We also examined service in Vietnam (because of possible
exposure to toxins) and self-reported use of marijuana and
2R. Arden et al. / Intelligence xxx (2008) xxx–xxx
ARTICLE IN PRESS
Please cite this article as: Arden, R., et al., Intelligence and semen quality are positively correlated, Intelligence (2008),
doi:10.1016/j.intell.2008.11.001

hard drugs (both past and current use for each) because all of
these might undermine semen quality—and more intelligent
men might more effectively avoid them. However, biserial
correlations for these dichotomous variables showed that
none of the five had a significant effect on any aspect of semen
quality, so we did not include them in further analyses.
We considered the possibility that an unmeasured third
variable might influence both intelligence and semen quality.
Variables that influence only intelligence or semen quality
would not explain any association between the two traits,
although we would still like to know about them. We
considered what factors (for which we have data) influence
the intelligence of middle-aged men, to see whether they
might also influence semen quality. The strongest influence
on intelligence differences in adulthood is genetic. The
heritability of intelligence in men of our sample age is around
70% (Plomin & Petrill, 1997). This does not rule out non
genetic influences, but we could not identify any specific
examples of environmental influences on intelligence in
middle-aged men from published reports, other than those
that would have led to the men being excluded from the Army
(such as lead poisoning, iodine deficiency). Quantitative
genetic analyses of intelligence find that the influence of
shared family background of twins and adoptees is zero in
adulthood (Plomin & Petrill, 1997; Scarr & Weinberg, 1978).
The relatively small non genetic contribution to intelligence
differences among middle-aged men may be non systematic,
arising from idiopathic influences, random biological noise
and measurement error (Jensen, 1997).
3. Results
3.1. Descriptive statistics
Table 1 provides basic descriptive statistics for intelligence
(g), the three semen measures (sperm concentration, count,
and motility), and the five key covariates (age, days
abstinence, BMI, alcohol use, and smoking) for the 425 men.
Sperm count, sperm concentration and alcohol use, were
highly skewed, so we log
10
transformed them for the analyses.
We also log
10
transformed BMI because its kurtosis deviated
significantly from normal.
3.2. Correlations
Table 2 shows that intelligence correlates significantly and
positively with all three measures of semen quality: sperm
concentration (r=.15, p= .002), sperm count (r=.19 , pb.001)
and percentage of motile sperm (% motility) (r=.14, p=.005).
The three semen measures also correlate substantially with
each other (r=.56 to .82). The sample size here (425) affords
above 80% power to detect a correlation of r=.14 at the p=.05
level (two tailed), which is the smallest of the three zero-order
correlations between gand semen quality (SISA website).
3.3. Multiple regressions
Table 3 shows the results fromthe multiple regressions. In
Model 1 we regressed each of the three semen quality
measures, in turn, as the outcome variable on all five
continuous covariates that might influence semen quality:
age, day's abstinence, BMI, alcohol, and smoking. In Model 2
we included intelligence (the gfactor) as an additional
predictor. The increase in Rand the R
2
show that adding the
gfactor improves the fit of the model. We also tested several
curve-fit regression models but none offered a better fit than
the linear model. These results show that the positive
correlations between gand semen quality were not mediated
by age or by three major health hazards: smoking cigarettes,
alcohol and adverse BMI. Nor were they mediated by
Table 1
Descriptive statistics (n=425)
Mean SD Skewness Kurtosis
gfactor −0.02 0.98 −0.61 −0.45
Sperm concentration (mil/ml) 102.39 80.13 1.24 1.52
Sperm concentration (mil/ml) (log
10
) 1.86 0.40 −0.59 −0.11
Sperm count (mil/ejac) 263.23 246.17 1.93 4.81
Sperm count (mil/ejac) (log
10
) 2.23 0.46 −0.61 0.16
Sperm motility (% motile) 58.98 23.37 −0.42 −0.76
Abstinence (days) 3.15 1.78 2.42 8.40
Abstinence (days) (log
10
) 0.45 0.20 0.68 0.48
Alcohol (current drinks/mo) 25.85 40.86 2.38 5.80
Alcohol (current drinks/mo) (log
10
) 0.47 1.21 −0.19 −1.65
Smoking (current cig/day) 10.66 14.03 1.01 −0.06
Age at interview 37.62 2.55 0.11 −0.13
Body mass index 25.60 3.64 1.07 3.01
Body mass index (log
10
) 1.40 0.06 0.42 1.05
Table 2
Correlations among g, semen measures and covariates (n=425)
Intelligence Sperm concentration (mil/ml) Sperm count (mil/ejac) Motility % Abstinence Alcohol Smoking Age BMI
gfactor 1.00
Sperm conc. (mil/ml)
+
0.15** 1.00
Sperm count (mil/ejac)
+
0.19** 0.82** 1.00
Motility (%) 0.14** 0.63** 0.56** 1.00
Abstinence (days) 0.01 0.16** 0.24** 0.08 1.00
Alcohol (drinks/mo) 0.08 0.05 0.04 0.07 –0.01 1.00
Smoking (cigs/day) −0.09 −0.06 −0.08 −0.02 –0.01 0.11* 1.00
Age −0.03 0.09 0.06 0.01 0.00 −0.05 −0.13** 1.00
Body mass index −0.03 −0.02 −0.03 −0.05 –0.09 −0.03 −0.10* 0.07 1.00
*pb05 (2-tailed)
**pb.01 (2-tailed)
+indicates log base 10
Notes: Abstinence (days), Alcohol (current drinks per month), BMI (Body Mass Index) were log
10
transformed. Smoking (current cigarettes per day), age at interview.
3R. Arden et al. / Intelligence xxx (2008) xxx–xxx
ARTICLE IN PRESS
Please cite this article as: Arden, R., et al., Intelligence and semen quality are positively correlated, Intelligence (2008),
doi:10.1016/j.intell.2008.11.001

abstinence, because abstinence and gwere uncorrelated and
therefore made independent contributions to variance in
predicting semen quality.
The assumptions of the multiple regression demand that
the errors, or residuals, are normally distributed (not that the
predictors are normally distributed (Field, 2005 p.170). For
each of the semen quality variables, we examined a
scatterplot of the standardised predicted values with the
standardised residuals to check for heteroscedasticity. We
found, in each case, that the residuals were homoscedastic—
having roughly equal variance at each level of the predictors.
We tested for influential cases that could bias the
regression model by examining the range of Mahalanobis
distances within the sample. Given our sample size and the
number of predictors in our model, Mahalanobis values above
25 may indicate distortion; the largest value in our data was
16.57. A second measure of leverage, Cook's distance, was .18
at the maximum, where values over 1 indicate possible
distortion. Together these analyses gave us confidence that
the model does not suffer from serious biases of those kinds.
4. Discussion
The observed correlations between intelligence and
semen quality may under-estimate the true correlations,
given the limited reliability of single semen samples. We
extracted a principal axis factor (unrotated) from the three
semen measures to see whether this improved the reliability
of the measures, but the correlation with gwas very similar
(r= .17). Semen quality varies across ejaculates not just
according to days abstinence, but also by collection method
(Gerris, 1999; Pellestor, Girardet, & Andreo, 1994), and
psychological context (such as cues of partner infidelity
(Baker & Bellis, 1989; Kilgallon & Simmons, 2005; Pound,
Javed, Ruberto, Shaikh, & Del Valle, 2002). Thus, each man's
single ejaculate as measured in this study is an imperfectly
reliable indicator of his average semen quality across actual
copulations. Finally, more functional aspects of semen quality
can be assayed today (such as capacitation, acrosomal
reactions and cervical mucus penetration (Aitken, 2006;
Weber, Dohle, & Romijn, 2005); these might influence fitness
more directly than do our measures.
The correlations between gand semen quality are small
(r=.14 to.19). This is not evidence of their irrelevance; on the
contrary the effect size is congruent with phenotypic
correlations observed for other bodily correlates of intelli-
gence such as height (r=.14, r=.15) (Silventoinen, Posthuma,
van Beijsterveldt, Bartels, & Boomsma, 2006; Sundet, Tambs,
Harris, Magnus, & Torjussen, 2005). Notably, these effect sizes
fit with average effect sizes (r=.18–.19) found across all
studies in evolution and ecology, as reported bya large survey
of meta-analyses (Jennions & Moller, 2003). Very small
correlations with fitness can, as Jennions and Moller put it,
“turn a mouse into an elephant”. This does not mean,
however, that if intelligence or semen quality is positively
correlated with fitness, high intelligence or semen quality
would become fixed within each mating population.
Although it used to be thought that any trait under selection
would ‘go to fixation’, this has been falsified empirically: an
extensive literature on the maintenance of additive genetic
variation among fitness-related traits now exists, including
for life-history traits (such as height, or time to first child)
(Houle, 1992; Merila & Sheldon, 1999).
The results in this paper are consistent with the hypothe-
sized general fitness factor (ffactor) which, if confirmed,
might help explain the many correlations between intelli-
gence and physical health measures now being documented
in cognitive epidemiology (Batty & Deary, 2004; Hart et al.,
2005; Kuh, Richards, Hardy, Butterworth, & Wadsworth,
2004; Osler, 2003). If a fitness factor does exist, and if gis a
component of a more general factor, we will want to know
how this fitness factor arises.
One possibility is that an ffactor emerges from individual
differences in mutation load (Houle, 2000; Miller, 2000). While
everyone carries many mildly harmful mutations (Nachman &
Crowell, 2000), people differ in mutation load—the overall
number of mutations that disrupt fitness-related traits. Assor-
tative mating for overall genetic quality in socially monoga-
mous species such as humans would amplify such variance in
mutation load. If most genes have pleiotropic effects on several
traits, then most mutations will harm several traits in parallel
and create positive genetic correlations among traits, as
manifest in an ffactor. The new field of phenomics (Oti,
Huynen, & Brunner, 2008) reveals correlations among complex
human phenotypic traits, often hinting at underlying genetic
correlations. Such correlations are also found throughout
research on body symmetry (Furlow, Armijo-Prewitt, Gang-
estad, & Thornhill, 1997) and mate preferences (Thornhill &
Table 3
Multiple regressions of semen measures on covariates and g (n=425)
Sperm conc. (mil/ml) (log
10
) Sperm count (mil/ejac) (log
10
) Motile sperm (%)
Model 1 Model 2 Model1 Model 2 Model 1 Model 2
ßpßpßpßpßpßp
Abstinence 0.16 0.00 0.16 0.00 0.24 0.00 0.24 0.00 0.07 0.13 0.07 0.13
Alcohol 0.07 0.17 0.05 0.27 0.05 0.30 0.03 0.48 0.08 0.12 0.07 0.18
Smoking −0.05 0.27 −0.04 0.42 –0.08 0.10 0.06 0.19 –0.03 0.49 –0.02 0.67
Age 0.09 0.08 0.09 0.06 0.05 0.29 0.06 0.23 0.01 0.81 0.02 0.75
BMI −0.01 0.78 −0.01 0.84 –0.02 0.76 –0.01 0.84 –0.04 0.39 –0.04 0.43
gfactor 0.14 0.0 0 0.18 0.0 0 0.13 0.01
R0.20 0.25 0.26 0.32 0.12 0.17
R
2
0.04 0.06 0.07 0.10 0.01 0.03
Adjusted R
2
0.03 0.05 0.06 0.09 0.00 0.02
Notes: Abstinence (days), Alcohol (current drinks permonth), BMI (Body Mass Index) were log
10
transformed. Smoking (current cigarettes per day), age at interview.
4R. Arden et al. / Intelligence xxx (2008) xxx–xxx
ARTICLE IN PRESS
Please cite this article as: Arden, R., et al., Intelligence and semen quality are positively correlated, Intelligence (2008),
doi:10.1016/j.intell.2008.11.001

Gangestad, 1999). The correlations between intelligence and
semen quality add to the plausibility of a phenotype-wide f
factor precisely because there is no obvious morphological or
functional relationship between the two traits.
Mutations acting on pleiotropic genes will result in correla-
tions between the traits affected by those genes. Gene expression
is more similar between brain and testes than between many
other human tissues investigated (Guo, Huang, Studholme, Wu, &
Zhao, 2005). Mutations in several genes on the X chromosome
result in both mental retardation and gonadal abnormalities
(Graves, 2006; Wilda et al., 2000). If a gene is very widely
expressed and its products have a pervasively useful and
important job to do, then massive pleiotropy should be detectable
with the appropriate sample size. Some genes now on the human
X chromosome were earlier on autosomes (Kohn, Kehrer-
Sawatzki, Steinbach, Graves, & Hameister, 2007). Evidence
concerning the recruitment of ancestrally old genes to new
functions (including spermatogenesis and intelligence) will
elucidate the connection between sperm quality and intelligence
(Kohn et al., 2007).
A second possibility is that traits under positive selection may
become linked—by assortative mating occurring over many
generations—resulting in what is called gametic-phase disequili-
brium. In a landmark study of people in 37 countries, intelligence
was found to be valued highly by both men and women as a mate
choice characteristic (Buss, 1989). If this preference has been
stable for an evolutionarily-relevant period (see Hawks, Wang,
Cochran, Harpending, & Moyzis, 2007 for a discussion of selection
rates), then trans-generational, cross-trait assortative mating
preferences may be another source of covariance between
intelligence and sperm quality.
The notion of the ffactor is under-explored empirically, yet it
is theoretically important for understanding human genomics,
phenomics, evolution, and intelligence itself. It is also practically
important for interpreting any observed correlations between
physical health, intelligence and socio-economic outcomes
affected by intelligence such as education, occupation, and
income.Researchonawiderrangeofpsychologicalandphysical
traits will be required to establish or refute the existence of a
genuine fitness matrix in our species.
References
Aitken, R. J. (2006). Sperm function tests and fertility. International Journal of
Andrology,29(1), 69−74.
Baker, R. R., & Bellis, M. A. (1989). Number of sperm in human ejaculates
varies in accordance with sperm competition theory. Animal Behaviour,
37, 867−869.
Batty, G. D., & Deary, I. J. (2004). Early life intelligence and adult health—
Associations, plausible mechanisms, and public health importance are
emerging. British Medical Journal,329(7466), 585−586.
Batty, G. D., Deary, I. J., & Gottfredson, L. S. (2007). Premorbid (early life) IQ
and later mortality risk: Systematic review. Annals of Epidemiology,17(4),
278−288.
Bonde, J.P. E., Ernst,E., Jensen, T.K., Hjollund, N. H.I., Kolstad, H.,Henriksen, T. B.,
et al. (1998). Relation between semen quality and fertility: A population-
based study of 430 first-pregnancy planners. Lancet,352( 913 5), 1172−117 7.
Buss, D. M. (1989). Sex differences in human mate selection: Evolutionary
hypothesis tested in 37 cultures. Behavioral and Brain Sciences,12,1−49.
Carroll, J. B. (1993). Human cognitive abilities. New York: Cambridge
University Press.
Centers for Disease Control (1989). Health status of Vietnam veterans, vol. 3.
Atlanta: US Department of health & Human services.
Centers for Disease Control. (2007). The Health Status of Vietnam Veterans.
from http://www.cdc.gov/nceh/veterans/default1c.htm
Field, A. P. (2005). Discovering statistics using SPSS: (and sex, drugs and
rock'n'roll). London; Thousand Oaks: SAGE.
Furlow, F. B., Armijo-Prewitt, T., Gangestad, S. W., & Thornhill, R. (1997).
Fluctuating asymmetry and psychometric intelligence. Proceedings of the
Royal Society of London. Series B, Biological sciences,264(1383), 823−829.
Gerris, J. (1999). Methods of semen collection not based on masturbation or
surgical sperm retrieval. Human Reproduction Update,5(3), 211−215.
Graves, J. A. M. (2006). Sex chromosome specialization and degeneration in
mammals. Cell,124(5), 901−914.
Guo, J. H., Huang, Q., Studholme, D. J., Wu, C. Q., & Zhao, S. Y. (2005).
Transcriptomic analyses support the similari ty of gene expression
between brain and testis in human as well as mouse. Cytogenetic and
Genome Research,111(2), 107−109.
Guzick, D. S., Overstreet, J. W., Factor-Litvak, P., Brazil, C. K., Nakajima, S. T.,
Coutifaris, C., et al. (2001). Sperm morphology, motility, and concentra-
tion in fertile and infertile men. New England Journal of Medicine,345
(19), 1388−1393.
Hart, C. L., Taylor, M. D., Smith, G. D., Whalley, L. J., Starr, J. M., Hole, D. J., et al.
(2005). Childhood IQ and all-cause mortality before and after age 65:
Prospective observational study linking the Scottish Mental Survey 1932
and the Midspan studi es. British Journal of Health Psychology,10,
153 −165.
Hawks, J., Wang, E. T., Cochran, G. M., Harpending, H. C., & Moyzis, R. K.
(2007). Recent acceleration of human adaptive evolution. Proceedings of
the National Academy of Sciences of the United States of America,104,
20753−20758.
Houle, D. (1992). Comparing evolvability and variability of quantitative traits.
Genetics,130 (1), 195−204.
Houle, D. (2000). Is there a gfactor for fitness? The nature of intelligence, Vol.
233 (pp. 149 −170). Chichester: John Wiley & Sons, Ltd.
Jastak, J. F., & Jastak, S. R. (1965). Wide range achievement test: WRAT. Wilmington,
Delaware: Guidance Associates.
Jennions, M. D., & Moller, A. P.(20 03). A survey of the statistical power of research in
behavioral ecology and animal behavior. Behavioral Ecology,14(3), 438−445.
Jensen, A. R. (1997). The puzzle of nongenetic variance. In R. J. Sternberg& E. L.
Grigorenko (Eds.), Intelligence, heredity, and environment (pp. 42−88).
Cambridge: Cambridge University Press.
Kilgallon, S. J., & Simmons, L. W. (2005). Image content influences men's
semen quality. Biology Letters,1(3), 253−255.
Kohn, M., Kehrer-Sawatzki, H., Steinbach, P., Graves, J. A. M., & Hameister, H.
(2007). Recruitment of old genes to new functions: Evidence obtained by
comparing the orthologues of human XLMR genes in mouse and chicken.
Cytogenetic and Genome Research,116 (3), 173−180.
Kuh, D., Richards, M., Hardy, R., Butterworth, S., & Wadsworth, M. E. J. (2004).
Childhood cognitive ability and deaths up until middle age: A post-war
birth cohort study. International Journal of Epidemiology,33(2), 408−413.
Levitas, E., Lunenfeld, E., Weiss, N., Friger, M., Har-Vardi, I., Koifman, A., et al.
(2005). Relationship between the duration of sexual abstinence and
semen quality: Analysis of 9,489 semen samples. Fertility and Sterility,83
(6), 1680−1686.
Little, W., Fowler, H. W., & Coulson, J. (1984). Theshorter Oxford English dictionary on
historic principles. Oxford: Clarend on.
Meizel, S. (2004). The sperm, a neuron with a tail: ‘Neuronal’receptors in
mammalian sperm. Biological Reviews,79(4), 713−732.
Merila, J., & Sheldon, B. C. (1999). Genetic architecture of fitness and
nonfitness traits: Empirical patterns and development of ideas. Heredity,
83,103−109.
Miller, G. F. (2000). Sexual selection for indicators of intelligence. The nature
of intelligence,Vol. 233. (pp. 260−275)Chichester: John Wiley & Sons Ltd.
Montague,E.K.,Williams,H.L.,Gieseking,C.F.,&Lubin,A.(1957).Armytestsfor
assessment of intellectual deficit. US Armed Forces Medical Journal,8, 883−892.
Nachman, M. W., & Crowell, S. L. (2000). Estimate of the mutation rate per
nucleotide in humans. Genetics,156 (1), 297−304.
Nguyen,R.H.N.,Wilcox,A.J.,Skjaerven,R.,&Baird,D.D.(2007).Men'sbodymass
index and infertility. Human Reproduction,22,2488−2493.
Osler, M. (2003). Socioeconomic position in early life, birth weight, childhood
cognitive function, and adult mortality. A longitudinalstudy of Danish men
born in 1953. Journal of Epidemiology and Community Health,57(12), 995.
Oti, M., Huynen, M. A., & Brunner, H. G. (2008). Phenome connections. Trends in
Genetics,24(3), 103−106.
Pellestor, F.,Girardet, A., & Andreo, B. (1994). Effect of long abstinence periods
on human sperm quality. International Journal of Fertility and Menopausal
Studies,39(5), 278−282.
Plomin, R., & Petrill, S. A. (1997). Genetics and intelligence: What's new?
Intelligence,24,53−77.
Pound, N., Javed, M. H., Ruberto, C., Shaikh, M. A., & Del Valle, A. P. (2002).
Duration of sexual arousal predicts semen parameters for masturbatory
ejaculates. Physiology & Behavior,76(4–5), 685−689.
Scarr, S., & Weinberg, R. A. (1978). The influence of “family background”on
intellectual attainment. American Sociological Review,43,674−692.
Shackelford, T. K., & Pound, N. (2006). Sperm competition in humans: Classic
and contemporary readings. New York: Springer.
5R. Arden et al. / Intelligence xxx (2008) xxx–xxx
ARTICLE IN PRESS
Please cite this article as: Arden, R., et al., Intelligence and semen quality are positively correlated, Intelligence (2008),
doi:10.1016/j.intell.2008.11.001

Silventoinen, K., Posthuma, D., van Beijsterveldt, T., Bartels, M., & Boomsma,
D. I. (2006). Genetic contributions to the association between height and
intelligence: Evidence from Dutch twin data from childhood to middle
age. Genes, Brain and Behavior,5(8), 585−595.
SISA. Simple Interactive Statistical Analysis. from http://www.quantitatives-
kills.com/sisa/
Soares, S. R., & Melo, M. A. (2008). Cigarette smoking and reproductive
function. Current Opinion in Obstetrics and Gynecology,20(3), 281−291.
Sundet, J. M., Tambs, K., Harris, J. R., Magnus, P., & Torjussen, T. M. (2005).
Resolving the genetic and environmental sources of the correlation
between height and intelligence: A study of nearly 2600 Norwegian male
twin pairs. Twin Research and Human Genetics,8(4), 307−311.
Thornhill, R., & Gangestad, S. W. (1999). Facial attractiveness. Trends in
Cognitive Sciences,3(12), 452−460.
van der Merwe, F. H., Kruger,T. F., Oehninger, S. C., & Lombard, C. J. (2005). The
use of semen parameters to identify the subfertile male in the general
population. Gynecologic and Obstetric Investigation,59(2), 86−91.
Weber, R. F. A., Dohle, G. R., & Romijn, J. C. (2005). Clinical laboratory
evaluation of male subfertility. Advances in Clinical Chemistry,40,
317−364.
Wechsler, D. (1981). WAIS-R: Manual: Wechsler adult intelligence scale—
revised. New York, NY: Harcourt Brace Jovanovich Psychological Corp.
Wilda, M., Bachner, D., Zechner, U., Kehrer-Sawatzki, H., Vogel, W., &
Hameister, H. (2000). Do the constraints of human speciation cause
expression of the same set of genes in brain, testis, and placenta? Cyto-
genetics and Cell Genetics,91(1–4), 300−302.
World Health Organization (1999). WHO laboratory manual for the examina-
tion of human semen and sperm-cervical mucus interaction. New York, NY.
6R. Arden et al. / Intelligence xxx (2008) xxx–xxx
ARTICLE IN PRESS
Please cite this article as: Arden, R., et al., Intelligence and semen quality are positively correlated, Intelligence (2008),
doi:10.1016/j.intell.2008.11.001