Human Pigmentation Variation: Evolution, Genetic Basis,
and Implications for Public Health
Esteban J. Parra*
Department of Anthropology, University of Toronto at Mississauga, Mississauga, ON, Canada L5L 1C6
KEY WORDS pigmentation; evolutionary factors; genes; public health
ABSTRACT Pigmentation, which is primarily determined
by the amount, the type, and the distribution of
melanin, shows a remarkable diversity in human populations,
and in this sense, it is an atypical trait. Numerous
genetic studies have indicated that the average proportion
of genetic variation due to differences among
major continental groups is just 10–15% of the total
genetic variation. In contrast, skin pigmentation shows
large differences among continental populations. The
reasons for this discrepancy can be traced back primarily
to the strong inﬂuence of natural selection, which has
shaped the distribution of pigmentation according to a
latitudinal gradient. Research during the last 5 years
has substantially increased our understanding of the
genes involved in normal pigmentation variation in
human populations. At least six genes have been identiﬁed
using genotype/phenotype association studies and/or
direct functional assays, and there is evidence indicating
that several additional genes may be playing a role in
skin, hair, and iris pigmentation. The information that is
emerging from recent studies points to a complex picture
where positive selection has been acting at different
genomic locations, and for some genes only in certain
population groups. There are several reasons why elucidating
the genetics and evolutionary history of pigmentation
is important. 1) Pigmentation is a trait that
should be used as an example of how misleading simplistic
interpretations of human variation can be. It is erroneous
to extrapolate the patterns of variation observed
in superﬁcial traits such as pigmentation to the rest of
the genome. It is similarly misleading to suggest, based
on the ‘‘average’’ genomic picture, that variation among
human populations is irrelevant. The study of the genes
underlying human pigmentation diversity brings to the
forefront the mosaic nature of human genetic variation:
our genome is composed of a myriad of segments with
different patterns of variation and evolutionary histories.
2) Pigmentation can be very useful to understand the
genetic architecture of complex traits. The pigmentation
of unexposed areas of the skin (constitutive pigmentation)
is relatively unaffected by environmental inﬂuences
during an individual’s lifetime when compared with
other complex traits such as diabetes or blood pressure,
and this provides a unique opportunity to study gene–
gene interactions without the effect of environmental
confounders. 3) Pigmentation is of relevance from a public
health perspective, because of its critical role in photoprotection
and vitamin D synthesis. Fair-skinned individuals
are at higher risk of several types of skin cancer,
particularly in regions with high UVR incidence, and
dark-skinned individuals living in high latitude regions
are at higher risk for diseases caused by deﬁcient or
insufﬁcient vitamin D levels. Yrbk Phys Anthropol
50:85–105, 2007. VVC 2007 Wiley-Liss, Inc.
INTRODUCTION
Pigmentation is one of the most variable phenotypes
in humans. The color of skin, hair, and eyes is primarily
determined by melanin, a generic term used to describe
a complex group of biopolymers synthesized by specialized
cells known as melanocytes. The available evidence
strongly suggests that natural selection is responsible
for the observed variation in this trait, but the ultimate
evolutionary factors have not been fully elucidated.
There have been impressive advances in our understanding
of the pigmentary system, mainly driven by studies
in animal models, and important breakthroughs have
also been made through studying pigmentation disorders
in humans. In contrast, we are just beginning to understand
the genetic basis of normal pigmentation variation
in our species. In this article, I review the major evolutionary
hypotheses that have been put forward to
explain the distribution of pigmentation and the current
state of our knowledge about the genes involved in the
variation of skin, hair, and eye pigmentation within and
among human populations. Finally, I devote the last section
of this review to the important public health implications
derived from the unique evolutionary history and
geographic distribution of skin pigmentation.
A REVIEW OF PIGMENTATION BIOLOGY
Melanin is the main pigment of our skin, hair, and
eyes, and other chromophores, such as hemoglobin, play
a minor role in skin pigmentation. Melanin is not a single
compound. Rather, it is a mixture of biopolymers synthesized
in melanocytes located in the basal layer of the
epidermis, the hair bulb, and the iris. Within the melanocytes,
melanin production takes place in the melanosomes,
lysosome-like organelles where melanin granules
are synthesized using the aminoacid tyrosine as the
major substrate. Tyrosinase (TYR) is the key enzyme in
melanogenesis and mediates the ﬁrst steps in melanin
synthesis, which involve the hydroxylation of tyrosine to
dopa and the subsequent oxidation to dopaquinone. Dopaquinone
is a key intermediate compound that undergoes
*Correspondence to: Esteban J. Parra, Department of Anthropology,
University of Toronto at Mississauga, 3559 Mississauga Road
North, Room 212 North Building, Mississauga, ON, Canada L5L
1C6. E-mail: esteban.parra@utoronto.ca
DOI 10.1002/ajpa.20727
Published online in Wiley InterScience
(www.interscience.wiley.com).
VVC 2007 WILEY-LISS, INC.
YEARBOOK OF PHYSICAL ANTHROPOLOGY 50:85–105 (2007)
further modiﬁcation in two alternative pathways. In the
absence of the amino acid cysteine, dopaquinone eventually
gives rise to the dark brown/black, insoluble DHI
(5,6-dihydroxyindole)-eumelanin and to the lighter, alkali-soluble,
DHICA (5,6-dihydroxyindole-2-carboxylic
acid)-eumelanin, while in the presence of cysteine, dopaquinone
gives rise to the alkali-soluble red-yellow pheomelanin.
Additional details about melanogenesis are
available in Bolognia and Orlow (2003) and Meredith
and Sarna (2006). The main steps involved in melanogenesis
are depicted in Figure 1. It is important to note
that eumelanin and pheomelanin seem to have different
properties with respect to phototoxic and oxidizing
potential, and this may explain the higher cancer risk
observed in individuals with high pheomelanin contents
(Hill and Hill, 2000; Takeuchi et al., 2004; Samokhvalov
et al., 2005; Ye et al., 2006).
The regulation of melanogenesis and the distribution
of melanin differ in the skin, hair, and iris. In the skin,
the melanocytes located in the basal layer of the epidermis
transfer melanosomes to adjacent keratinocytes
through dentritic structures, and the keratinocytes eventually
migrate to the upper layers of the epidermis
(Rees, 2003). Within the keratinoctyes, melanosomes are
typically aggregated over the nucleus, providing protection
against ultraviolet radiation (UVR). In the hair, the
hair bulb is the only site of melanin production. Here,
highly melanogenic melanocytes transfer melanosomes
to surrounding immature precortical keratinocytes that
will eventually differentiate and migrate to form the pigmented
hair shaft (Slominski et al., 2005). The epidermal-melanin
unit (skin) and the follicular-melanin unit
(hair) differ in some important ways. In the hair, melanogenesis
only takes place during the anagen phase of
the hair growth cycle (the period of hair shaft formation
that lasts on average 3–5 years), while in the epidermis,
it appears to be continuous (Ortonne and Prota, 1993).
There seem to be additional differences related to the
morphology of the melanocytes and the microenvironment
in which melanogenesis takes place (Tobin and
Bystryn, 1996; Tobin and Paus, 2001; Slominski et al.,
2005). This may explain the discordant pigmentation
patterns observed in some individuals (e.g., dark-pigmented
hair and pale blue eyes and pale skin). In contrast
to hair and skin, in the iris the melanosomes are
found only within the melanocytes and the type of melanin
and the density and distribution of melanosomes are
the major determinants of eye color (Sturm and Frudakis,
2004). The iris contains two tissue layers. The thin
inner layer is called the iris pigment epithelium (IPE)
and does not contribute to normal iris color variation
because the melanin in this heavily pigmented layer has
a similar distribution in individuals with different iris
colors (albinos are an exception). The outermost layer is
Fig. 1. Major steps in melanin synthesis. Reprinted from Sturm and Frudakis, 2004. Eye colour: portals into pigmentation
genes and ancestry. Trends in Genetics. 20:327–332, with permission from Elsevier. [Color ﬁgure can be viewed in the online issue,
which is available at www.interscience.wiley.com.]
86 E.J. PARRA
Yearbook of Physical Anthropology—DOI 10.1002/ajpa
known as the iris stroma, and the melanin content in
this layer is the primary cause of normal iris color variation
(Sturm and Frudakis, 2004; Wielgus and Sarna,
2005).
Although there are differences in melanocyte density
depending on body site (Whiteman et al., 1999), variation
in the number of melanocytes does not seem to be a
major factor explaining the pigmentation differences
among human populations. Rather, differences in pigmentation
are due to two major factors: the amount and
type of melanin synthesized in the melanocytes (e.g., ratio
of eumelanin to pheomelanin), and the shape and distribution
of the melanosomes. Lightly pigmented skin is
enriched in light-brown DHICA-eumelanin and yellow/
red pheomelanins, and the melanosomes tend to be less
pigmented, smaller in size, and packaged in groups.
Darkly pigmented skin has more melanin, which is
enriched in DHI-eumelanin (Alaluf et al., 2002). Additionally,
the melanosomes of dark-skinned individuals
are more pigmented, larger, and distributed as single
units (Szabo et al., 1969; Alaluf et al., 2002). Figure 2
summarizes the population differences observed in the
pattern of melanosome size and distribution in the skin.
It has also been described that individuals with black
hair have the highest eumelanin-to-pheomelanin ratio,
those with blond or light brown hair show intermediate
ratios, and those with red hair have the highest pheomelanin
levels (Ortonne and Prota, 1993; Rees, 2003).
With respect to eye color, brown irises have large
amounts of melanin and high numbers of melanosomes,
and absorb a substantial proportion of the incoming
light, particularly at short wavelengths. In contrast, blue
irises have low melanin content and few melanosomes.
As a result of this, long-wavelength light penetrates the
stroma and is absorbed in the IPE, while short-wavelength
light (blue) is scattered by the collagen matrix of
the iris. Green and hazel irises have intermediate
amounts of melanin (Sturm and Frudakis, 2004).
MEASURING PIGMENTATION
The ﬁrst attempts to use standardized methods to
measure skin pigmentation were based on color-matching
techniques. The most widely used was Von Luschan’s
chromatic scale, where a subject’s skin pigmentation was
compared with 36 small ceramic tiles ranging from white
to black (Robins, 1991). However, these early methods,
which were popular in the ﬁrst half of the 20th century,
had a high degree of subjectivity and were largely abandoned
in the 1950s, when the ﬁrst portable reﬂectance
spectrometers became available for ﬁeld work. Reﬂectance
spectrometers measure the percentage of light
reﬂected from the skin at different wavelengths and represented
an important technological improvement in
terms of objectivity and accuracy. The most widely used
instrument was the EEL reﬂectance spectrophotometer
(Evans Electroselenium Ltd., currently distributed by
Diffusion Systems, UK), which measured the reﬂectance
of the skin using nine ﬁlters corresponding to different
wavelengths of the visible spectrum, from violet (601 ﬁlter,
425 nm) to deep red (609 ﬁlter, 685 nm). The readings
of the instrument represented the reﬂectance of the
skin at each wavelength relative to that of a white
standard (typically white magnesium carbonate). Dozens
of populations from around the world were sampled
using the EEL reﬂectance spectrophotometer (although
not always using all the available ﬁlters) from the time
of its ﬁrst use by Weiner (Weiner, 1951) for measuring
pigmentation in 1951. More information about studies
using the EEL reﬂectance spectrophotometer can be
found in Byard (1981), Robins (1991), and Jablonski and
Chaplin (2000). Other portable spectrometers were available
to the scientiﬁc community, such as the Photovolt
instrument. The Photovolt used principles similar to
those of the EEL spectrophotometer, but the two instruments
differed in the number and characteristics of the
ﬁlters used to sample the visible spectrum: The EEL
spectrophotometer had nine narrow wavelength ﬁlters
while the Photovolt employed six ﬁlters with broader
band transmittance. This made direct comparison of the
results obtained with both instruments difﬁcult,
although conversion formulae were developed to overcome
this issue (Lees and Byard, 1978; Lees et al.,
1979). Currently, there are three strategies based on reﬂectance
technologies to study the main chromophores of
the skin (melanin and hemoglobin): tristimulus colorimetry,
specialized narrow-band reﬂectometry, and diffuse
reﬂectance spectroscopy. The modern instruments used
to measure skin pigmentation offer signiﬁcant advantages
over previously available instruments in terms of
portability and accuracy. In addition to these indirect
approaches to measure pigmentation based on reﬂectometry,
it is also possible to directly measure melanin content
using biochemical methods. I brieﬂy describe the
available strategies to measure pigmentation below.
Tristimulus colorimetry
Tristimulus colorimetry was developed as a means of
objectively representing color in a manner analogous to
the way the eye perceives color (Hunter, 1942). The reﬂectance
level of light through three particular broad
wavelength ﬁlters (photodiode arrays on newer instruments)
is determined. Color parameters are then deﬁned
by the levels of, and differences among, the reﬂectance
levels of these three ﬁlters. The most commonly used
color parameters are the Commission International
d’Eclairage (CIE) L*a*b* system established in 1976. In
Fig. 2. Population differences in the pattern of melanosome
size and distribution in the skin. Reproduced from Barsh, 2003.
What controls variation in human skin color? PLoS Biology
1:19–22, with permission from PLoS Biology. [Color ﬁgure
can be viewed in the online issue, which is available at www.
interscience.wiley.com.]
87HUMAN PIGMENTATION VARIATION
Yearbook of Physical Anthropology—DOI 10.1002/ajpa
the CIELab color system, any color can be represented
by three variables: L*, the lightness-darkness axis; a*,
the red-green axis; and b*, the blue-yellow axis, which
can be plotted in three-dimensional space (Weatherall
and Coombs, 1992). Tristimulus colorimeters like the
Minolta Chroma Meter series (Konica Minolta Sensing,
Osaka, Japan) are usually able to report color values for
a number of other color systems as well. In studies using
tristimulus colorimeters, erythema (redness of the skin
primarily due to hemoglobin) is typically evaluated using
the a* parameter, and pigmentation is evaluated using
L*, b*, or a combination of the three parameters (Stamatas
et al., 2004).
Specialized narrow-band reﬂectometry
This strategy is based on the work of Diffey et al.
(1984), who developed a method to speciﬁcally measure
hemoglobin and melanin in the skin. Hemoglobin is present
in the blood vessels of the dermis in two major
forms, deoxy-hemoglobin (hemoglobin with the oxygenbinding
sites unoccupied by oxygen) and oxy-hemoglobin
(oxygen-binding sites occupied by oxygen molecules). Hemoglobin
primarily absorbs in the lower wavelengths of
visible light, with typical absorption peaks in the greenyellow
region of the spectrum (deoxy-hemoglobin absorption
maxima, 555 nm; oxy-hemoglobin absorption maxima,
540 and 577 nm), and absorbs very little in the red
wavelengths, which is why blood is red (Kollias and Stamatas,
2002). In contrast, melanin shows absorbance of
light of all wavelengths. Melanin’s absorption decreases
exponentially in the range 400–600 and then linearly in
the range of 720–620 nm (Kollias et al., 1991). Based on
these differences in the spectral curves of hemoglobin
and melanin, Diffey et al. (1984) suggested that the reﬂectance
of narrow-band light in the red spectrum would
yield reasonable estimates of the melanin content of the
skin, following the equation:
M ðmelanin indexÞ ¼ log10
1
% red reflectance
 
The degree of skin redness or erythema can be calculated
by subtracting the absorbance due to melanin from
the absorbance of the green ﬁlter and is calculated as:
E ðerythema indexÞ ¼ log10
1
% green reflectance
 
À log10
1
% red reflectance
 
A number of narrow-band spectrometers that provide
measurements of the melanin and erythema indices are
commercially available, including the DermaSpectrometer
(Cortex Technology, Hadsund, Denmark), the Erythema/Melanin
Meter (DiaStron, DiaStron, Hampshire,
UK), and the Mexameter (Courage-Khazaka Electronic,
Ko¨ln, Germany). Shriver and Parra (2000) carried out a
comparative study of hair and skin pigmentation in individuals
of diverse ancestry using a tristimulus colorimeter
(Photovolt ColorWalk) and a narrow-band spectrometer
(DermaSpectrometer). They showed that there is a
high correlation between L* and M, the parameters typically
used to measure melanin content, and concluded
that, while both instruments provide accurate estimates
of melanin levels in skin and hair, measurements using
narrow-band instruments may be less affected by the
greater redness of certain body sites due to increased
vascularization.
Diffuse reﬂectance spectroscopy
Diffuse reﬂectance spectroscopy (DRS) is considered
the most versatile and accurate noninvasive strategy to
measure skin pigmentation (Stamatas et al., 2004). In
contrast to the two methods described before, which are
based on reﬂectance values in a subset of the visible
spectrum, DRS reﬂectance values are measured at high
optical resolution throughout the full visible spectrum.
Currently, several companies manufacture portable spectrometers
at reasonable prices, including OceanOptics
(Boca Raton, FL), Newport (Irvine, CA), and Stellarnet
(Tampa, FL). The illumination light is delivered to the
skin at the site of interest using a bifurcated ﬁber optic
bundle or an integrating sphere. An analyzer-detector
system (typically a diffraction grating coupled to a detector
array) allows simultaneous acquisition of the whole
spectrum. Figure 3 shows an example of skin reﬂectance
curves from 12 individuals of diverse ancestry measured
on the inner upper arm at 0.5 nm intervals along the
visible spectrum (400–700 nm) using a Stellarnet
EPP2000 reﬂectance spectrometer (E.J. Parra, unpublished
data). The availability of data from the full visible
spectrum means that it is possible to separate the effects
of the two major chromophores of the skin, melanin and
hemoglobin, because of their different spectral properties.
Kollias and Baqer (1985, 1987) have shown that the
amount of melanin in the skin can be estimated by DRS
from the slope of the absorbance spectrum in the region
between 620 and 720 nm. The amount of hemoglobin can
be determined by considering the absorbance of the skin
at 560–580 nm, where absorption by hemoglobin is more
prominent, corrected by the melanin absorption spectrum
(Stamatas et al., 2004). An additional advantage of
the new generation of portable spectrometers is that
they can be used to calculate all types of pigmentation
measures, such as CIELab and other standard color systems,
and the reﬂectance levels at narrow-bands (such
as those used by the DermaSpectrometer) can be determined
using appropriate formulae.
Biochemical measurements of melanin content
As mentioned above, pigmentation can be studied
using noninvasive techniques based on reﬂectometry.
However, using these techniques it is not possible to
directly quantify the three major types of melanin: yellow-red
pheomelanin, light brown DHICA-enriched
eumelanin, and dark brown-black DHI-enriched eumelanin.
The most common strategy for directly determining
melanin composition entails the measurement of chemically
degraded melanin by-products using high-performance
liquid chromatography (HPLC). Pheomelanin and
DHICA-enriched eumelanin contents can be determined
using HPLC, and the amount of DHI-enriched eumelanin
can be estimated as the difference between the total
melanin content, measured using spectophotometry, and
the amounts of pheomelanin and DHICA-enriched melanin,
measured using HPLC (Alaluf et al., 2001). HPLCbased
methods have been used to estimate eumelanin
and pheomelanin contents in human hair and skin (Ito
88 E.J. PARRA
Yearbook of Physical Anthropology—DOI 10.1002/ajpa
and Fujita, 1985; Thody et al., 1991; Ozeki et al., 1996;
Alaluf et al., 2001). The major disadvantage of this
approach for determining pigment levels in the skin is
that it is an invasive method, requiring suction blisters
or punch biopsies. There have been comparative studies
showing that there is a high correlation between epidermal
melanin content estimated using biochemical and
reﬂectance methods (Alaluf et al., 2002, Kongshoj et al.,
2006).
Although methods based on reﬂectance technologies
have gained widespread acceptance during the last decade,
alternative strategies are still widely used in some
ﬁelds. For example, the skin classiﬁcation system developed
in 1975 by Thomas B. Fitzpatrick and coworkers is
still widely used today in the ﬁeld of dermatology (Fitzpatrick,
1988). This system involves the classiﬁcation of
a person into one of six sun-reactive skin types (I,
always burns/never tans; II, usually burns/tans less than
average; III, sometimes mildly burns/tans average; IV,
rarely burns/tans more than average; V, never burns/
brown skin; VI, never burns/black skin) (Fitzpatrick,
1988). This scale has been criticized for various reasons,
including its poor predictive value, poor correlation with
skin color, and poor correlation with UV sensitivity
(Westerhof et al., 1990; Leenutaphong, 1995; Snellman
et al., 1995; Damian et al., 1997).
I would like to ﬁnish this section devoted to the techniques
available for the measurement of pigmentation
by discussing strategies used to determine iris color. The
conventional methods based on reﬂectometry or biochemical
analysis cannot be applied to evaluate iris pigmentation
in anthropological or genetic studies. Traditionally,
studies of iris color have been based on classiﬁcation of
iris pigmentation in broad categories (e.g., blue, gray,
green, hazel, light brown, dark brown) by trained observers
(Eiberg and Mohr, 1996; Rebbeck et al., 2002; Zhu
et al., 2004; Duffy et al., 2007). However, this qualitative
approach fails to capture the quantitative nature of iris
pigmentation variation. Recently, there have been
attempts to develop quantitative methods to estimate
iris color based on photographs taken under standardized
conditions. German et al. (1998) and Niggemann
et al. (2003) reported iris pigmentation in the RedGreen-Blue
(RGB) color space, and Melgosa et al. (2000)
quantiﬁed iris color using the CIELab color system.
More recently, Fan et al. (2003) have developed a method
that automatically extracts the iris region from photographs,
computes the iris color, and corrects the color
based on a standard calibration target. These and other
studies are paving the way for much more objective and
precise measurement of eye color, which will allow quantitative
evaluation of even subtle pigmentation variation
within the iris (e.g., darker pigmentation in the peripupillary
area).
DISTRIBUTION OF PIGMENTATION IN
HUMAN POPULATIONS
The distribution of skin pigmentation differs from that
of other phenotypic traits and most genetic markers.
Relethford (2002) estimated that 88% of the total skin
pigmentation variation can be explained by pigmentation
differences among major geographic groups. This value
stands in sharp contrast to what has been described in
numerous autosomal genetic studies, which indicate that
for an ‘‘average’’ genetic marker in humans, variation
among major geographic groups typically accounts for
only 10–15% of the total diversity (Lewontin 1972; Jorde
and Wooding, 2004; Tishkoff and Kidd, 2004). Other
quantitative traits, such as craniometric traits, show a
pattern of diversity broadly consistent with the genetic
data; only 13% of the total variation is due to differences
among major geographic regions (Relethford, 2002).
Clearly, skin pigmentation shows an atypical distribution,
especially considering the recent origin of anatomically
modern humans ($200,000 years ago, Foley, 1998).
Figure 4 shows the global distribution of skin pigmentation,
based on the map of the Italian geographer Renato
Biasutti. Biasutti (1953) did not use reﬂectometry to
measure skin pigmentation (his research took place in
the ﬁrst half of the 20th century) and his sampling of
Fig. 3. Skin reﬂectance
curves for individuals of European,
East Asian, South Asian,
and African ancestry. Skin reﬂectance
was measured on the
inner upper arm at 0.5-nm
intervals along the visible spectrum
(400–700 nm) using a
Stellarnet EPP2000 reﬂectance
spectrometer. Note the decrease
in skin reﬂectance in the greenyellow
wavelengths (540–580
nm) due to hemoglobin absorption,
which is particularly evident
in the curves of the individuals
with the highest skin
reﬂectance.
89HUMAN PIGMENTATION VARIATION
Yearbook of Physical Anthropology—DOI 10.1002/ajpa
human populations was not comprehensive. However,
Biasutti’s map is still useful to describe the general geographic
patterns of skin pigmentation in human populations,
which show a strong correlation with latitude.
Skin color tends to be darker in equatorial and tropical
areas (sub-Saharan Africa, South Asia, Australia, and
Melanesia) than in areas located far from the equator
(see Fig. 4). The correlation of skin pigmentation and
latitude was also demonstrated in a study carried out by
Relethford (1997). Importantly, Relethford used more
precise pigmentation estimates based on reﬂectometry,
and he found a strong relationship of skin reﬂectance
with latitude (see Fig. 5). The underlying factor explaining
this remarkable relationship between skin pigmentation
and latitude seems to be UVR intensity, which is
greater at the equator and progressively diminishes with
increasing latitude. More recently, Jablonski and Chaplin
(2000) published a comprehensive study, which evaluated
the relationship of skin reﬂectance measurements
in populations throughout the world with UVR levels
estimated using remote sensing technology. These
authors observed a strong correlation between skin reﬂectance
and UVR levels and found an excellent agreement
of the skin reﬂectance values predicted by a
regression model using average UVR as an independent
variable with the observed reﬂectance values. In summary,
the available body of data clearly indicates that
the distribution of skin pigmentation in humans has
been strongly inﬂuenced by UVR levels. I will discuss
this issue in more detail in the next section, which is
devoted to the evolution of human skin pigmentation.
In contrast to skin pigmentation, which shows a
strong correlation with latitude, variation in hair and
iris color is much more restricted geographically. Most
human populations have dark hair and irises. Red and
blond hair are mainly found in European populations
(the highest frequencies of red hair occur in Great Britain
and Ireland and the highest frequencies of blond
hair in Nordic countries), although blondism is a trait
commonly found in some Australian and Melanesian
populations (Birdsell, 1993). Similarly, the lighter iris
colors (blue, green, and hazel) are primarily found in
European populations, although they are also present in
Middle Eastern, North African, West Asian, and South
Asian populations (Sturm and Frudakis, 2004).
HYPOTHESES REGARDING THE EVOLUTION OF
SKIN PIGMENTATION IN HUMANS
Numerous hypotheses have been put forward to
explain the evolution of skin pigmentation in human
populations, taking into account the observed relationship
between melanin levels and latitude. In this section,
I brieﬂy discuss the major hypotheses. Additional information
can be found in Robins (1991) and Jablonski
(2004). According to most authors, the underlying factor
explaining the unique geographic distribution of skin
pigmentation in humans appears to be UVR exposure
(Relethford, 1997, 2002; Jablonski and Chaplin, 2000
and references therein). Melanin acts as a key photoprotective
layer in the skin. Of particular importance is
melanin’s role in ﬁltering the harmful UVR from the sun
(280–400 nm). Additionally, melanin (in particular eumelanin)
also exerts a protective effect by scavenging reactive
free radicals and other oxidants (Bustamante et
al., 1993; Krol and Liebler, 1998). There are several
Fig. 4. World map showing
skin pigmentation distribution.
The map is based on the work
of the Italian geographer R.
Biasutti. Higher numbers represent
darker skin color.
Source: D. O’Neil (Behavioral
Sciences Department, Palomar
College, San Marcos, CA; http://
anthro.palomar.edu/vary/vary_1.
htm). [Color ﬁgure can be
viewed in the online issue,
which is available at www.
interscience.wiley.com.]
Fig. 5. Relationship of skin reﬂectance with latitude.
Source: John Relethford (The human species, 6th ed., McGrawHill,
2005, p. 182), reproduced with permission of the McGrawHill
Companies. [Color ﬁgure can be viewed in the online issue,
which is available at www.interscience.wiley.com.]
90 E.J. PARRA
Yearbook of Physical Anthropology—DOI 10.1002/ajpa
potential selective factors that could have driven the evolution
of highly melanized skin in equatorial and tropical
regions with high incidence of UVR.
Melanin and protection from sunburn
and skin cancer
Melanin acts as a natural sunblock and is especially
effective in protecting against the harmful effects of the
shorter wavelengths of the electromagnetic radiation
($300 nm), which are the most damaging to DNA and
proteins (Rees, 2003). The initial effect of skin exposure
to UVR is sunburn, which is characterized by erythema
(redness), edema, and possibly pain and blistering (Rees,
2004). Severe sunburn can cause damage to the sweat
glands, which can lead to a disruption in thermoregulation,
and can also result in an increased risk of infection
in damaged skin cells (Jablonski and Chaplin, 2000;
Rees, 2004). Therefore, in the tropical environments in
which our species ﬁrst evolved, dark skin with high
amounts of UVR absorbing eumelanin would have been
highly advantageous, while lighter skin that was prone
to sunburn, damaged sweat glands, and infections would
have been selected against. It is also well known that
extended exposure to the sun can lead to skin cancer
when UVR damages genes that normally inhibit cancerous
growth (Halliday, 2005; Ullrich, 2005). There is evidence
indicating that there are major differences in skin
cancer risk depending on skin type, with dark skin being
much less susceptible to several types of skin cancer
(Rees, 2004). For this reason, some authors have indicated
that natural selection favored darkly pigmented
skin in tropical and equatorial regions to protect against
skin cancer (Robins, 1991 and references therein). However,
others argue that since some types of skin cancer
(basal and squamous cell carcinomas) are rarely fatal
and since most individuals do not develop cancer
(including malignant melanomas, the most serious type
of skin cancer) until they are past their reproductive
years, it is unlikely that skin cancer had a signiﬁcant
role in the evolution of skin color (Jablonski and Chaplin,
2000). In this respect, it is important to mention
that albinos living in areas with high incidence of UVR
(e.g., sub-Saharan Africa) typically develop premalignant
lesions or skin cancer as teenagers or in early adulthood,
and in Nigeria and Tanzania, less than 10% of albinos
survived beyond the age of 30 (Robins, 1991; Yakubu
and Mabogunje, 1993; Rees, 2003).
Melanin and protection against nutrient
(folate) photolysis
Sunlight is not only damaging to human skin, but also
to some essential nutrients, particularly folate. Folate is
necessary for DNA synthesis and repair, and folate deﬁciency
can result in complications during pregnancy and
multiple fetal abnormalities, including neural tube
defects, such as spina biﬁda and anencephalus (Lucock,
2000; Off et al., 2005). Folate deﬁciency was an important
cause of perinatal and postnatal mortality in some
populations before the introduction of preventive supplementation
(Jablonski and Chaplin, 2000). Folate also
plays a key role in spermatogenesis (Mathur et al., 1977;
Cosentino et al., 1990; Wong et al., 2002; Ebisch et al.,
2006). Central to this discussion is the fact that folate is
extremely sensitive to UVR (Off et al., 2005). Branda
and Eaton (1978) found that folate concentrations in
human plasma decreased signiﬁcantly after brief exposure
to UVR and they also observed that light-skinned
patients undergoing therapeutic UVR exposure had
lower folate levels than controls. This suggests that in
geographic areas with high levels of UVR, light-skinned
individuals will be more prone to folate deﬁciency than
those with dark skin. Due to the important role of folate
in several key biological processes, it is likely that maintenance
of optimal folate levels has been under the inﬂuence
of natural selection.
It is important to mention that the two hypotheses
described above are compatible and either singly or in
combination can explain the eumelanin-rich, dark pigmentation
observed in regions with high UVR incidence.
The genetic data available for some pigmentation candidate
genes (e.g., MC1R) are consistent with natural
selection favoring dark pigmentation in tropical areas,
and will be discussed in more detail in the next section.
However, these hypotheses are insufﬁcient to explain the
global distribution of skin pigmentation (Figs. 4 and 5).
In order to explain the strong correlation of skin pigmentation
with latitude (or UVR), it is necessary to clarify
the evolutionary factors responsible for the lightening of
the skin in regions with low UVR. I summarize the two
main hypotheses below.
Melanin and vitamin D synthesis
Although the effects of UVR on the skin are for the
most part harmful, there is one important exception:
UVB radiation is essential for the synthesis of vitamin D
in the skin. While some dietary sources have substantial
amounts of vitamin D (particularly fatty ﬁsh and ﬁsh oil,
egg yolk, and organ meats), cutaneous synthesis of vitamin
D via exposure to the sun is the main source of vitamin
D (Holick, 2003, 2005). Vitamin D plays a key role
in bone metabolism and vitamin D deﬁciency results in
rickets in children and osteomalacia in adults. In recent
years, other functions of vitamin D have been recognized,
including immunoregulation and regulation of cell
differentiation and proliferation (Holick, 2004).
According to the original version of the vitamin D hypothesis
advanced by Loomis (1967), the correlation
between skin pigmentation and latitude is the result of
selection favoring dark skin near the equator in order to
prevent an excess of cutaneous synthesis of vitamin D,
which could be potentially toxic, and selection favoring
light skin far from the equator in order to maximize
vitamin D synthesis in regions with low UVR incidence.
Loomis’ views on vitamin D toxicity have since then
been disproved, as it has been demonstrated that sun exposure
does not result in vitamin D toxicity (Holick,
2007). The current versions of the vitamin D hypothesis
explain the modern distribution of human pigmentation
as the result of a balance between natural selection
favoring protection against sunburn and folate destruction
in areas where UVR exposure is high (see earlier),
and selection for lighter pigmentation in regions far
from the equator in order to facilitate vitamin D synthesis.
People with dark skin (e.g., Fitzpatrick’s skin phototype
V/VI) require at least 10 times more exposure to
sunlight than those with light skin (e.g., skin phototype
II/III) to produce the same amount of vitamin D in their
skin (Holick, 2003). This suggests that individuals with
darker skin would be at a selective disadvantage in synthesizing
vitamin D in areas of reduced UVR. Some
authors (e.g., Robins, 1991) have cast doubt on the valid-
91HUMAN PIGMENTATION VARIATION
Yearbook of Physical Anthropology—DOI 10.1002/ajpa
ity of the vitamin D hypothesis, mainly on the grounds
that in high latitude regions, synthesis of vitamin D during
the spring and summer months would be sufﬁcient
to produce stores of vitamin D in fat and muscle to
secure adequate levels of vitamin D in the winter. However,
the analysis of UVR levels using remote sensing
technology (Jablonski and Chaplin, 2000) indicates that
there is a broad region in the Northern Hemisphere
where there is not enough UVR to permit adequate cutaneous
synthesis of vitamin D, even in lightly pigmented
skin (see Fig. 6 for map corresponding to worldwide
UVR zones, based on the analysis of Jablonski and
Chaplin). Importantly, Robins’ assertion that vitamin D
synthesis in the spring and summer months would be
sufﬁcient to maintain adequate levels of vitamin D during
the winter is contradicted by numerous clinical studies
indicating that vitamin D insufﬁciency is present in
epidemic proportions at high latitudes, and that darkskinned
populations are at particular risk of vitamin D
insufﬁciency (Calvo and Whiting, 2003; Holick, 2005). In
fact, there is research indicating that vitamin D levels
are much lower throughout the year in dark-skinned
than light-skinned individuals living at high latitudes,
and these lower vitamin D levels cannot be explained by
differences in vitamin D intake (Harris and DawsonHughes,
1998). This issue will be discussed in more
detail in the section Implications of pigmentation for
public health. Overall, the vitamin D hypothesis seems
to be a reasonable explanation for the lower melanin levels
observed at high latitudes, although other evolutionary
factors may also have played a role (see following). It
is quite possible that the major selective mechanism was
the well-known effect of vitamin D on bone growth and
development (Beall and Steegmann, 2000). However, it
is important to note that research carried out in the last
decade clearly indicates that the role of vitamin D goes
well beyond calcium homeostasis. Vitamin D receptors
are present in kidney, keratinocytes, osteoblasts, activated
T and B lymphocytes, b-islet cells, small intestine,
prostate, colon, and most organs in the body (including
brain, heart, skin, gonads, prostate and breast) and vitamin
D plays an important role in the prevention of autoimmune
diseases, control of invading pathogens, and
regulation of cell growth and differentiation (MacDonald,
1999; Hansen et al., 2001; Omdahl et al., 2002; Holick,
2003, 2004; Zasloff, 2006).
Sexual selection
The idea that sexual selection could be responsible for
skin color variation has a long history. In his book The
descent of man, Charles Darwin (1871) stated that the
population differences observed for several human traits,
including pigmentation, could be the result of sexual
selection. Indeed, there is evidence indicating that pigmentation
is an important criterion of mate choice in
humans (Aoki, 2002 and references therein), leading
some authors to postulate that sexual selection has been
an important factor shaping the distribution of not only
skin, but also hair and eye color in humans (Diamond,
1991; Aoki, 2002, Frost et al., 2006). Similarly, the observation
that females tend to be lighter than males in
most populations has been interpreted as a result of sexual
selection (Van den Berghe and Frost, 1986, Frost,
1994), although this sexual dimorphism has been interpreted
by other authors to be a result of natural selection,
because females have higher calcium requirements
than males and this would favor lighter skin pigmentation
in order to increase vitamin D synthesis (Jablonski
and Chaplin, 2000). Of course, natural selection and sexual
selection are not mutually exclusive processes, and it
is possible that both have shaped the current distribution
of pigmentation in human populations, either as primary
or secondary factors. This possibility has been suggested
by multiple authors (Cavalli Sforza et al., 1994;
Frost, 1994; Aoki, 2002; Jablonski and Chaplin, 2000,
among others), although some authors put more emphasis
on natural selection and assign sexual selection a secondary
role (Jablonski and Chaplin, 2000), while others
believe that sexual selection had a more prominent role
(Diamond, 1991). In 2002 Aoki reformulated the sexual
selection hypothesis to explain the global distribution of
skin pigmentation in humans. According to Aoki (2002:
p. 592), the observed gradient of skin pigmentation is
due to ‘‘1) ubiquitous natural selection against light skin
that is strong near the equator, becomes progressively
weaker at higher latitudes, and is perhaps negligible at
the limits of human habitation; and 2) sexual preference
for light skin that is everywhere of roughly the same intensity.’’
Aoki (2002) criticized the vitamin D hypothesis,
using arguments similar to those used by Robins (1991)
and based his argument for sexual selection primarily on
a study by Van den Berghe and Frost (1986), indicating
that there is a preference for a lighter-than-average skin
color in 47 out of 51 societies, and that the preference
for lighter skin color is more strongly expressed in
males. Madrigal and Kelly (2006) recently tested, using
reﬂectance data, whether there is a positive correlation
between increasing distance from the equator and
increased sexual dimorphism, as predicted from Aoki’s
hypothesis, and they found no evidence to support his
prediction. In my judgment, the photoprotection and vitamin
D hypotheses satisfactorily explain the general latitudinal
gradient observed in human skin pigmentation
Fig. 6. Map showing the potential for synthesis of previtamin
D3 in lightly pigmented human skin, computed from annual
average UVMED values (which are indicated with different colors
in the online version of the article, and different shades of
gray in the printed version). UVMED is the quantity of UV radiation
(UVR) required to produce a barely perceptible reddening
of lightly-pigmented skin. In zone 1, in the tropics, there is
adequate UVR throughout the year to produce previtamin D3.
In zone 2 (light stippling), there is insufﬁcient UVR during at
least 1 month of the year to produce previtamin D3 in the skin.
In zone 3 (heavy stippling), there is not enough UVR for previtamin
D3 synthesis on average for the whole year. Reprinted
from Jablonski and Chaplin, 2000. The evolution of human skin
coloration. Journal of Human Evolution 39:57–106, with permission
from Elsevier. [Color ﬁgure can be viewed in the online issue,
which is available at www.interscience.wiley.com.]
92 E.J. PARRA
Yearbook of Physical Anthropology—DOI 10.1002/ajpa
and Aoki’s hypothesis, which requires a universal preference
for light skin color, both in time and space, has limited
support. However, this does not exclude the possibility
that sexual selection has shaped, to an extent, the pigmentation
variation observed in human populations.
In closing this section, it is necessary to mention that
there are other hypotheses explaining the distribution
of skin pigmentation in humans. For example, Wassermann
(1965) and more recently MacKintosh (2001)
linked dark skin pigmentation with resistance to bacterial,
parasitic, and viral infections, and Post et al. (1975)
suggested that depigmented skin would provide resistance
to cold injury.
THE GENETIC BASIS OF NORMAL
PIGMENTATION VARIATION
In spite of recent breakthroughs in our knowledge of
the pigmentary system, and the physiological and evolutionary
importance of pigmentation, very little is known
about the genetic basis of normal variation in human
pigmentation (Barsh, 2003). The pigmentation of unexposed
areas of the skin (constitutive pigmentation) is
fairly stable throughout the life of an individual and
does not change much due to environmental factors
(Robins, 1991). Constitutive pigmentation is a polygenic
trait, but the number of genes and the exact nature of
the allelic variants determining melanin content remain,
for the most part, unknown. Studies of human pigmentation
disorders, combined with studies of animal pigmentation
models, have greatly increased our knowledge of
the pigmentary system. These studies have brought a
new understanding of the diverse pathways driving melanin
production and regulation, including the following:
1) the tyrosinase enzyme complex (TYR, TRP1, DCT)
located on the membrane of the melanosome, which is
responsible for the enzymatic conversion of the amino
acid tyrosine into melanin; 2) other proteins located
within the melanosomes that play a critical role in melanogenesis
(MATP, P gene, SILV, SLC24A5); 3) the signaling
pathways affecting the regulation of melanin synthesis,
including hormones and receptors (a-MSH, MC1R,
ASIP, ATRN); 4) the transcription factors involved in
melanin production (PAX3, MITF, SOX10); 5) proteins
involved in melanosome transport (MYO5A, MYO7A,
RAB27A) and melanosome construction/protein routing
(CHS1, HPS1-6); and 6) developmentally important
ligands (EDN3, KITLG) and receptors (KIT, ENDRB),
which control melanoblast migration and differentiation.
Figure 7 shows a representation of a melanocyte indicating
some of the proteins involved in the pigmentary
system. A more complete list of the genes involved in
pigmentation is available at http://albinismdb.med.umn.
edu/genes.htm. Comprehensive reviews about the pigmentary
system and the genetics of pigmentation can be
found in Barsh (1996, 2003), Nordlund et al. (1998), and
Sturm et al. (2001). Given the polygenic inheritance of
pigmentation, and the complexity of the pigmentary
Fig. 7. Graphical representation of a melanocyte showing some important proteins involved in melanocyte development and the
pigmentation pathway, including the following: 1) the tyrosinase enzyme complex (TYR, TRP1, DCT) located on the membrane of
the melanosome, which is responsible for the enzymatic conversion of the amino acid tyrosine into melanin; 2) other proteins
located within the melanosomes that play a critical role in melanogenesis (MATP, P gene, SILV, SLC24A5); 3) the signaling pathways
affecting the regulation of melanin synthesis, including hormones and receptors (a-MSH, MC1R, ASIP, ATRN); 4) the transcription
factors involved in melanin production (PAX3, MITF, SOX10); 5) proteins involved in melanosome transport (MYO5A,
MYO7A, RAB27A) and melanosome construction/protein routing (CHS1, HPS1-6); and 6) developmentally important ligands
(EDN3, KITLG) and receptors (KIT, EDNRB) which control melanoblast migration and differentiation. Modiﬁed from Heather Norton
and Mark Shriver, with permission (personal communication, August 2007).
93HUMAN PIGMENTATION VARIATION
Yearbook of Physical Anthropology—DOI 10.1002/ajpa
system, it is not surprising that it has been very challenging
to identify the genes responsible for the variation
observed in skin, hair, and iris pigmentation. However,
the situation has changed dramatically in the last
decade, and several genes have been associated with normal
pigmentation variation. I summarize the available
evidence below, and also discuss novel strategies of data
analysis taking advantage of the large-scale catalogs of
genome variation (Hinds et al., 2005; The International
HapMap Consortium 2005) that will undoubtedly
advance our understanding of pigmentation variation in
years to come.
Genes involved in normal variation in skin, hair,
and eye color in human populations
MC1R. The melanocortin 1 receptor gene (MC1R) is
without a doubt the most exhaustively studied pigmentation
candidate gene. The role of this gene in normal
human pigmentation was clariﬁed by studies showing an
association of variants of MC1R with red hair and fair
skin (Valverde et al., 1995; Bastiaens et al., 2001;
Schaffer and Bolognia, 2001; Naysmith et al., 2004). The
melanocortin 1 receptor is a member of the family of G
protein-coupled receptors known as melanocortin receptors
and has a critical role in switching between eumelanin
and pheomelanin synthesis in the melanocytes.
Binding of the melanocyte-stimulating hormone (a-MSH)
to MC1R results in activation of adenylyl cyclase, and
increased levels of intracellular cAMP and tyrosinase
activity. The ultimate outcome is the production of eumelanin
within the melanocytes. Conversely, when the
antagonist agouti signaling protein (ASIP) binds to
MC1R, there is a decrease in tyrosinase activity and a
switch to pheomelanin production. The MC1R gene
shows an interesting pattern of polymorphism worldwide
(for a recent review, see Makova and Norton, 2005).
Rana et al. (1999) and Harding et al. (2000) sequenced
this gene in samples from several continents, and they
found a surprising lack of diversity in sub-Saharan African
populations; in particular, nonsynonymous variants
were absent in all the African samples analyzed. Similarly,
the frequency of amino acid variants in other darkskinned
populations, such as Papuans and South Asians,
was very low. This lack of diversity in dark-skinned populations
can be explained as a result of a strong functional
constraint on MC1R. Although a recent study by
John et al. (2003) reported three nonsynonymous mutations
in individuals from South Africa (farther south of
the equator than the previous African samples analyzed),
the overall picture is consistent with the action of
purifying selection removing MC1R mutations that could
promote pheomelanin synthesis in sub-Saharan Africa
and possibly in other regions with high UVR incidence
(Makova and Norton, 2005). The situation is very different
in Europe, East Asia, and Southeast Asia, where
MC1R is highly polymorphic. In fact, the levels of nucleotide
diversity observed in the MC1R gene in these populations
are higher than the values observed in other
autosomal genes (Makova and Norton, 2005). More than
30 alleles have been reported in European populations,
with more than 20 nonsynonymous variants (Sturm et
al., 2001). Importantly, at least nine nonsynonymous
variants are present at frequencies of 1% or higher in
European populations. Four of these mutations have a
strong association with the red hair/fair skin phenotype
(Asp84Glu, Arg151Cys, Arg160Trp, and Asp294His),
while three show only a weak association (Val60Leu,
Val92Met, and Arg163Gln) (Duffy et al., 2004). In vitro
studies have shown that some of these variants have
reduced ability to bind a-MSH (e.g., Val92Met) or to activate
adenylyl cyclase (e.g., Arg151Cys, Arg160Trp, and
Asp294His) (Xu et al., 1996; Fra¨ndberg et al., 1998;
Schio¨th et al., 1999). This explains the association of
these variants with hair and skin phenotypes characterized
by a rich content in pheomelanin. It is important to
mention that individuals harboring these functional variants
burn easily and have poor tanning capacity, and
numerous studies have indicated that these variants
increase the risk of melanoma and nonmelanoma skin
cancers (reviewed in Rees, 2004). Similarly, studies in
cultured melanocytes expressing these functional variants
of MC1R show a decrease in eumelanin production
and an increased sensitivity to the cytotoxic effects of
UVR (Scott et al., 2002). The importance of skin type
and MC1R variants in skin cancer risk will be further
discussed in the section Implications of pigmentation for
public health. The variation of the gene MC1R in Asian
populations has not been studied as extensively as in
European populations, but the available data indicate
high levels of polymorphism in East Asia and Southeast
Asia (Rana et al., 1999; Harding et al., 2000; Yao et al.,
2000; Peng et al., 2001; Nakayama et al., 2006). Asian
populations are characterized by high frequencies of two
nonsynonymous variants, Arg163Gln and Val92Met,
which are also present at much lower frequencies in European
populations (see Nakayama et al., 2006 for a
review of MC1R allele frequencies in East and Southeast
Asia). Interestingly, Nakayama et al. (2006) have
recently described three novel functional variants showing
very dramatic reduction in MC1R activities. These
variants are restricted to high latitudes, suggesting once
again that adaptation to UVR levels may have played an
active role in shaping the distribution of MC1R polymor-
phisms.
SLC24A5. The ‘‘golden’’ gene (SLC24A5) has been the
target of selection during the evolutionary process that
resulted in the lightening of the skin of European (and
closely related) populations (Lamason et al., 2005). The
role of this gene in human pigmentation variation was
identiﬁed in a somewhat atypical fashion; the story
began with researchers working with zebraﬁsh (Danio
rerio), a model organism used in laboratories around the
world. One of the zebraﬁsh mutants, the so-called golden
mutant, is characterized by a delayed and reduced development
of pigmentation with respect to the wild-type
zebraﬁsh (see Fig. 8). Using linkage analysis combined
with morpholino-knockdown experiments, Lamason et
al. (2005) identiﬁed slc24a5 as the gene responsible for
the zebraﬁsh golden mutant. The golden phenotype is
due to a premature stop codon that truncates the protein,
resulting in lightening of the pigmented stripes of
the zebraﬁsh. Interestingly, this protein is conserved in
other vertebrates and it was possible to rescue melanin
pigmentation by injecting human mRNA in golden
zebraﬁsh embryos. The golden gene encodes a cation
exchanger that seems to have a critical role in melanosome
morphogenesis and melanogenesis. When the
authors retrieved the information available at the HapMap
database (http://www.hapmap.org) about the human
SLC24A5 gene, they found a very unusual pattern of
polymorphism. Several SNPs, including a nonsynonymous
polymorphism (rs1426654) encoding alanine or
94 E.J. PARRA
Yearbook of Physical Anthropology—DOI 10.1002/ajpa
threonine, respectively, at amino acid 111 of the protein,
show extreme frequency differences between the HapMap
European sample and the West African and East
Asian samples. In the SNP rs1426654, the ancestral
allele, encoding alanine, is the most frequent allele in
African, and East Asian populations (93–100%), while
the derived allele encoding threonine is nearly ﬁxed
(98.7–100%) in European populations (Lamason et al.,
2005). In addition to this atypical pattern of differentiation,
in the European sample (but not the West African
or East Asian samples), there is a dramatic reduction of
heterozygosity encompassing a 150 kb region that includes
SLC24A5, indicating that there has been a recent
selective sweep in the ancestral European population
(Fig. 9A). The authors then decided to investigate
whether SNP rs1426654 plays a role in normal pigmentation
variation. With this aim, they tested whether the
rs1426654 polymorphism was associated with melanin
levels (measured by reﬂectometry) in two admixed populations:
African Americans and African Caribbeans.
Lamason et al. (2005) found that individuals with one or
two threonine alleles had lighter skin than homozygotes
for the ancestral alanine allele (Fig. 9B). It was estimated
that SLC24A5 explains 25–38% of the differences
in skin melanin index between populations of European
and African ancestry (Lamason et al., 2005). This
research emphasizes the usefulness of studies using
model organisms to understand the diversity of our species.
Additionally, the distribution of the SLC24A5 polymorphism
has important evolutionary implications. The
presence of the ancestral alanine allele at very high frequencies
and the lack of a selective signature in East
Asian populations in the SLC24A5 gene strongly suggest
that the decrease in melanin content took place, at least
in part, by different mechanisms in Europe and East
Asia. More recently, Norton et al. (2007) characterized
the rs1426654 polymorphism in the samples of the
Centre d’Etude du Polymorphisme Humain (CEPH) Diversity
Panel, which comprise 1059 individuals from 53
populations from different continents (Cann et al., 2002).
Figure 10 depicts the distribution of the ancestral alanine
and the derived threonine alleles throughout the
world. The ancestral alanine allele is present at very
high frequencies in sub-Saharan Africa, East Asia,
Southeast Asia, the Americas, and Melanesia. In contrast,
the derived threonine allele is ﬁxed in European
populations (frequency 5 100%) and is also present at
high frequencies in geographically proximate populations
in the Middle East, North Africa, and Pakistan (frequency
ranging from 62% to 100%).
Fig. 8. Photographs showing the differences in pigmentation
between the adult wild-type zebraﬁsh (A) and the golden zebraﬁsh
(B). Note the lighter stripes in the golden zebraﬁsh. The
melanophores of the golden zebraﬁsh (D) are thinner and contain
fewer melanosomes than do those of the wild-type (C).
Source: Lamason et al., 2005. Science 310:1782–1786, with permission
from AAAS. [Color ﬁgure can be viewed in the online
issue, which is available at www.interscience.wiley. com.]
Fig. 9. A: Evidence of a selective sweep in the region of the
SLC24A5 gene in European populations. Note the dramatic
reduction of heterozygosity in the European sample in the region
encompassing the SLC24A5 gene (enclosed by two arrows in the
ﬁgure), but not in the East Asian or West African samples. B:
Histograms showing the distribution of pigmentation in an African-American
and African-Caribbean sample, after adjusting for
ancestry for each genotype of the SLC24A5 rs1426654 SNP. GG
homozygotes have higher melanin levels than AG heterozygotes
and AA homozygotes (the difference is 9.5 and 7 melanin units,
respectively). Source: Lamason et al., 2005. Science 310:1782–
1786, with permission from AAAS. [Color ﬁgure can be viewed in
the online issue, which is available at www.interscience.wiley.
com.]
95HUMAN PIGMENTATION VARIATION
Yearbook of Physical Anthropology—DOI 10.1002/ajpa
MATP (also known as SLC45A2 or AIM1). The MATP
(membrane-associated transporter protein) plays an important
role in murine pigmentation, and some variants
of this gene lead to generalized hypopigmentation of the
eyes and fur. In humans, mutations in the MATP gene
are responsible for oculocutaneous albinism type 4
(OCA4). Recent research indicates that MATP shows a
pattern of polymorphism that is remarkably similar to
what is observed in SLC24A5: the leucine variant of the
nonsynonymous SNP rs16891982 (encoding leucine or
phenylalanine, respectively, at amino acid position 374 of
the protein) is mainly restricted to European (and closely
related) populations (Nakayama et al., 2002; Graf et al.,
2005; Norton et al., 2007), and MATP shows a strong signature
of selection in European populations. Figure 11
shows the distribution of the rs16891982 polymorphism
in the samples of the CEPH Diversity panel, based on
Norton et al. data (2007). Two recent studies indicate
that MATP plays a role in normal pigmentation variation
in humans. Graf et al. (2005) have shown that the
SNP rs16891982 is strongly associated with dark skin,
dark hair and dark eye color in Caucasians. Another
nonsynonymous polymorphism (rs26722) was also associated
with darker pigmentation. Norton et al. (2007) have
also shown that the derived G (leucine) allele of the
rs16891982 polymorphism is strongly associated with
lighter skin pigmentation in a sample of African Americans
(P \ 0.0001). It was estimated that the ancestral C
(phenyalanine) allele increases skin pigmentation by
approximately ﬁve melanin units and the distribution of
pigmentation values per genotype is consistent with an
additive mode of action.
OCA2 (also known as the P gene). The human OCA2
gene is associated with a form of albinism (oculocutaneous
albinism type 2). However, evidence is accumulating
that this gene also may play a role in normal variation
in skin and, more particularly, in iris pigmentation.
Akey et al. (2001) described that interaction between the
MC1R and P genes contributes to interindividual variation
in skin pigmentation in Tibetans, while Shriver et
al. (2003) reported that a variant in the P gene was associated
with skin pigmentation in a sample of African
Americans and African Caribbeans. OCA2 also seems to
have an important role in iris color. In 1996, Eiberg and
Mohr found strong indication of linkage of a region on
chromosome 15q, where OCA2 is located, with eye color
and attributed this signal to the OCA2 gene. Zhu et al.
(2004) carried out a sib pair linkage analysis in 502 twin
families, and most of the variation in eye color was due
to a quantitative trait locus (QTL) on chromosome 15q,
near OCA2 (LOD score 5 19.2). This result was recently
replicated by Posthuma et al. (2006). In 2002, Rebbeck
et al. reported that two OCA2 polymorphisms
(Arg305Trp and Arg419Gln) were associated with eye
color. Similarly, Frudakis et al. (2003) reported that several
SNPs and haplotypes located within the pigmentation
candidate genes OCA2, MYO5A, TYRP1, MATP,
DCT, and TYR were associated with iris pigmentation.
Associations of SNPs located within OCA2 were by far
the most signiﬁcant of any gene tested, with 13 SNPs
showing strong associations. Recently, Duffy et al. (2007)
have described a very strong association of three SNPs
within intron 1 of the OCA2 gene (rs7495174 T/C,
rs6497268 G/T, and rs11855019 T/C) with blue eye color.
These three SNPs are located in a major haplotype
block, and the diplotype (combination of haplotypes)
TGT/TGT seems to be a major determinant of iris color,
with a frequency of more than 90% in subjects with blue
or green iris compared to less than 10% in individuals
with brown eye color. In summary, the current evidence
clearly indicates that OCA2 is a major determinant of
variation in eye color and also plays a role in skin pigmentation
variation.
ASIP. As discussed in the previous section, the agouti
signaling protein (ASIP) is the antagonist of MC1R, and
upon binding to this receptor, ASIP promotes the syntheFig.
10. Distribution of the SLC24A5 A111G polymorphism (rs1426654) in the samples of the CEPH-Diversity panel. The G
allele (Ala) is indicated in blue in the online version of the article, and dark gray in the printed version; the A allele (Thr) is indicated
in yellow in the online version, and light gray in the printed version. Reproduced from Norton et al., 2007. Genetic evidence
for the convergent evolution of light skin in Europeans and East Asians. Molecular Biology and Evolutuion 24:710–722 by permission
of Oxford University Press. [Color ﬁgure can be viewed in the online issue, which is available at www.interscience.wiley.com.]
96 E.J. PARRA
Yearbook of Physical Anthropology—DOI 10.1002/ajpa
sis of pheomelanin. A SNP located in the 30
-untranslated
region (UTR) of this gene, g.8818A?G, has been associated
with pigmentation phenotypes in humans. In particular,
it appears that the ASIP 8818G allele is signiﬁcantly
associated with dark hair, brown eyes, and dark
skin (Kanetsky et al., 2002; Bonilla et al., 2005). Interestingly,
a recent study by Voisey et al. (2006) indicates
that the level of ASIP mRNA is around 12 times lower in
melanocyte cell lines carrying the G allele (homozygotes
GG or heterozygotes AG) than in cell lines with the AA
genotype. Lower levels of the agouti protein would result
in decreased antagonism of a-MSH binding to MC1R and
therefore increased production of eumelanin.
TYR. Recent research indicates that the tyrosinase gene
(TYR), which encodes a key enzyme in melanogenesis
and is responsible for one of the two most common types
of albinism (oculocutaneous albinism type 1, OCA1), is
also involved in normal pigmentation variation. Shriver
et al. (2003) described that a nonsynonymous TYR polymorphism
(rs1042602) was associated with skin pigmentation
(measured by reﬂectometry) in a sample of African
Americans and African Caribbeans.
Detecting signatures of selection in pigmentation
candidate genes: A promising strategy to identify
genes involved in normal pigmentation variation
I would like to close this section by reviewing a promising
new strategy for identifying genes involved in normal
pigmentation variation using tests to detect signatures
of selection in pigmentation candidate genes. The
idea behind this approach is that pigmentation candidate
genes showing evidence of selection are, or have
been in the past, important from the functional point of
view, and can potentially explain the variation observed
in this phenotype. Indeed, the example of the SLC24A5
gene clearly shows that this approach has promise to
uncover genes involved in pigmentation variation. The
action of natural selection leaves a ‘‘signature’’ in the genome,
distorting the patterns of genetic variation with
respect to neutral expectations. For example, when
directional selection has been active in a genomic region,
increasing the frequency of an advantageous mutation
and linked neutral sites (e.g., a selective sweep), there
will be a decrease in nucleotide diversity, an excess of
rare variants and an excess of linkage disequilibrium
with respect to what is expected under a neutral model
(Fay and Wu, 2000; Przeworski, 2002). Numerous tests
have been proposed to detect positive selection using
data on genetic variation within a species. Some tests
evaluate the departure of the allele frequency distribution
with respect to neutral expectations (Tajima’s D, Fu
and Li’s D and F, Fay and Wu’s H). Other tests evaluate
the decrease in heterozygosity (natural log ratio of heterozygosities,
LnRH), or the presence of atypical levels of
population differentiation (FST; Locus-Speciﬁc Branch
Length test; LSBL; Informativeness of Assignment
index, In). Finally, another group of tests evaluates the
decay of linkage disequilibrium (Extented Haplotype
Homozygosity test, EHH; iHS test; Linkage Disequilibrium
Decay test, LDD). For further information about
these tests, refer to Biswas and Akey (2006), McEvoy et
al. (2006), and Harris and Meyer (2006). However, the
identiﬁcation of loci that have been the target of selection
(directional or balancing) is not straightforward
because demographic factors can mimic the patterns of
variation produced by selection. As an example, an
excess of rare variants with respect to neutral expectations
can be due to selective sweeps or population expansions.
Conversely, an excess of intermediate-frequency
alleles can be due to balancing selection or population
bottlenecks (Akey et al., 2004; Williamson et al., 2005).
There are two main strategies to solve these potential
problems. The ﬁrst strategy is to incorporate demographic
history in the statistical models (Toomajian
et al., 2003; Williamson et al., 2005). The difﬁculty with
this approach is that demographic history can be very
Fig. 11. Distribution of the MATP C374G polymorphism (rs16891982) in the samples of the CEPH-Diversity panel. The C allele
(Phe) is indicated in red in the online version of the article, dark gray in the printed version; the G allele (Leu) is indicated in
yellow in the online version, and light gray in the printed version. Reproduced from Norton et al., 2007. Genetic evidence for the
convergent evolution of light skin in Europeans and East Asians. Molecular Biology and Evolutuion 24:710–722 by permission of
Oxford University Press. [Color ﬁgure can be viewed in the online issue, which is available at www.interscience.wiley.com.]
97HUMAN PIGMENTATION VARIATION
Yearbook of Physical Anthropology—DOI 10.1002/ajpa
complex (and therefore difﬁcult to model), and it is generally
unknown. The second strategy is to compare the
patterns of genetic variation of one region with those
found in the rest of the genome (Biswas and Akey, 2006).
The underlying assumption of this approach is that demographic
history will affect the entire genome, while
selection will have an effect on speciﬁc loci. This method
also has potential problems and under certain demographic
and selective scenarios, the false discovery rate
can be quite large (Kelley et al., 2006; Teshima et al.
2006). However, a recent simulation study indicates that
this is a reliable approach to identify genes that have
been the target of natural selection (Kelley et al., 2006).
In fact, there have been several recent genome-wide
scans for positive selection based on the Perlegen (Hinds
et al., 2005) and HapMap data (http://www.hapmap.org)
that have identiﬁed hundreds of genes showing evidence
of positive selection, and in spite of the different methods
used to detect selection signatures, there is considerable
overlap between the studies (Carlson et al., 2005;
The International HapMap Consortium, 2005; Voight et
al., 2006; Wang et al., 2006; Kelley et al., 2006). Of particular
relevance for this review are four recent studies
that speciﬁcally applied outlier approaches in order to
identify pigmentation candidate genes showing evidence
of positive selection in samples of East Asian, European,
and West African ancestry (Izagirre et al., 2006; McEvoy
et al., 2006; Lao et al., 2007; Myles et al., 2007). These
studies used a variety of tests to identify signatures of
selection, including the Tajima’s D- (TD), FST-, LSBL-,
In-, LnRH-, and EHH-based tests. Several pigmentation
candidate genes had exceptionally unusual patterns of
variation, and in many cases the results were consistent
using different tests. Some pigmentation candidate genes
were also identiﬁed as outliers in a previous genomewide
scan published by Voight et al. (2006), using the
iHS test. Table 1 shows the list of pigmentation genes
for which the results were signiﬁcant in two or more
studies, using different types of tests (for a full list of
genes, refer to Izagirre et al., 2006; McEvoy et al., 2006;
Myles et al., 2007; and Lao et al., 2007). The fact that
these genes are outliers using tests based on different
characteristics of the data (allele frequency distribution,
heterozygosity, genetic differentiation, and decay of linkage
disequilibrium) strongly indicates that they have
been the target of positive selection. There are 10 genes
in the list, which can be classiﬁed in the following categories:
1) genes involved in melanogenesis (SLC24A5,
MATP, OCA2, DCT, and TYRP1); 2) genes involved in
melanosome construction/protein routing (LYST); and 3)
genes involved in melanocyte development and differentiation
(MITF, KITLG, ADAM17, and ADAMTS20). Previous
association studies have indicated that 3 of these
10 genes (SLC24A5, MATP, and OCA2, labeled in boldface
in Table 1) play a role in pigmentation variation.
One gene shows signatures of selection in the three samples
analyzed (MITF), ﬁve genes in only two populations
(OCA2 and KITLG in the East Asian and European sample,
TYRP1 and DCT in the European and West African
sample, and LYST in the West African and East Asian
sample), and four genes show signatures of selection in
only one sample (SLC24A5 and MATP in the European
sample and ADAM17 and ADAMTS20 in the East Asian
sample). It is important to emphasize that these results
have to be interpreted with caution: although the evidence
of positive selection appears to be strong, some of
these genes have been implicated in biological processes
other than pigmentation and it is possible that the selection
signatures identiﬁed in these genes may be the
result of positive selection related to other gene functions.
For example, the ADAM17 gene has been implicated
in many processes involved in cell–cell and cell–
matrix interactions, including fertilization, muscle development,
and neurogenesis. It is also possible that some
of these results are just false positives (although this
possibility is minimized by the inclusion in the list of
genes that are outliers for at least two tests based on different
summary statistics). In order to conﬁrm the role
of these genes in normal pigmentation variation, genotype/phenotype
association studies and/or direct functional
studies will be required. To my knowledge, this
research has not yet been completed for 7 of the 10
genes (DCT, TYRP1, LYST, MITF, KITLG, ADAM17,
and ADAMTS20). However, these studies focusing on
the identiﬁcation of potential signatures of selection offer
great promise to identify new genes involved in normal
population variation. The preliminary data point to a
complex picture where positive selection has been acting
at different genome locations, and for some genes only in
certain population groups. It is interesting to note that
although some genes show signatures of selection in
more than one population, in some cases there is evidence
indicating that these selective events were independent.
For example, the gene OCA2 shows signatures
of selection in European and East Asian populations, but
the core haplotypes showing evidence of extended homoTABLE
I. List of pigmentation candidate genes showing signatures of natural selection, using different testsa
Gene Test Population Referenceb
SLC24A5 (Solute carrier family 24, member 5) LSBL, TD, lnRH, FST, EHH, iHS Eur 1–4
MATP (SLC45A2, Solute carrier family 45, member 2) LSBL, TD, lnRH, FST, In, EHH Eur 2–5
OCA2 (P gene) LSBL, FST, In, EHH Eur, Eas 2–5
DCT (Dopachrome tautomerase) LSBL, FST, In, EHH Eas, Afr 2–5
TYRP1 (Tyrosinase-related Protein 1) LSBL, FST, In, EHH, iHS, Eur, Afr 1–5
LYST (Lysosomal Trafﬁcking Regulator) LSBL, TD, lnRH, FST, EHH Eas, Afr 2,3
MITF (Microphthalmia-associated transcription factor) LSBL, TD, lnRH, FST, EHH Eur, Eas, Afr 2,4
KITLG (KIT ligand) TD, In, EHH Eur, Eas 2,5
ADAM17 (ADAM metallopeptidase domain 17) LSBL, TD, lnRH, FST Eas 2,3
ADAMTS20 (ADAM metallopeptidase with
thrombospondin type 1 motif)
LSBL, TD, lnRH, FST Eas 2,3
a
Polymorphisms within the genes SLC24A5, MATP, and OCA2 (in boldface) have been associated with normal pigmentation variation
in previous studies.
b
1, Voight et al. (2006); 2, McEvoy et al. (2006); 3, Izagirre et al. (2006); 4, Myles et al. (2007); 5, Lao et al. (2007).
Abbreviations: LSBL, Locus-Speciﬁc Branch Length test; TD, Tajima’s D test; ln RH, natural logarithm of the Ratio Of Heterozygosities
test; FST, FST test; In, Informativeness of Assignment test; EHH, Extended Haplotype Homozygosity test; his, iHS test.
98 E.J. PARRA
Yearbook of Physical Anthropology—DOI 10.1002/ajpa
zygosity are different in each population (Lao et al.,
2007). Similarly, the two haplotypes showing evidence of
selection at the DCT gene in East Asians and West Africans
are different (Lao et al., 2007). These results highlight
the complex evolutionary history of skin pigmentation
in human populations, which we are just beginning
to understand. A tentative evolutionary model for skin
pigmentation pertaining to the three populations (East
Asian, European, and West African) that have been studied
using these outlier approaches has been recently proposed
by McEvoy et al. (2006).
IMPLICATIONS OF PIGMENTATION FOR
PUBLIC HEALTH
In a previous section, I summarized the main evolutionary
hypotheses that have been put forward to
explain the distribution of skin pigmentation in human
populations. Although there is still some debate regarding
the selective factors involved, it is generally accepted
that the strong association between latitude and pigmentation
in humans is primarily the result of the action of
natural selection promoting the adaptation of human
populations to local environmental conditions (in particular
UVR incidence), a process that likely took hundreds
of generations. However, as a consequence of recent
human migrations, many individuals are now living in
geographic regions with different environmental conditions
than those in which their populations evolved.
Because of the key role that pigmentation plays in photoprotection
and vitamin D synthesis, these recent
migrations have important implications for public
health: fair-skinned individuals are at higher risk of several
types of skin cancer, particularly in regions with
high UVR incidence, and dark-skinned individuals living
in high latitude regions are at higher risk for diseases
caused by deﬁcient or insufﬁcient vitamin D levels.
These issues are further discussed below.
Fair pigmentation and increased risk of
skin cancer
Most of the harmful effects of sunlight on the skin
(e.g., erythema and DNA damage) arise from exposure to
the UV wavelength (Rees, 2004). Melanins are an efﬁcient
natural sunblock, particularly in the UV spectrum,
so it is not surprising that the risk of developing skin
cancer can vary by as much as 100-fold, and is strongly
related to skin color (Rees and Flanagan, 1999). The
incidences of skin cancer and skin cancer mortality correlate
strongly with latitude, with both decreasing further
away from the equator (Mielke et al., 2006). Furthermore,
skin cancer incidence is higher in fair-skinned
populations than dark-skinned populations at the same
latitude. Nowhere is this more evident than in Australia,
the country with the highest incidence of skin cancer in
the world. This applies to both nonmelanoma skin cancer
(basal cell carcinoma and squamous cell carcinoma) and
malignant melanoma. The incidence of these three types
of skin cancer is around 10 times higher in Australia
(particularly the northern regions such as Queensland
and the Northern Territory) than in Northern Europe
(Diepgen and Mahler, 2002; Staples et al., 2006). It has
been estimated that one in two Australians will develop
some form of skin cancer in his (or her) lifetime (http://
www.cancercouncil.com.au/). However, skin cancer disproportionately
affects fair-skinned individuals and the
Australian aboriginals have a much lower incidence of
skin cancers (Sturm, 2002). A recent study has indicated
that cutaneous melanin density at the upper inner arm,
measured quantitatively by spectrophotometry, is a
strong predictor of skin cancer risk, particularly in men
(Dwyer et al., 2002). Similarly, skin cancer risk is
strongly associated with skin type; individuals with skin
type I or II (who burn easily and do not tan) have higher
risk than individuals with other skin types (Armstrong
and Kricker, 1995). The gene MC1R seems to be a key
link in this relationship. As discussed in the section
devoted to the genetic basis of pigmentation, several
MC1R variants present in polymorphic frequencies
([1%) in European populations show a substantial reduction
in MC1R activity and are associated with red
hair and fair skin. Individuals with these MC1R variants
have poor tanning ability and are prone to burn
and to freckle, which are well known risk factors for
nonmelanoma skin cancers and malignant melanoma
(Sturm, 2002). It is not surprising, then, that the major
‘‘red-hair’’ alleles Arg151Cys, Arg160Trp, and Asp294His
signiﬁcantly increase the risk of developing melanoma
(having one allele increases the risk twofold, and having
two alleles fourfold) (Sturm, 2002). Similarly, these variants
also increase the risk for basal cell carcinoma,
squamous cell carcinoma, and solar keratosis (odds ratio
approximately 3, Sturm, 2002). The role of MC1R polymorphisms
in skin cancer risk is also supported by
recent research indicating that melanocytes expressing
the variants with decreased MC1R function are more
sensitive to the cytotoxic effects of UVR (Scott et al.,
2002). Overall, there is very strong evidence indicating
that the fair skin typical of northern latitudes (e.g.,
Northern Europe) is a signiﬁcant risk factor for different
types of skin cancer, particularly in regions with high
UVR incidence. The incidence of skin cancers has been
increasing dramatically for several decades and has
become an important public health concern not only in
Australia, but also in Europe and North America
(Diepgen and Mahler, 2002; Tucker and Goldstein,
2003). Consequently, numerous prevention programs
have been established to lessen the impact of this disease
(Saraiya et al., 2004; Stanton et al., 2004).
Dark pigmentation and increased risk of vitamin
D insufﬁciency and deﬁciency
Synthesis of vitamin D via exposure of the skin to the
sun is the most abundant source of vitamin D for most
people (Holick, 2003, 2005). UVB radiation (280–320 nm)
penetrates the epidermis and photolyzes 7-dehydrocholesterol
(7-DHC) to previtamin D3, which then undergoes
thermal isomerization to form vitamin D3 (Webb, 2006;
Chen et al., 2007). Vitamin D3 enters the circulation
bound to the vitamin D-binding protein (DBP), and
undergoes two obligate hydroxylations, the ﬁrst in the
liver to 25-hydroxyvitamin D3 [25(OH)D3], and the second
in the kidney to its active form, 1,25-dihydroxyvitamin
D3 [1,25(OH)D3] (Holick, 2005). Given the properties
of melanin as a natural UVR ﬁlter, it is not surprising
that the amount of melanin is inversely related to vitamin
D production in the skin. According to Holick et al.
(2003), people with dark skin (e.g., skin phototype V/VI)
require at least 10 times more exposure to sunlight than
those with light skin (e.g., skin phototype II/III) to produce
the same amount of vitamin D in their skin. Similarly,
a recent report by Chen et al. (2007) has demon-
99HUMAN PIGMENTATION VARIATION
Yearbook of Physical Anthropology—DOI 10.1002/ajpa
strated that the conversion of epidermal 7-DHC to previtamin
D3 is 5–10 times more efﬁcient in skin phototype
II than in skin phototype V. Therefore, vitamin D synthesis
can be compromised by large amounts of melanin,
particularly under conditions of limited UVR exposure.
Such is the case in high-latitude regions, where there is
not enough UVR to synthesize vitamin D for a substantial
period of the year (see Fig. 6). Jablonski and Chaplin
(2000) did a systematic analysis in which they described
the geographic distribution of the potential for vitamin
D synthesis. At latitudes around 408 north (Boston),
there is insufﬁcient UVB radiation for vitamin D synthesis
from November to early March, while at latitudes
farther north, like Edmonton, Canada (528 north), this
‘‘vitamin D winter’’ extends from mid-October to midMarch
(Webb et al., 1988). Recent research indicates
that there is a surprisingly high prevalence of vitamin D
insufﬁciency (typically deﬁned as serum levels of 25hydroxyvitamin
D [25(OH)D] lower than 40–50 nmol/l)
in high-latitude countries, even among light-skinned
individuals (Lamberg-Allardt et al., 2001; Vieth et al.,
2001; Rucker et al., 2002; Tangpricha et al., 2002;
Rejnmark et al., 2004). However, the prevalence of vitamin
D insufﬁciency is higher in population groups with
higher average levels of melanin. A recent review indicates
that in the United States, at equivalent latitudes
(25–34.58N), the prevalence of vitamin D insufﬁciency is
highest in African Americans (52–76%), intermediate in
Hispanics (18–50%), and lowest in European Americans
(8–31%) (Calvo and Whiting, 2003). Scragg et al. (1995)
also reported that in New Zealand (latitude 408S), Paciﬁc
Islanders and Maories had signiﬁcantly lower concentrations
of 25(OH)D than individuals of European ancestry,
after controlling for age, sex, and time of the year. The
majority of these studies have not controlled for vitamin
D intake, which is known to vary signiﬁcantly among
ethnic groups (Calvo et al., 2005). However, a study published
by Harris and Dawson-Hughes (1998) reported
that in Boston (latitude 408 north), plasma 25(OH)D levels
were substantially lower in African American women
than in European American women throughout the year,
even after adjusting for body weight and vitamin D
intake. Unfortunately, none of these studies has evaluated
directly the effect of skin pigmentation, measured
quantitatively using spectrometry, on vitamin D levels.
Therefore, there is very limited information on how melanin
levels inﬂuence the serum concentration of vitamin
D metabolites, after controlling for relevant covariates,
such as vitamin D intake (from diet and supplements)
and UV exposure.
The current epidemic of vitamin D insufﬁciency is of
public health relevance not only because of the wellknown
effect of vitamin D on bone metabolism, but also
its signiﬁcant role in protection against numerous
chronic conditions. Recent research indicates that many
tissues other than the liver and kidney have vitamin D
receptors, and are capable of locally synthesizing the
active vitamin D metabolite, 1,25(OH)2D3, from the precursor
25(OH)D (Peterlik and Cross, 2005; Hollis and
Wagner, 2006). This extrarenal pathway is critical to
understanding the effects of vitamin D on immunomodulation
and regulation of cell growth and development
(see Fig. 12). Adequate vitamin D levels are important
not only for protection against rickets, osteomalacia, or
osteoporosis, but also for protection against several types
of cancer (e.g., breast, colon, prostate cancer), autoimmune
diseases (e.g., rheumatoid arthritis, systemic lupus
erythematosus, multiple sclerosis), cardiovascular disease
and microbial infections (e.g., tuberculosis) (for
recent reviews, see Holick, 2004; Peterlik and Cross,
2005; and Giovannucci, 2005). According to most vitamin
D experts, desirable 25(OH)D serum concentration is
!75 nmol/l, and the current vitamin D supplementation
recommendations in Canada and the United States (200
international units or IU/day up to age 50 years, 400 IU/
day for age 50–70 years, and 600 IU/day over age 70
years) are insufﬁcient for optimal health, so there is
urgent need to re-evaluate these recommendations (Vieth
et al., 2007). A recent randomized clinical trial (Lappe
et al., 2007) has shown that treatment of postmenopausal
women with calcium plus 1,100 IU of vitamin D3 per day
dramatically decreases cancer risk compared to women
in the placebo control group (RR: 0.402, P 5 0.01). This
and other studies have motivated a recent announcement
by the Canadian Cancer Society, recommending that
adults living in Canada should consider taking 1,000 IU/
day of vitamin D supplementation during the fall and the
winter, and adults at higher risk of having lower vitamin
D levels (people who are older, those with dark skin,
those who do not go outside very often, and/or people who
wear clothing that covers most of their skin) should consider
taking 1,000 IU/day all year around.
Because of the important autocrine and endocrine
functions of vitamin D, ensuring optimal vitamin D levFig.
12. Organs and tissues
capable of synthesis of the active
form of vitamin D, 1,25(OH2)D3,
showing its multiple physiological
roles. Reproduced from Hollis
and Wagner, 2006. Nutritional
vitamin D status during pregnancy:
reasons for concern—
reprinted from, CMAJ 174(9):
1287–1290 by permission of the
publisher. VVC 2006 Canadian Medical
Association. [Color ﬁgure
can be viewed in the online
issue, which is available at www.
interscience.wiley. com.]
100 E.J. PARRA
Yearbook of Physical Anthropology—DOI 10.1002/ajpa
els has become an important public health issue. More
research is needed to increase our understanding of how
factors such as age, diet, skin pigmentation, geographic
location, and intensity of the sun affect the production of
vitamin D and to better deﬁne the optimal amount of
vitamin D supplementation required to prevent health
problems in individuals of diverse ancestry.
CONCLUSIONS AND INSIGHTS: WHY STUDY
OF PIGMENTATION IS IMPORTANT
I have summarized our current understanding of the
evolution and the genetic basis of pigmentation in our
species. Pigmentation shows a remarkable diversity in
human populations, and in this sense, it is an atypical
trait; numerous genetic studies have indicated that the
average proportion of genetic variation due to differences
among major continental groups is just 10–15% of the
total genetic variation. In contrast, skin pigmentation
shows large differences among continental populations.
The reasons for this discrepancy can be traced back primarily
to the strong inﬂuence of natural selection, which
has shaped the distribution of pigmentation according to
a clear latitudinal gradient. It is therefore erroneous to
assume that the large population differences observed
for skin pigmentation can be extrapolated to most other
human traits. In fact, traits that have been strongly
shaped by natural selection are notoriously unsuitable to
decipher population relationships. Classiﬁcations based
on traits subject to strong natural selection reﬂect the
underlying environmental factors, instead of population
history. In this particular case, a classiﬁcation derived
from skin pigmentation mainly captures population differences
in UVR exposure, and shows very poor concordance
with classiﬁcations based on neutral markers,
which are the proper tools to use when mapping human
history. It is therefore unfortunate that pigmentation
has been an omnipresent trait in ‘‘racial’’ classiﬁcations
that fail to capture or explain the diversity of the human
species. However, the study of the genes underlying
human pigmentation diversity brings to the forefront the
mosaic nature of human genetic variation. Our genome
is composed of a myriad of segments with different patterns
of variation and evolutionary histories. Due to the
recent origin of our species, most of these segments show
small differences among human populations, but some
genomic regions, particularly those that have been under
the inﬂuence of natural selection, depart considerably
from the average picture of the human genome and
show large differences among populations. These
genomic regions probably played a key role in the process
of adaptation of human populations to different
environments and may also explain some of the population
differences that have been described in disease risk
or drug metabolism. Therefore, in the same way that it
is erroneous to extrapolate the patterns of variation
observed in superﬁcial traits such as pigmentation to the
rest of the genome, it is misleading to suggest, based on
the ‘‘average’’ genomic picture, that variation among
human populations is irrelevant. In my view, it is paradoxical
that, well into the 21st century and into the
post-human genome project era, we know so little about
the genetic basis of eye, hair, and skin pigmentation.
There are many reasons why elucidation of the genetics
and evolutionary history of pigmentation is important. 1)
Pigmentation is a trait that should be used as an example
of how misleading simplistic interpretations of human
variation can be, and to educate the general public
on the importance of studying variation using an evolutionary,
rather than a typological, approach. 2) Pigmentation
can be very useful to understand the genetic
architecture of complex traits. The pigmentation of unexposed
areas of the skin (constitutive pigmentation) is relatively
unaffected by environmental inﬂuences during
an individual’s lifetime, when compared with other complex
traits such as diabetes or blood pressure, and this
provides a unique opportunity to study gene–gene interactions
without the effect of environmental confounders.
3) Pigmentation is of relevance from a public health perspective,
because of its critical role in photoprotection
and vitamin D synthesis. Our understanding of the
genes involved in normal pigmentation variation in
human populations has increased substantially in the
last 5 years, and recent studies have identiﬁed several
pigmentation candidate genes that show strong signatures
of natural selection and therefore, may be functionally
important in pigmentation. At the current pace of
discovery, the next decade will clarify many of the
remaining gaps in our knowledge of the genetic architecture
and evolutionary history of this fascinating trait.
ACKNOWLEDGMENTS
The author thanks Emily Cameron for administrative
assistance; and Heather Norton, Heather MacLean,
Agnes Gozdzik, Jodi Barta, and three anonymous reviewers
for their comments and suggestions on a previous
version of the manuscript. The author has research
support from the Natural Sciences and Engineering
Research Council of Canada (NSERC), the Canada Foundation
for Innovation (CFI), the Ontario Innovation
Trust (OIT), the Canadian Institutes of Health Research
(CIHR), and the Government of Ontario Early Research
Award (ERA).
LITERATURE CITED
Akey JM, Eberle MA, Rieder MJ, Carlson CS, Shriver MD,
Nickerson DA, Kruglyak L. 2004. Population history and
natural selection shape patterns of genetic variation in 132
genes. PLoS Biol2:e286.
Akey JM, Wang H, Xiong M, Wu H, Liu W, Shriver MD, Jin L.
2001. Interaction between the melanocortin-1 receptor and P
genes contributes to inter-individual variation in skin pigmentation
phenotypes in a Tibetan population. Hum Genet 108:
516–520.
Alaluf S, Atkins D, Barrett K, Blount M, Carter N, Heath A.
2002. The impact of epidermal melanin on objective measurements
of human skin colour. Pigment Cell Res 15:119–126.
Alaluf S, Heath A, Carter N, Atkins D, Mahalingam H, Barrett
K, Kolb R, Smit N. 2001. Variation in melanin content and
composition in type V and VI photoexposed and photoprotected
human skin: the dominant role of DHI. Pigment Cell
Res 14:337–347.
Aoki K. 2002. Sexual selection as a cause of human skin color
variation: Darwin’s hypothesis revisited. Ann Hum Biol 29:
589–608.
Armstrong BK, Kricker A. 1995. Skin cancer. Dermatol Clin. 13:
583–594.
Barsh GS. 1996. The genetics of pigmentation: from fancy genes
to complex traits. Trends Genet 12:299–305.
Barsh GS. 2003. What controls variation in human skin color?
PLoS Biol 1:19–22.
Bastiaens M, ter Huurne J, Gruis N, Bergman W, Westendorp
R, Vermeer BJ, Bouwes Bavinck JN. 2001. The melanocortin-
1 receptor gene is the major freckle gene. Hum Mol Genet
10:1701–1708.
101HUMAN PIGMENTATION VARIATION
Yearbook of Physical Anthropology—DOI 10.1002/ajpa
Beall CM, Steegmann Jr. T. 2000. Human adaptation to climate:
temperature, ultraviolet radiation, and altitude. In: Stinson S,
Bogin B, Huss-Ashmore R, O’Rourke D, editors. Human biology:
an evolutionary and biocultural perspective. New Cork:
Wiley-Liss. p 163–224.
Biasutti R. 1953. Le razze e i popoli della terra. Torino: Unione
Tipograﬁco Editrice Torinese.
Birdsell JB. 1993. Microevolutionary patterns in aboriginal Australia.
New York: Oxford University Press.
Biswas S, Akey JM. 2006. Genomic insights into positive selection.
Trends Genet 22:437–446.
Bolognia JL, Orlow SJ. 2003. Melanocyte biology. In: Bolognia
JL, Jorizzo JL, Rapini RP, editors. Dermatology. London:
Elsevier-Mosby. p 935–945.
Bonilla C, Boxill LA, Donald SA, Williams T, Sylvester N, Parra
EJ, Dios S, Norton HL, Shriver MD, Kittles RA. 2005. The
8818G allele of the agouti signaling protein (ASIP) gene is ancestral
and is associated with darker skin color in African
Americans. Hum Genet 116:402–406.
Branda RF, Eaton JW. 1978. Skin color and nutrient photolysis—evolutionary
hypothesis. Science 201:625–626.
Bustamante J, Bredeston L, Malanga G, Mordoh J. 1993. Role
of melanin as a scavenger of active oxygen species. Pigment
Cell Res 6:348–353.
Byard PJ. 1981. Quantitative genetics of human skin color.
Yrbk Phys Anthropol 24:123–137.
Calvo MS, Whiting SJ. 2003. Prevalence of vitamin D insufﬁciency
in Canada and the United States: importance to health
status and efﬁcacy of current food fortiﬁcation and dietary
supplement use. Nutr Rev 61:107–113.
Calvo MS, Whiting SJ, Barton CN. 2005. Vitamin D intake: a
global perspective of current status. J Nutr 135:310–316.
Cann HM, de Toma C, Cazes L, Legrand MF, Morel V, Piouffre
L, Bodmer J, Bodmer WF, Bonne-Tamir B, Cambon-Thomsen
A, Chen Z, Chu J, Carcassi C, Contu L, Du R, Excofﬁer L,
Ferrara GB, Friedlander JS, Groot H, Gurwitz D, Jenkins T,
Herrera RJ, Huang X, Kidd J, Kidd KK, Langaney A, Lin AA,
Mehdi SQ, Parham P, Piazza A, Pistillo MP, Qian Y, Shu Q,
Xu J, Zhu S, Weber JL, Greely HT, Feldman MW, Thomas G,
Dausset J, Cavalli-Sforza LL. 2002. A human genome diversity
cell line panel. Science 296:261–262.
Carlson CS, Thomas DJ, Eberle MA, Swanson JE, Livingston
RJ, Rieder MJ, Nickerson DA. 2005. Genomic regions exhibiting
positive selection identiﬁed from dense genotype data.
Genome Res 15:1553–1565.
Cavalli Sforza LL, Menozzi P, Piazza A. 1994. The history and geography
of human genes. Princeton: Princeton University Press.
Chen TC, Chimeh F, Lu Z, Mathieu J, Person KS, Zhang A,
Khon N, Martinello S, Berkowitz R, Holick MF. 2007. Factors
that inﬂuence the cutaneous synthesis and dietary sources of
vitamin D. Arch Biochem Biophys 460:213–217.
Cosentino MJ, Pakyz RE, Fried J. 1990. Pyrimethamine—an
approach to the development of a male contraceptive. Proc
Natl Acad Sci USA 87:1431–1435.
Damian DL, Halliday GM, Barnetson RS. 1997. Prediction of
minimal erythema dose with a reﬂectance melanin meter. Br
J Dermatol 136:714–718.
Darwin C. 1871. The descent of man. Murray: London.
Diamond J. 1991. The rise and fall of the third chimpanzee.
London: Radius.
Diepgen TL, Mahler V. 2002. The epidemiology of skin cancer.
Br J Dermatol 146 (Suppl 61):24–30.
Diffey BL, Oliver RJ, Farr PM. 1984. A portable instrument for
quantifying erythema induced by ultraviolet radiation. Br J
Dermatol 3:663–672.
Duffy D, Box N, Chen W, Palmer JS, Montgomery GW, James
MR, Hayward NK, Martin NG, Sturm RA. 2004. Interactive
effects of MC1R and OCA2 on melanoma risk phenotypes.
Hum Mol Genet 13:447–461.
Duffy DL, Montgomery GW, Chen W, Zhao ZZ, Le L, James
MR, Hayward NK, Martin NG, Sturm RA. 2007. A three-single-nucleotide
polymorphism haplotype in intron 1 of OCA2
explains most human eye-color variation. Am J Hum Genet
80:241–252.
Dwyer T, Blizzard L, Ashbolt R, Plumb J, Berwick M, Stankovich
JM. 2002. Cutaneous melanin density of Caucasians
measured by spectrophotometry and risk of malignant melanoma,
basal cell carcinoma, and squamous cell carcinoma of
the skin. Am J Epidemiol 155:614–621.
Ebisch IM, Pierik FH, DE Jong FH, Thomas CM, SteegersTheunissen
RP. 2006. Does folic acid and zinc sulphate intervention
affect endocrine parameters and sperm characteristics
in men? Int J Androl 29:339–345.
Eiberg H, Mohr J. 1996. Assignment of genes coding for brown
eye colour (BEY2) and brown hair colour (HCL3) on chromosome
15q. Eur J Hum Genet 4:237–241.
Fan S, Dyer CR, Hubbard L. 2003. Quantiﬁcation and correction
of iris color. Technical Report 1495, Department of Computer
Sciences, University of Wisconsin-Madison.
Fay JC, Wu CI. 2000. Hitchhiking under positive Darwinian
selection. Genetics 155:1405–1413.
Fitzpatrick TB. 1988. The validity and practicality of sun-reactive
skin types I through VI. Arch Dermatol 124:869–871.
Foley R. 1998. The context of human genetic evolution. Genome
Res 8:339–347.
Fra¨ndberg P, Doufexis M, Kapas S, Chhajlani J. 1998. Human
pigmentation phenotype: a point mutation generates nonfunctional
MSH receptor. Biochem Biophys Res Commun 245:490–
492.
Frost P. 1994. Geographic distribution of human skin colour: a
selective compromise between natural selection and sexual
selection? Hum Evol 9:141–153.
Frost P. 2006. European hair and eye color: a case of frequencydependent
sexual selection? Evol Hum Behav 27:85–103.
Frudakis T, Thomas M, Gaskin Z, Venkateswarlu K, Chandra
KS, Ginjupalli S, Gunturi S, Natrajan S, Ponnuswamy VK,
Ponnuswamy KN. 2003. Sequences associated with human
iris pigmentation. Genetics 165:2071–2083.
German EJ, Hurst MA, Wood D, Gilchrist J. 1998. A novel system
for the objective classiﬁcation of iris colour and its correlation
with response to 1% tropicamide. Ophthal Physiol Opt
18:103–110.
Giovannucci E. 2005. The epidemiology of vitamin D and cancer
incidence and mortality: a review (United States). Cancer
Causes Control 16:83–95.
Graf J, Hodgson R, van Daal A. 2005. Single nucleotide polymorphisms
in the MATP gene are associated with normal
human pigmentation variation. Hum Mutat 25:278–284.
Halliday GM. 2005. Inﬂammation, gene mutation and photoimmunosuppression
in response to UVR-induced oxidative damage
contributes to photocarcinogenesis. Mutat Res 571:107–120.
Hansen CM, Binderup L, Hamberg KJ, Carlberg. 2001. Vitamin
D and cancer: effects of 1,25(OH)2D3 and its analogs on
growth control and tumorigenesis. Front Biosci 6:D820–D848.
Harding RM, Healy E, Ray AJ, Ellis NS, Flanagan N, Todd C,
Dixon C, Sajantila A, Jackson IJ, Birch-Machin MA, Rees JL.
2000. Evidence of variable selective pressures at MC1R. Am J
Hum Genet 66:1351–1361.
Harris EE, Meyer D. 2006. The molecular signature of selection
underlying human adaptations. Am J Phys Anthropol (Suppl
43):89–130.
Harris SS, Dawson-Hughes B. 1998. Seasonal changes in
plasma 25-hydroxyvitamin D concentrations of young American
black and white women. Am J Clin Nutr 67:1232–1236.
Hill HZ, Hill GJ. 2000. UVA, pheomelanin and the carcinogenesis
of melanoma. Pigment Cell Res 13 (8 Suppl):140–
144.
Hinds DA, Stuve LL, Nilsen GB, Halperin E, Eskin E, Ballinger
DG, Frazer KA, Cox DR. 2005. Whole-genome patterns of
common DNA variation in three human populations. Science
307:1072–1079.
Holick MF. 2003. Vitamin D: a millennium perspective. J Cell
Biochem 88:296–307.
Holick MF. 2004. Vitamin D: importance in the prevention of
cancers, type 1 diabetes, heart disease and osteoporosis. Am J
Clin Nutr 79:362–371.
Holick MF. 2005. The vitamin D epidemic and its health consequences.
J Nutr 135:2739S–2748S.
102 E.J. PARRA
Yearbook of Physical Anthropology—DOI 10.1002/ajpa
Holick MF. 2007. Vitamin D deﬁciency. N Engl J Med 357:266–
281.
Hollis BW, Wagner CL. 2006. Nutritional vitamin D status during
pregnancy: reasons for concern. Can Med Assoc J 174:
1287–1290.
Hunter RS. 1942. Photoelectric tristimulus colorimetry with
three ﬁlters. Circular of the National Bureau of Standards
C429. J Opt Soc Am 32:509–538.
Ito S, Fujita K. 1985. Microanalysis of eumelanin and pheomelanin
in hair and melanomas by chemical degradation and liquid
chromatography. Anal Biochem 144:527–536.
Izagirre N, Garcia I, Junquera C, de la Rua C, Alonso S. 2006.
A scan for signatures of positive selection in candidate loci for
skin pigmentation in humans. Mol Biol Evol 23:1697–1706.
Jablonski NG. 2004. The evolution of human skin and skin
color. Annu Rev Anthropol 33:585–623.
Jablonski NG, Chaplin G. 2000. The evolution of human skin
coloration. J Hum Evol 39:57–106.
John P, Makova K, Li WH, Jenkins T, Ramsay M. 2003. DNA
polymorphism and selection at the melanocortin-1 receptor
gene in normally pigmented southern African individuals.
Ann N Y Acad Sci 994:299–306.
Jorde LB, Wooding SP. 2004. Genetic variation, classiﬁcation
and ‘‘race.’’ Nat Genet 36 (11 Suppl):S28–S33.
Kanetsky PA, Swoyer J, Panossian S, Holmes R, Guerry D,
Rebbeck TR. 2002. A polymorphism in the agouti signaling
protein gene is associated with human pigmentation. Am J
Hum Genet 70:770–775.
Kelley JL, Madeoy J, Calhoun JC, Swanson W, Akey JM. 2006.
Genomic signatures of positive selection in humans and the
limits of outlier approaches. Genome Res 16:980–989.
Kollias N, Baqer A. 1985. Spectroscopic characteristics of human
melanin in vivo. J Invest Dermatol 85:38–42.
Kollias N, Baqer AH. 1987. Absorption mechanism of human
melanin in the visible, 400-720 nm. J Invest Dermatol 89:
384–388.
Kollias N, Sayer RM, Zeise L, Chedekel MR. 1991. Photoprotection
by melanin. J Photochem Photobiol B Biol 9:135–160.
Kollias N, Stamatas GN. 2002. Optical non-invasive approaches
to diagnosis of skin diseases. J Invest Dermatol 7:64–75.
Kongshoj B, Thorleifsson A, Wulf HC. 2006. Pheomelanin and
eumelanin in human skin determined by high-performance
liquid chromatography and its relation to in vivo reﬂectance
measurements. Photodermatol Photoimmunol Photomed 22:
141–147.
Krol ES, Liebler DC. 1998. Photoprotective actions of natural
and synthetic melanins. Chem Res Toxicol 11:1434–1440.
Lamason RL, Mohideen MA, Mest JR, Wong AC, Norton HL,
Aros MC, Jurynec MJ, Mao X, Humphreville VR, Humbert JE,
Sinha S, Moore JL, Jagadeeswaran P, Zhao W, Ning G, Makalowska
I, McKeigue PM, O’Donnell D, Kittles R, Parra EJ,
Mangini NJ, Grunwald DJ, Shriver MD, Canﬁeld VA, Cheng
KC. 2005. SLC24A5, a putative cation exchanger, affects pigmentation
in zebraﬁsh and humans. Science 310: 1782–1786.
Lamberg-Allardt CJ, Outila TA, Karkkainen MU, Rita HJ, Valsta
LM. 2001. Vitamin D deﬁciency and bone health in
healthy adults in Finland: could this be a concern in other
parts of Europe? J. Bone Miner Res 16:2066–2073.
Lao O, de Gruijter JM, van Duijn K, Navarro A, Kayser M.
2007. Signatures of positive selection in genes associated with
human skin pigmentation as revealed from analyses of single
nucleotide polymorphisms. Ann Hum Genet 71:354–369.
Lappe JM, Travers-Gustafson D, Davies KM, Recker RR, Heaney
RP. 2007. Vitamin D and calcium supplementation
reduces cander risk: results of a randomized trial. Am J Clin
Nutr 85:1586–1591.
Leenutaphong V. 1995. Relationship between skin color and cutaneous
response to ultraviolet radiation in Thai. Photodermatol
Photoimmunol Photomed 11:198–203.
Lees FC, Byard PJ. 1978. Skin colorimetry in Belize. I. Conversion
formulae. Am J Phys Anthropol 48:515–522.
Lees FC, Byard PJ, Relethford JH. 1979. New conversion formulae
for light-skinned populations using photovolt and
E.E.L. reﬂectometers. Am J Phys Anthropol 51:403–408.
Lewontin R. 1972. The apportionment of human diversity. In:
Hecht MK, Steere WS, editors. Evolutionary biology, Vol. 6.
New York: Plenum.
Loomis WF. 1967. Skin-pigment regulation of vitamin-D biosynthesis
in man. Science 157:501–506.
Lucock M. 2000. Folic acid: nutritional biochemistry, molecular
biology, and role in disease processes. Mol Genet Metab 71:
121–138.
MacDonald PN. 1999. Molecular biology of the vitamin D receptor.
In: Holick MF, editor. Vitamin D: molecular biology, physiology,
and clinical applications. Totowa, NJ: Humana. p 109–128.
MacKintosh JA. 2001. The antimicrobial properties of melanocytes,
melanosomes and melanin and the evolution of black
skin. J Theor Biol 211:101–113.
Madrigal L, Kelly W. 2006. Human skin-color sexual dimorphism:
a test of the sexual selection hypothesis. Am J Phys
Anthropol 132:470–482.
Makova K, Norton H. 2005. Worldwide polymorphism at the
MC1R and normal pigmentation variation in humans. Peptides
26:1901–1908.
Mathur U, Datta SL, Mathur BBL. 1977. Effect of aminopterininduced
folic-acid deﬁciency on spermatogenesis. Fertil Steril
28:1356–1360.
McEvoy B, Beleza S, Shriver MD. 2006. The genetic architecture
of normal variation in human pigmentation: an evolutionary
perspective and model. Hum Mol Genet 15:R176–R181.
Melgosa M, Rivas MJ, Gomez L, Lita E. 2000. Towards a colorimetric
characterization of the human iris. Ophthal Physiol
Opt 20:252–260.
Meredith P, Sarna T. 2006. The physical and chemical properties
of eumelanin. Pigment Cell Res 19:572–594.
Mielke JH, Konisberg LW, Relethford JH. 2006. Human biological
variation. New York: Oxford University Press.
Myles S, Somel M, Tang K, Kelso J, Stoneking M. 2007. Identifying
genes underlying pigmentation differences among human
populations. Hum Genet 120:613–621.
Nakayama K, Fukamachi S, Kimura H, Koda Y, Soemantri A,
Ishida T. 2002. Distinctive distribution of AIM1 polymorphism
among major human populations with different skin color.
J Hum Genet 47:92–94.
Nakayama K, Soemantri A, Jin F, Dashnyam B, Ohtsuka R,
Duanchang P, Isa MN, Settheetham-Ishida W, Harihara S,
Ishida T. 2006. Identiﬁcation of novel functional variants of
the melanocortin 1 receptor gene originated from Asians.
Hum Genet 119:322–330.
Naysmith L, Waterston K, Ha T, Flanagan N, Bisset Y, Ray A,
Wakamatsu K, Ito S, Rees JL. 2004. Quantitative measures of
the effect of the melanocortin 1 receptor on human pigmentary
status. J Invest Dermatol 122:423–428.
Niggemann B, Weinbauer G, Vogel F, Korte R. 2003. A standardized
approach for iris color determination. Int J Toxicol 22:
49–51.
Nordlund JJ, Bossy RE, Hearing VJ, King RA, Ortonne J-P,
editors. 1998. The pigmentary system: physiology and pathophysiology.
New York: Oxford University Press.
Norton HL, Kittles RA, Parra E, McKeigue P, Mao X, Chen K,
Canﬁeld VA, Bradley D, McEvoy B, Shriver MD. 2007.
Genetic evidence for the convergent evolution of light skin in
Europeans and East Asians. Mol Biol Evol 24:710–722.
Off MK, Steindal AE, Porojnicu AC, Juzeniene A, Vorobey A,
Johnsson A, Moan J. 2005. Ultraviolet degradation of folic
acid. J Photochem Photobiol B 80:47–55.
Omdahl JL, Morris HA, May BK. 2002. Hydroxylase enzymes of
the vitamin D pathway: expression, function, and regulation.
Annu Rev Nutr 22:139–166.
Ortonne JP, Prota G. 1993. Hair melanins and hair color: ultrastructural
and biochemical aspects. J Invest Dermatol 101 (1
Suppl):82S–89S.
Ozeki H, Ito S, Wakamatsu K. 1996. Chemical characterization
of melanins in sheep wool and human hair. Pigment Cell Res
9:51–57.
Peng S, Lu XM, Luo HR, Xiang-Yu JG, Zhang YP. 2001. Melanocortin-1
receptor gene variants in four Chinese ethnic populations.
Cell Res 11:81–84.
103HUMAN PIGMENTATION VARIATION
Yearbook of Physical Anthropology—DOI 10.1002/ajpa
Peterlik M, Cross HS. 2005. Vitamin D and calcium deﬁcits predispose
for multiple chronic diseases. Eur J Clin Invest 35:
290–304.
Post PW, Daniels F, Binford RT. 1975. Cold injury and the evolution
of ‘white’ skin. Hum Biol 39:131–143.
Posthuma D, Visscher PM, Willemsen G, Zhu G, Martin NG,
Slagboom PE, de Geus EJ, Boomsma DI. 2006. Replicated
linkage for eye color on 15q using comparative ratings of sibling
pairs. Behav Genet 36:12–17.
Przeworski M. 2002. The signature of positive selection at randomly
chosen loci. Genetics 160:1179–1189.
Rana BK, Hewett-Emmett D, Jin L, Chang BH, Sambuughin N,
Lin M, Watkins S, Bamshad M, Jorde LB, Ramsay M, Jenkins
T, Li WH. 1999. High polymorphism at the human melanocortin
1 receptor locus. Genetics 151:1547–1557.
Rebbeck TR, Kanetsky PA, Walker AH, Holmes R, Halpern AC,
Schuchter LM, Elder DE, Guerry D. 2002. P gene as an inherited
biomarker of human eye color. Cancer Epidemiol Biomarkers
Prev 11:782–784.
Rees JL. 2003. Genetics of hair and skin color. Annu Rev Genet
37:67–90.
Rees JL. 2004. The genetics of sun sensitivity in humans. Am J
Hum Genet 75:739–51.
Rees JL, Flanagan N. 1999. Pigmentation, melanocortins, and
red hair. Q J Med 92:125–131.
Rejnmark L, Jorgensen ME, Pedersen MB, Hansen JC, Heickendorff
L, Lauridsen AL, Mulvad G, Siggaard C, Skjoldborg
H, Sorensen TB, Pedersen EB, Mosekilde L. 2004. Vitamin D
insufﬁciency in Greenlanders on a westernized fare: ethnic
differences in calcitropic hormones between Greenlanders and
Danes. Calcif Tissue Int 74:255–263.
Relethford JH. 1997. Hemispheric difference in human skin
color. Am J Phys Anthropol 104:449–457.
Relethford JH. 2002. Apportionment of global human genetic diversity
based on craniometrics and skin color. Am J Phys
Anthropol 118:393–398.
Robins AH. 1991. Biological perspectives on human pigmentation.
Cambridge, NY: Cambridge University Press.
Rucker D, Allan JA, Fick GH, Hanley DA. 2002. Vitamin D
insufﬁciency in a population of healthy western Canadians.
Can Med Assoc J 166:1517–1524.
Samokhvalov A, Hong L, Liu Y, Garguilo J, Nemanich RJ,
Edwards GS, Simon JD. 2005. Oxidation potentials of human
eumelanosomes and pheomelanosomes. Photohem Photobiol
81:145–148.
Saraiya M, Glanz K, Briss PA, Nichols P, White C, Das D,
Smith SJ, Tannor B, Hutchinson AB, Wilson KM, Gandhi N,
Lee NC, Rimer B, Coates RC, Kerner JF, Hiatt RA, Bufﬂer P,
Rochester P. 2004. Interventions to prevent skin cancer by
reducing exposure to ultraviolet radiation: a systematic
review. Am J Prev Med 27:422–466.
Schaffer JV, Bolognia JL. 2001. The melanocortin-1 receptor:
red hair and beyond. Arch Dermatol 137:1477–1485.
Schio¨th HB, Phillips SR, Rudzish R, Birch-Machin MA, Wikberg
JE, Rees JL. 1999. Loss of function mutations of the
human melanocortin 1 receptor are common and are associated
with red hair. Biochem Biophys Res Commun 260:488–
491.
Scott MC, Wakamatsu K, Ito S, Kadekaro AL, Kobayashi N,
Groden J, Kavanagh R, Takakuwa T, Virador V, Hearing VJ,
Abdel-Malek ZA. 2002. Human melanocortin 1 receptor variants,
receptor function and melanocyte response to UVR radi-
ation.115:2349–2355.
Scragg R, Holdaway I, Singh V, Metcalf P, Baker J, Dryson E.
1995. Serum 25-hydroxyvitamin D3 is related to physical activity
and ethnicity but not obesity in a multicultural workforce.
Aust N Z J Med 25:218–223.
Shriver MD, Parra EJ. 2000. Comparison of narrow-band reﬂectance
spectroscopy and tristimulus colorimetry for measurements
of skin and hair color in persons of different biological
ancestry. Am J Phys Anthrop 112:17–27.
Shriver MD, Parra EJ, Dios S, Bonilla C, Norton H, Jovel C,
Pfaff C, Jones C, Massac A, Cameron N, Baron A, Jackson T,
Argyropoulos G, Jin L, Hoggart CJ, McKeigue PM, Kittles
RA. 2003. Skin pigmentation, biogeographical ancestry and
admixture mapping. Hum Genet 112:387–399.
Slominski A, Wortsman J, Plonka PM, Schallreuter KU, Paus
R, Tobin DJ. 2005. Hair follicle pigmentation. J Invest Dermatol
124:13–21.
Snellman E, Jansen CT, Leszczynski K, Visuri R, Milan T,
Jokela K. 1995. Ultraviolet erythema sensitivity in anamnestic
(I-IV) and phototested (1-4) Caucasian skin phototypes: the
need for a new classiﬁcation system. Photochem Photobiol
62:769–772.
Stamatas GN, Zmudzka BZ, Kollias N, Beer JZ. 2004. Non-invasive
measurements of skin pigmentation in situ. Pigment Cell
Res 17:618–626.
Stanton WR, Janda M, Baade PD, Anderson P. 2004. Primary
prevention of skin cancer: a review of sun protection in Australia
and internationally. Health Promot Int 19:369–378.
Staples MP, Elwood M, Burton RC, Williams JL, Marks R, Giles
GG. 2006. Non-melanoma skin cancer in Australia: the 2002
national survey and trends since 1985. Med J Aust 184:6–10.
Sturm RA. 2002. Skin colour and skin cancer—MC1R, the
genetic link. Melanoma Res 12:405–416.
Sturm RA, Frudakis TN. 2004. Eye colour: portals into pigmentation
genes and ancestry. Trends Genet 20:327–332.
Sturm RA, Teasdale RD, Box NF. 2001. Human pigmentation
genes: identiﬁcation, structure and consequences of polymorphic
variation. Gene 277:49–62.
Szabo G, Gerarld AB, Pathak MA, Fitzpatrick TB. 1969. Racial
differences in the fate of melanosomes in human epidermis.
Nature 222:1081–1082.
Takeuchi S, Zhang W, Wakamatsu K, Ito S, Hearing VJ,
Kraemer KH, Brash DE. 2004. Melanin acts as a potent UVB
photosensitizer to cause an atypical mode of cell death in murine
skin. Proc Natl Acad Sci USA 101:15706–15801.
Tangpricha V, Pearce EN, Chen TC, Holick MF. 2002. Vitamin
D insufﬁciency among free-living healthy young adults. Am J
Med 112:659–662.
Teshima KM, Coop G, Przeworski M. 2006. How reliable are
empirical genomic scans for selective sweeps? Genome Res.
16:702–712.
The International HapMap Consortium. 2005. A haplotype map
of the human genome. Nature 437:1299–1320.
Thody AJ, Higgins EM, Wakamatsu K, Ito S, Burchill SH,
Marks JM. 1991. Pheomelanin as well as eumelanin is present
in human epidermis. J Invest Dermatol 97:340–344.
Tishkoff SA, Kidd KK. 2004. Implications of biogeography of
human populations for ‘‘race’’ and medicine. Nat Genet 36 (11
Suppl):S21–S27.
Tobin DJ, Bystryn JC. 1996. Different populations of melanocytes
are present in hair follicles and epidermis. Pigment Cell
Res 9:304–310.
Tobin DJ, Paus R. 2001. Graying: gerontology of the hair follicle
pigmentary unit. Exp Gerontol 36:29–54.
Toomajian C, Ajioka RS, Jorde LB, Kushner JP, Kreitman M.
2003. A method for detecting recent selection in the human
genome from allele age estimates. Genetics 165:287–297.
Tucker MA, Goldstein AM. 2003. Melanoma etiology: where are
we? Oncogene 22:3042–3052.
Ullrich SE. 2005. Mechanisms underlying UV-induced immune
suppression. Mutat Res 571:185–205.
Valverde P, Healy E, Jackson I, Rees JL, Thody AJ. 1995. Variants
of the melanocyte-stimulating hormone receptor gene
are associated with red hair and fair skin in humans. Nat
Genet 11:328–330.
Van Den Berghe PL, Frost P. 1986. Skin color preference, sexual
dimorphism and sexual selection: a case of gene culture coevolution?.
Ethnic Racial Stud 9:87–113.
Vieth R, Bischoff-Ferrari H, Boucher BJ, Dawson-Hughes B,
Garland CF, Heaney RP, Holick MF, Hollis BW, LambergAllardt
C, McGrath JJ, Norman AW, Scragg R, Whiting SJ,
Willett WC, Zittermann A. 2007. The urgent need to recommend
an intake of vitamin D that is effective. Am J Clin Nutr
85:649–650.
Vieth R, Cole DE, Hawker GA, Trang HM, Rubin LA. 2001.
Wintertime vitamin D insufﬁciency is common in young Cana-
104 E.J. PARRA
Yearbook of Physical Anthropology—DOI 10.1002/ajpa
dian women, and their vitamin D intake does not prevent it.
Eur J Clin Nutr 55:1091–1097.
Voight BF, Kudaravalli S, Wen X, Pritchard JK. 2006. A map of
recent positive selection in the human genome. PLoS Biol 4:e72.
Voisey J, Gomez-Cabrera M del C, Smit DJ, Leonard JH, Sturm
RA, van Daal A. 2006. A polymorphism in the agouti signaling
protein (ASIP) is associated with decreased levels of
mRNA. Pigment Cell Res 19:226–231.
Wang ET, Kodama G, Baldi P, Moyzis RK. 2006. Global landscape
of recent inferred Darwinian selection for Homo sapiens.
Proc Natl Acad Sci USA 103:135–140.
Wassermann HP. 1965. Human pigmentation and environmental
adaptation. Arch Environ Health 11:691–694.
Weatherall IL, Coombs BD. 1992. Skin color measurements in
terms of CIELAB color space values. J Invest Dermatol 99:
468–473.
Webb AR. 2006. Who, what, where and when—inﬂuences on cutaneous
vitamin D synthesis. Prog Biophys Mol Biol 92:17–25.
Webb AR, Kline L, Holick MF. 1988. Inﬂuence of season and latitude
on the cutaneous synthesis of vitamin-D3—exposure to
winter sunlight in Boston and Edmonton will not promote
vitamin-D3 synthesis in human-skin. J Clin Endocrinol Metab
67:373–378.
Weiner JS. 1951. A spectrophotometer for measurement of skin
colour. Man 51:563–568.
Westerhof W, Estevez-Uscanga O, Meens J, Kammeyer A, Durocq
M, Cario I. 1990. The relation between constitutional
skin color and photosensitivity estimated from UV-induced erythema
and pigmentation dose-response curves. J Invest Dermatol
94:812–816.
Whiteman DC, Parsons PG, Green AC. 1999. Determinants of
melanocyte density in adult human skin. Arch Dermatol Res
291:511–516.
Wielgus AR, Sarna T. 2005. Melanin in human irides of different
color and age of donors. Pigment Cell Res 18:454–464.
Williamson SH, Hernandez R, Fledel-Alon A, Zhu L, Nielsen R,
Bustamante CD. 2005. Simultaneous inference of selection
and population growth from patterns of variation in the
human genome. Proc Natl Acad Sci USA 102:7882–7887.
Wong WY, Merkus HM, Thomas CM, Menkveld R, Zielhuis GA,
Steegers-Theunissen RP. 2002. Effects of folic acid and zinc
sulfate on male factor subfertility: a double-blind, randomized,
placebo-controlled trial. Fertil Steril 77:491–498.
Xu X, Thornwall M, Lundin L, Chhajlani V. 1996. Val92Met variant
of the melanocyte stimulating hormone receptor gene.
Nat Genet 14:384.
Yakubu A, Mabogunje OA. 1993. Skin cancer in African albinos.
Acta Oncol 32:621–622.
Yao YG, Lu XM, Luo HR, Li WH, Zhang YP. 2000. Gene admixture
in the Silk Road region of China: evidence from mtDNA
and melanocortin 1 receptor polymorphisms. Genes Genet
Syst 75:173–178.
Ye T, Hong L, Garguilo J, Pawlak A, Edwards GS, Nemanich
RJ, Sarna T, Simon JD. 2006. Photoionization thresholds of
melanins obtained from free electron laser-photoelectron emission
microscopy, femtosecond transient absorption spectroscopy
and electron paramagnetic resonance measurements of
oxygen photoconsumption. Photochem Photobiol 82:733–737.
Zasloff M. 2006. Fighting infections with vitamin D. Nat Med
12:388–390.
Zhu G, Evans DM, Duffy DL, Montgomery GW, Medland SE,
Gillespie NA, Ewen KR, Jewell M, Liew YW, Hayward NK,
Sturm RA, Trent JM, Martin NG. 2004. A genome scan for
eye color in 502 twin families: most variation is due to a QTL
on chromosome 15q. Twin Res 7:197–210.
105HUMAN PIGMENTATION VARIATION
Yearbook of Physical Anthropology—DOI 10.1002/ajpa