Original Research Article
Sex Ratios in the Arctic—Do Man-Made Chemicals Matter?
PETER BJERREGAARD,1*SUSAN CHATWOOD,2,3BRYANY DENNING,2,4LAWRENCE JOSEPH,5AND T. KUE YOUNG3
1Centre for Health Research in Greenland, National Institute of Public Health, University of Southern Denmark
2Institute for Circumpolar Health Research, Yellowknife, Northwest Territories, Canada
3Dalla Lana School of Public Health, University of Toronto, Toronto, Canada
4Canadian Public Health Service, Office of Public Health Practice, Public Health Agency of Canada Yellowknife, Northwest Territories, Canada
5Department of Epidemiology, Biostatistics and Occupational Health, McGill University, Montre ´al, Canada
Objectives: The objective was to analyze the variation of secondary sex ratios across the Arctic and to estimate the
time trend. The rationale for this was claims in news media that, in the Arctic, sex ratios have become reduced due to
exposure to anthropogenic contaminants in the environment.
Methods: Data was collected from 27 circumpolar jurisdictions from public websites of the eight Arctic countries.
Sex ratios at birth were calculated for each jurisdiction and each available year. Linear regression models of the sex
ratios across time were fit within each jurisdiction to estimate the change in sex ratio over time.
Results: All male:female sex ratios were close to 1.05 with time trends close to 0. In a Bayesian hierarchical model
overall sex ratio was estimated at 1.054 (95% confidence interval 1.048, 1.058). The estimate for the 10-year slope across
all jurisdictions was 0.0010 (95% confidence interval 20.0021, 0.0046). Separate analyses of indigenous populations in
Alaska and Greenland gave similar results and similar sex ratios were found among Greenland Inuit in 1900 and today.
Conclusions: The absence of deviation of the secondary sex ratio in any of the Arctic jurisdictions indicates that the
contaminants that are present are not disrupting endocrine systems to the extent that sex ratios are being affected.
Am. J. Hum. Biol. 24:165–169, 2012.
' 2012 Wiley Periodicals, Inc.
The Arctic environment is undergoing rapid changes as
a result of global warming and resource development,
especially in minerals, oil, and gas. It is also affected by
long-range transportation by air and ocean currents of
contaminants from distant industrial sources. The physi-
cal and social environments interact with human biology,
lifestyles, and behaviors to influence the health of human
populations in the Arctic (Young and Bjerregaard, 2008).
In 2007, the British newspaper Guardian declared in a
headline that twice as many girls than boys were being
born in Greenland and Arctic Russia, and that man-made
chemicals were to blame (Brown, 2007). We traced the
source of the alarming story to anecdotal information
from one village in northern Greenland and a report from
the Arctic Monitoring and Assessment Programme on per-
sistent toxic substances among indigenous people of the
Russian North (Arctic Monitoring and Assessment Pro-
gramme, 2004). While we share the increasing concern of
many northern residents over environmental pollution in
many parts of the Arctic, we believe that concern over
health effects should be based on facts. We wanted to ana-
lyze this before it became a generally acknowledged truth
that man-made chemicals were responsible for changing
sex ratios in the Arctic. The first step was to investigate if
there had indeed been any deviation from the ‘‘normal’’
secondary sex ratio (i.e., ratio of male to female live births)
across 27 Arctic jurisdictions.
The secondary sex ratio is remarkably consistent across
human populations, between 1.05 and 1.07, i.e., 105–107
males for every 100 females born (Hesketh and Xing,
2006). In the absence of selective termination of preg-
nancy favoring one sex over another for sociocultural rea-
sons, substantial deviation of the ratio from this average
is suggestive of threats to human reproductive health
from the external environment or health conditions in the
parents. Trends in declining proportions of male births
have been reported from a number of countries (Grech
et al., 2003), including those in high latitudes such as Nor-
way (Irgens and Irgens, 2003), Finland (Vartianen et al.,
1999), and Canada (Allan et al., 1997), though the magni-
tude of the decline varies. Among potential explanatory
mechanisms proposed, the endocrine disruptive effects of
persistent organic pollutants (POPs), including polychlori-
nated biphenyls (PCBs), during conception and pregnancy
have received much attention (James, 2006).
In addition to the sources cited above, much of the evi-
dence for the association between pollutants and sex ratio
has been based on small community studies or the after-
math of environmental incidents (Hertz-Picciotto et al.,
2008; Mackenzie et al., 2005; Mocarelli et al., 1996). Other
studies, however, failed to demonstrate an association
between sex ratio and exposure to environmental or toxic
agents (Karmaus et al., 2002; Taylor et al., 2006; Terrell
et al., 2009; Vartianinen et al., 1999; Yoshimura et al., 2001).
The exposure to environmental contaminants and its
potential health effects are issues of major concern in the
Arctic, especially among its indigenous populations (Arc-
tic Monitoring and Assessment Programme, 2009). Con-
taminants are transported to the Arctic by ocean and air
currents. They are also produced and released by local
activities, such as mining or the disposal of waste. Con-
taminants enter the marine and terrestrial food chains,
posing a risk to the animals as well as to humans who
consume these animals. For PCBs, it is especially the
blubber of marine mammals that contribute to the expo-
sure (Kraemer et al., 2005).
*Correspondence to: Peter Bjerregaard, National Institute of Public
Health, University of Southern Denmark, Øster Farimagsgade 5A, 1353
Copenhagen K, Denmark. E-mail: firstname.lastname@example.org
Received 10 September 2011; Revision received 22 November 2011;
Accepted 22 November 2011
Published online 27 January 2012 in Wiley Online Library (wileyonlinelibrary.
AMERICAN JOURNAL OF HUMAN BIOLOGY 24:165–169 (2012)
C2012 Wiley Periodicals, Inc.
In the circumpolar region, diet is the most important
source of contaminant exposure. Indigenous people are at
particular risk because traditional or ‘‘country’’ foods,
including fish, birds, terrestrial, and marine mammals,
continue to constitute an important component of the diet.
The harvesting, preparation and consumption of such
foods, is an integral part of their culture and important
for social and spiritual health. POPs such as PCBs have
been found in high levels in the blood, cord blood or milk
of pregnant Faroese women (Fangstrom et al., 2002),
pregnant Inuit women in Canada (Butler-Walker et al.,
2003), Greenlanders (Johansen et al., 2004), and nursing
mothers in Arctic and sub-Arctic regions of Russia (Polder
et al., 2003).
There is ample evidence for high concentrations of
POPs in the serum of Arctic residents from certain regions
but only a few studies on time trends. The percentage of
women of reproductive age who exceeded the Health Can-
ada level of concern for PCBs (5 lg/L Aroclor 1260) ranged
from less than 10% in Iceland and several Russian Arctic
regions to more than 70% in Greenland, the eastern Cana-
dian Arctic, and Chukotka (Oostdam et al., 2009). Mean
PCB levels of adult women of the Que ´bec City area were
reported to be only one-tenth of those of Canadian Inuit
(Lebel et al., 1998). Trends of legacy POPs in biota were in
general decreasing during about 1970–2006 (Riget et al.,
2010). Similarly, data for pregnant women suggested that
the serum concentrations of legacy POPs have decreased
from 1990 to 2007 (Oostdam et al., 2009). Despite the high
concentrations of POPs in some Arctic regions, reports of
health effects are few and inconsistent.
We have analyzed secondary sex ratios across the Arctic
in 27 jurisdictions that differ among other things with
regard to their proportion of indigenous people in the pop-
ulation, the type of traditional diet consumed (marine and
terrestrial) and hence exposure to heavily contaminated
marine mammals. If there is an association between POPs
exposure and secondary sex ratios, we would expect a sig-
nificant variation in secondary sex ratios among the juris-
dictions, as would be evidenced by a non-zero slope for the
sex ratio over time.
MATERIAL AND METHODS
Because health and demographic data are collected by
government agencies, we define the Arctic as consisting of
27 political-administrative divisions (Fig. 1), rather than
use geophysical or ecological criteria. These jurisdictions
include the state of Alaska, the three Canadian territories
The Arctic with the 27 jurisdictions studied within eight Arctic countries.
P. BJERREGAARD ET AL.
American Journal of Human Biology
of Yukon, Northwest Territories and Nunavut, the self-
governing territories of Greenland and the Faroe Islands
which are part of the Kingdom of Denmark, Iceland, the
northern regions of Norway, Sweden and Finland, and
various republics, oblasts, and autonomous okrugs (AO)
in the European North and Siberia of the Russian Federa-
tion. All or parts of these regions lie above latitude 608N.
These regions vary considerably in area, population size,
climate, and socioeconomic conditions.
The vital statistics databases of statistical agencies of
the various countries and regions were accessed from
their public websites to extract the number of live births
by sex. A complete list of the data sources is provided in
the Appendix A. The number of years of data available
varies. Data are generally available from the early 1970s,
with the exception of Russian regions, which are available
only from the 1990s. For Alaska and Greenland, the vital
statistics databases permitted us to create additional data
series on Alaska Natives (separate from the Alaska all-
race series) and individuals born to mothers who were
born in Greenland (separate from the series of all births
Graphs of the sex ratio over time were prepared for each
of the 27 jurisdictions. Within each jurisdiction, a linear
regression line was fit to these data. Any deviations of the
slopes of these regression lines from 0 would be evidence of a
change in the sex ratio over time. Time was measured with
the year 2000 as time 0, so that the intercept could be inter-
preted as the estimated sex ratio in that year.
To combine data from the 27 jurisdictions into an overall
estimate, a Bayesian hierarchical model was used. At the
first level of this model, the sex ratio data over time for
each jurisdiction were assumed to follow a linear regres-
sion line, as described above. At the second level of the
hierarchy, the intercepts from the first level were assumed
to follow a normal density, with a mean representing the
overall sex ratio across all regions in the year 2000, and a
variance parameter representing the between region vari-
ability in sex ratios at that time. The slopes across regions
were assumed to follow a separate normal density, with a
mean representing the overall estimate of the change in
sex ratio over time across regions, and a variance parame-
ter estimating the variability in these slopes across
regions. Small variance estimates for the intercepts and
slopes at this second level would indicate little differences
between regions in terms of sex ratio at the year 2000 and
in changes over time, while larger values would indicate
higher between region differences. Ninety-five percent
intervals are presented throughout.
Aside from the general populations of the 27 jurisdictions,
separate analyses were also run for the Alaska Natives and
children born to mothers who were born in Greenland as a
proxy for the indigenous Inuit in Greenland.
Figure 2 shows 5-year moving averages for the five
main regions in the circumpolar area, i.e., Alaska, north-
ern Canada, Greenland, the northern parts of the Nordic
countries, and the northern part of the Russian Federa-
tion. The overall visual impression is that there is no gen-
eral trend over time, as indicated by the flat regression
slopes, and that all averages cluster around a male:female
ratio of 1.05. Separate graphs for individual jurisdictions
are shown in Appendix B [available as web extra materi-
als]. There were no obvious deviations from a flat line
(slope of zero) within any jurisdiction.
This visual conjecture is supported by the estimates of
the slopes from each jurisdiction, as presented in Table 1
that shows the ratio and trend estimates from all 27 Arctic
jurisdictions. Although slopes for Magadan Oblast and Norr-
botten had 95% confidence intervals which excluded 0, the
estimated slopes were small, and it can be expected that in
calculating 27 intervals that one or two may exclude 0 by
chance alone. The sex ratios were all close to 1.05. Only for
the Koryak AO did the 95% confidence interval not span
1.05 but this estimate was based on only 7,200 births. Fur-
thermore, in calculating 27 confidence intervals, it is
expected that one falls outside 1.05 by chance alone.
To check this, we examined estimates from the hier-
archical model. Here all 27 of the intercepts were esti-
mated to be close to 1.05, with overall point estimate from
the hierarchical model of 1.054 (95% CI 1.048, 1.058). All
the slope estimates included 0 within their range, includ-
ing those for Magadan Oblast and Norrbotten. The overall
estimate for the 10-year slope across the 27 jurisdictions
from this model was 0.0010 (95% CI 20.0021, 0.0046).
Results for Alaskan and Greenland native popula-
tions were similar, with sex ratios close to 1.05 and
slopes close to 0.
Our data, which cover the entire Arctic and extend over
several decades, fail to confirm media claim or the results
of studies on single communities that are limited in time
duration, of decreased sex ratios of live births in the Arctic
or significant time trends. For a region such as Greenland,
the population which is among the most exposed to PCBs,
the secondary sex ratio today (1.043) is no different than
that in the early 20th century (1.048) as reported by Ber-
The lack of change in sex ratios among 27 Arctic juris-
dictions over the period studied provides evidence that
the contaminants that are present are not disrupting
in five Arctic regions, 1997–2007.
Five-year moving averages of male:female sex ratios at birth
SEX RATIOS IN THE ARCTIC
American Journal of Human Biology
endocrine systems to the extent that sex ratios are being
affected, or widespread enough to be detected at the popu-
lation level. By analyzing separately data on Alaska
Natives and indigenous Greenlanders, populations which
have been described to have higher levels of contaminant
exposure (Arctic Monitoring and Assessment Programme,
2001), we were still unable to demonstrate any deviation
from the global human norm or change in the secondary
sex ratios over time.
Our finding is at odds with national studies in several
circumpolar countries which have demonstrated a down-
ward trend in secondary sex ratio (Allan et al., 1997;
Irgens and Irgens, 2003; Vartianinen et al., 1999), for
which the global increase in the use of POPs has been
cited as a potential cause. The existing evidence for a
causative role of POPs and other environmental contami-
nants in changing the sex ratio is inconsistent, but the
association likely varies based on the type and amount of
exposure, the timing during pregnancy and how the expo-
sure is measured. Legacy POPs show a decreasing trend
in the Arctic due to legislation banning these widely meas-
ured POPs. There are, however, newer POPs that are not
studied to the same extent as the legacy POPs (Oostdam
et al., 2009) and for which, therefore, we do not have infor-
mation about time trends.
We recognize that our study is not designed to test the
association between environmental contaminants as the
exposure and secondary sex ratio as the outcome. It is,
however, logical to first determine if an outcome exists at
all before finding out what causes it.
While we cannot conclude that PCBs have no effect on
the sex ratio at birth in the northern jurisdictions that we
have analyzed, we can conclude that the regions studied
do not differ from each other with regard to sex ratio at
birth, although they differ with regard to contaminant ex-
posure. We can also conclude that there is no trend over
time toward a decreasing sex ratio in any of the regions
studied or in all of them together.
Many factors have been studied for their effects on the
secondary sex ratio—one review listed 30 demographic
and environmental factors, including family size, parental
age, parental occupation, birth order, race, coital rate, hor-
monal treatments, exposure to toxins, stress, and several
diseases (Hesketh and Xing, 2006). Because of the nature
of our data, we were not able to include these variables,
but further studies including this information may help
understand the stability of the sex ratio in the Arctic.
From a public health monitoring perspective the sex ra-
tio is not of primary interest. Given the fact that the sec-
ondary sex ratio has not changed significantly over the
past several decades when the Arctic has experienced sub-
stantial social, climatic, and environmental change, the
ratio is unlikely to be a useful summary indicator to moni-
tor future changes in these areas. Instead, these changes
are best monitored directly at the environmental level.
Public health responses to social, climatic, and environ-
mental change and the monitoring of human health
impacts are more complex and require multidisciplinary
Allan BB, Brant R, Seidel JE, Jarrell JF. 1997. Declining sex ratios in Can-
ada. Can Med Assoc J 156:37–41.
Arctic Monitoring and Assessment Programme. 2004. Persistent toxic sub-
stances, food security and indigenous peoples of the Russian North. Final
Report. Oslo: Arctic Monitoring and Assessment Programme;175 p.
TABLE 1. Male:female sex ratio and 10 year increase in ratios in 27 jurisdictions across the Arctic
Male:female ratio Increase per 10 year period
EstimateLower CL Upper CL EstimateLower CLUpper CL
All regions combined
Estimates with 95% confidence intervals.
P. BJERREGAARD ET AL.
American Journal of Human Biology
Arctic Monitoring and Assessment Programme. 2009. AMAP Assessment
2009: Human Health in the Arctic. Oslo: AMAP. 254 p.
Bertelsen A. 1935. Grønlandsk medicinsk statistic og nosografi: Undersø-
gelser og erfaringer fra 30 aars grønlandsk lægevirksomhed. Bd. I:
Grønlands befolkningsstatistik 1901–30. [Greenland population statis-
tics 1901–30. In Danish]. Meddelelser om Grønland 117:1–83.
Brown P. 2007. Man-made chemicals blamed as many more girls than boys
are born in Arctic. Guardian 12 September 2007. Available at: www.
guardian.co.uk/world/2007/sep/12/gender.sciencenews. Accessed June 6,
Butler Walker J, Seddon L, McMullen E, Houseman J, Tofflemire K, Corri-
veau A, Weber JP, Mills C, Smith S, Van Oostdam J. 2003. Organochlor-
ine levels in maternal and umbilical cord blood plasma in Arctic Canada.
Sci Total Environ 302:27–52.
Fangstrom B, Athanasiadou M, Grandjean P, Weihe P, Bergman A. 2002.
Hydroxylated PCB metabolites and PCBs in serum from pregnant Faro-
ese women. Environ Health Perspectives 110:895–899.
Grech V, Vassallo-Agius P, Savona-Ventura C. 2003. Secular trends in sex
ratios at birth in North America and Europe over the second half of the
20th century. J Epidemiol Commun Health 57:612–615.
Hertz-Picciotto I, Jusko TA, Willman EJ, Baker RJ, Keller JA, Teplin SW,
Charles MJ. 2008. A cohort study of in utero polychlorinated biphenyl
(PCB) exposures in relation to secondary sex ratio. Environ Health 7:37.
Hesketh T, Xing ZW. 2006. Abnormal sex ratios in human populations:
causes and consequences. Proc Nat Acad Sci USA 103:13271–13275.
Irgens A, Irgens LM. 2003. Secular trends in uncertain-sex births and pro-
portion of male births in Norway, 1967–1998. Arch Environ Health
James WH. 2006. Offspring sex ratios at birth as markers of paternal en-
docrine disruption. Environ Res 100:77–85.
Johansen P, Muir D, Asmund G, Riget F. 2004. Human exposure to con-
taminants in the traditional Greenland diet. Sci Total Environ
Karmaus W, Huang S, Cameron L. 2002. Parental concentration of
dichlorodiphenyl dichloroethene and polychlorinated biphenyls in
Michigan fish eaters and sex ratio in offspring. J Occupat Environ
Kraemer LD, Berner JE, Furgal CM. 2005. The potential impact of climate
on human exposure to contaminants in the Arctic. Int J Circumpolar
Lebel G, Dodin S, Ayotte P, Marcoux S, Dewailly E´. 1998. Organochlorine
exposure and the risk of endometriosis. Fertil Steril 69:221–228.
Mackenzie CA, Lockridge A, Keith M. 2005. Declining sex ratio in a First
Nation community. Environ Health Perspect 113:1295–1298.
Mocarelli P, Brambilla P, Gerthoux PM, Patterson DG, Needham LL. 1996.
Change in sex ratio with exposure to dioxin. Lancet 348:409.
Oostdam JV, Donaldson SG. 2009. Human tissue levels of environmental
contaminants. In: AMAP. AMAP Assessment 2009: Human health in the
Arctic, Chapter 5. Oslo, Arctic Monitoring and Assessment Programme.
Polder A, Odland JO, Tkachev A, Føreid S, Savinova TN, Skaare JU. 2003.
Geographic variation of chlorinated pesticides, toxaphenes and PCBs in
human milk from sub-Arctic and Arctic locations in Russia. Sci Total En-
Riget F, Bignert A, Braune B, Stow J, Wilson S. 2010. Temporal trends
of legacy POPs in Arctic biota, an update. Sci Total Environ 408:
Taylor KC, Jackson LW, Lynch CD, Kostyniak PJ, Buck Louis GM. 2006.
Preconception maternal polychlorinated biphenyl concentrations and
the secondary sex ratio. Environ Res 103:99–105.
Terrell ML, Berzen AK, Small CM, Cameron LL, Wirth JJ, Marcus M.
2009. A cohort study of the association between secondary sex ratio and
parental exposure to polybrominated biphenyl (PBB) and polychlori-
nated biphenyl (PCB). Environ Health 8:35.
Vartianinen T, Kartovaara L, Tuomisto J. 1999. Environmental chemicals
and changes in sex ratio: analysis over 250 years in Finland. Environ
Health Perspect 107:813–815.
Yoshimura T, Kaneko S, Hayabuchi H. 2001. Sex ratio in offspring of those
affected by dioxin and dioxin-like compounds: the Yusho, Seveso, and
Yucheng incidents. Occupat Environ Med 58:540–541.
Young TK, Bjerregaard P, editors. 2008. Health transitions in Arctic popu-
lations. Toronto: University of Toronto Press. 485 p.
APPENDIX A: DATA SOURCES
Data was collected from vital statistics repositories,
most of which were accessible online.
National Centre for Health Statistics: VitalStats data
files (http://www.cdc.gov/nchs/VitalStats.htm) from 1994
to 2006, and earlier data from past volumes of the Vital
Statistics of the United States annual reports (http://
Statistics Canada: 2000–2006 data available online at
obtained by special request.
Statistics Greenland’s Statbank for the years 1973–
2009 at bank.stat.gl.
Statistics Faroe Islands for the years 1985–2008 at
Statistics Iceland for the years 1951–2008 at www.
Statistics Norway’s Statbank for the years 1972–2008,
Statistics Sweden Database for the years 1968–2009, at
Statistics Finland data is available at pxweb2.stat.fi/
Central Statistical Database available at: www.gks.ru/
All websites were accessed in February 2011.
Graphs of male:female sex ratios over time in individual
Arctic jurisdictions. Point estimates and 95% confidence
SEX RATIOS IN THE ARCTIC
American Journal of Human Biology