Human Reproduction, Vol.24, No.12 pp. 3196 – 3204, 2009
Advanced Access publication on October 3, 2009
ORIGINAL ARTICLE Reproductive epidemiology
Physical activity and fertility in women:the North-Trøndelag Health Study
S.L. Gudmundsdottir1, W.D. Flanders2, and L.B. Augestad1,3
1Human Movement Science Programme, Faculty of Social Sciences and Technology Management, Norwegian University of Science andTechnology (NTNU), NO-7491 Trondheim, Norway 2Departments of Epidemiology and Biostatistics, Emory University, Atlanta,GA 30345, USA
3Correspondence address. Tel: þ47 7359-1780; Fax: þ47 7359-1770; E-mail: firstname.lastname@example.org
background: Changes in the state of energy balance owing to changes in physical activity may affect the reproductive system. Weevaluated the association between physical activity (PA) and fertility and parity in healthy women.
methods: A population-based health survey (HUNT 1) was conducted during 1984 – 1986 in Nord-Trøndelag county, Norway, withfollow-up from 1995 to 1997 (HUNT 2). The study included 3887 women, ,45 years old in HUNT 2. PA was assessed by baseline ques-
tionnaire, and fertility and parity by questionnaire at follow-up. Data focused on overall occurrence of infertility in the population (withoutbiological conﬁrmation).
results: Increased frequency, duration and intensity of PA were associated with increased subfertility, and frequency of PA was associ-ated with voluntary childlessness (P , 0.01). After adjusting for age, parity, smoking, and marital status, women who were active on mostdays were 3.2 times more likely to have fertility problems than inactive women. Exercising to exhaustion was associated with 2.3 times theodds of fertility problems versus low intensity. Women with highest intensity of PA at baseline had the lowest frequency of continuing nulli-parity and highest frequency of having three or more children during follow-up (P , 0.05). Sensitivity analysis including body mass index asconfounder did not alter the results. No associations were found between lower activity levels and fertility or parity.
conclusion: Increased risk of infertility was only found for the small group of women reporting the highest levels of intensity and fre-quency of PA. Awareness of the possible risks of infertility should be highlighted among non-athletic women who exercise vigorously.
Key words: epidemiology / exercise / infertility / physical activity / women
Kravdal, 2008). It is suggested that one-third of the incidences ofcouple infertility can be related to male reproductive problems
Infertility is deﬁned as the inability of a couple to conceive within
1 year of trying to become pregnant (Kumar et al., 2007). Recent
Reproductive dysfunction is reported to have a higher prevalence in
estimates of the prevalence of infertility range from 3.5 to 16.7% in
athletes than non-athletes (Russell et al., 1984; Otis et al., 1997), with
more developed nations and from 6.9 to 9.3% in less developed
clinical consequences that may include infertility (Warren et al.,
nations (Boivin et al., 2007). In Norway, the lifetime infertility preva-
2001). Less is known about the effects of physical activity (PA) on
lence has been estimated to be 6.6% (Rostad et al., 2006).
female fertility in the general population, in contrast to the relatively
A variety of lifestyle factors, such as smoking (Hughes and Brennan
well-known beneﬁcial effects of regular PA for numerous undesirable
1996; Augood et al., 1998; Hull et al., 2000), psychological stress
health outcomes, including the prevention of premature death
(Hjollund et al., 1999), caffeine (Wilcox et al., 1988) and alcohol con-
(Bouchard and Blair 1999; Warburton et al., 2006). Only a few
sumption (Gill, 2000; Eggert et al., 2004) and extremely low or high
studies have focused on the general population. In a cohort study, the
body mass index (BMI) (Rich-Edwards et al., 2002; Hassan and
Nurses’ Health Study II, more hours of vigorous exercise were associ-
Killick 2004; Norman et al., 2004) have been proposed as causes of
ated with reduced risk of ovulatory infertility (Rich-Edwards et al., 2002;
infertility. Consequences of infertility may include psychosocial pro-
Chavarro et al., 2007) while Morris et al. (2006) found that women who
blems (Greil, 1997; Boivin, 2003), higher risks of breast and ovarian
had enrolled on a fertility treatment programme and reported exercis-
cancer (Jensen et al., 2008) and high ﬁnancial costs for those trying
ing 4 h or more per week for 1 – 9 years previously were 40% less likely
infertility treatment (Jain and Hornstein 2003). In addition, childless-
to have a live birth, almost three times more likely to experience cycle
ness has been related to mortality disadvantages (Grundy and
cancellation, and twice as likely to have an implantation failure or
& The Author 2009. Published by Oxford University Press on behalf of the European Society of Human Reproduction and Embryology. All rights reserved.
For Permissions, please email: email@example.com
pregnancy loss than women who had reported not exercising. Moremoderate exercise together with weight loss has been found to be posi-tive in fertility treatment in obese women (Clark et al., 1995). The effectof PA on fertility may therefore be positive up to a certain level and havea negative effect above that threshold level of activity. A similar pattern issuggested for bone health: physically active individuals have been foundto have a lower risk of low bone mineral density and fractures, while therisk of stress fractures is increased with high-volume training (Kohrtet al., 2004). This may occur through the direct effects of activity on hor-mones which stimulate bone formation or indirectly via effects on estro-gen and menstrual function, a condition often coupled with negativeenergy availability in an interrelationship known as the female athletetriad (Nattiv et al., 2007).
PA plays an important role in maintaining energy balance and it has
been suggested that changes in the status of energy balance due tochanges in PA may affect the reproductive system (Redman, 2006).
In addition, there are some indications that weight loss via PA mayprotect ovarian function by increasing insulin resistance and changingthe hormone proﬁle (Norman et al., 2004).
The purpose of this study is to evaluate the association between PA
and fertility status and parity in healthy, premenopausal, Norwegian
All residents in the county of Nord-Trøndelag, Norway, men and womenaged 20 years, were invited to participate in the Nord-Trøndelag Health
Figure 1 Sequential exclusion of participants in the study of phys-
Study in 1984 – 1986 (HUNT 1) and in the follow-up study in 1995 – 1997
ical activity and fertility. *No response to questions needed for
Among those eligible for HUNT 1, 77 310 (90.8%) returned a
health-related questionnaire and 74 977 (88.1%) participated in thehealth examination. At the examination, participants ﬁlled out a secondquestionnaire, including questions regarding PA.
Fertility status and reproductive history were assessed by a self-
In HUNT 2, 71.3% of the original cohort from HUNT 1 participated. Par-
reported questionnaire in HUNT 2. Questions included inability to con-
ticipants were sent a comprehensive health-related questionnaire by mail.
ceive within 1 year of trying (and at what age), parity, age at childbirth(s),
Assessments at the screening station included measurements of height and
contraceptive use and menstruation and pregnancy status.
weight. At the examination, participants received a more detailed question-
We classiﬁed women as fertile if they conceived within 1 year of
naire evaluating demographics, medication use, diet, reproductive history,
attempting to become pregnant and gave birth, and infertile if they did
infertility, menstrual history, history of gynaecological surgery and pregnancy.
not conceive within 1 year, regardless of any subsequent pregnancies
In total, 24 837 women participated in both surveys. Of those, 5986
(Kumar et al., 2007). Infertile women were further classiﬁed as
women were ,45 years old at HUNT 2, and therefore considered pre-
subfertile if the time to pregnancy exceeded 1 year of trying, and as
menopausal. We excluded women who reported serious conditions or
involuntarily childless if they reported problems with conceiving within
diagnoses known to affect capability for PA or reproductive function at
1 year and there had been no childbirth. Voluntarily childlessness was
baseline, as summarized in Fig. 1, leaving 3887 women in our study.
assumed if neither problems with becoming pregnant nor childbirth
Weekly frequency, intensity and duration of leisure time PA was
assessed using a self-reported questionnaire in HUNT 1 (http://www.
Fertility status and parity for different levels of baseline variables were
compared using descriptive and chi-square statistics. We modelled the
questionnaires/) (Table I). The questions regarding PA have been vali-
probability of fertility problems (subfertility or involuntary childlessness)
dated for men with good reliability and test and retest validity (Kurtze
using unadjusted and adjusted logistic regression. For infertility, odds
et al., 2008). Based on the reported intensity, duration and frequency,
ratio (ORs) estimates were obtained by maximum likelihood with associ-
an index of PA was calculated: detailed descriptions are presented in pre-
ated 95% conﬁdence intervals (CIs). We were concerned there might be
viously published papers (Augestad et al., 2004; Augestad et al., 2006).
some link between PA and age and therefore performed subgroup ana-
With division at the 33.3th and 66.6th percentiles of the index, leisure
lyses for young women, ,30 years of age, because of the higher preva-
time PA was categorized into low, moderate and high levels. Occupational
PA was assessed by asking the women if they felt physically tired
The associations of baseline measurements with parity were examined
from occupational work (almost never, seldom, often, almost always).
in women who were nulliparous in HUNT 1. Comparisons were made
Age, education, marital status, smoking, and alcohol consumption were
using ordinal regression models, adjusting for potential confounders.
reported at baseline. BMI (kg/m2) was computed from measured height
In these analyses parity was treated as a three-level variable: nulliparous,
one to two children and more than two children.
Table I Baseline characteristics of the healthy Norwegian female study population
PA, physical activity. Fertility status based on self-reported problems with conceiving within 1 year and childbirths at follow-up. P-values indicate results of the Pearson’s chi-square test. n (%).
In all of the above analyses, models were run separately for each of the
subfertility and voluntary and involuntary childlessness were approxi-
PA variables (frequency, intensity, duration and index). Full multivariate
mately J-shaped, and subfertility was least common in the normal
models were run at ﬁrst. Variables were subsequently removed from
weight group (BMI: 18.5 – 24.9). Based on the univariate analyses,
the models if their removal did not affect the coefﬁcient estimates of
women classiﬁed as fertile were more often married and had less
the remaining variables in the model or the overall signiﬁcance of the
than 10 years of education. Voluntarily childless women were more
often unmarried, non-smokers, with more than 12 years of education.
Based primarily on a priori considerations, we considered the following
Subfertility was least common among married women (Table I).
baseline measurements as potential confounders: age (5-year categories),parity, i.e. when assessing the probability of infertility (nulliparous, not nul-
Increased frequency of PA was associated with increased subfertility
liparous) (Sternfeld et al., 1999; Miller et al., 2002), smoking (never,
and voluntary childlessness. A similar relationship was observed
former, current), frequency of alcohol consumption during the 14 days
between intensity and longer duration of PA and subfertility but less
prior to participation in the study (never, 1 – 4 times, 5 – 10 times, .10
so with voluntary childlessness. In the unadjusted analyses, a high
times) marital status (unmarried, married, previously married) and edu-
level of PA, according to our calculated index, was associated with
increased subfertility and voluntary childlessness. Due to missing
Women with an unusually high or low BMI tend to have greater risk of
answers however, the PA index could not be calculated for just
infertility. Because women with high or low BMI may be either less likely or
over 1000 women. Tiredness from occupational activity was not
more likely to be physically active, baseline BMI may be considered a
found to be associated with fertility status (Table I).
potential confounder of the relationship between PA and fertility.
The observed relationships between PA variables and fertility status
However, BMI might also be considered an intermediate variable in that
followed the same trend as those reported above, after controlling for
PA could affect BMI and, in turn, fertility status. We therefore did notinclude BMI as an independent variable in the primary (a priori) analyses,
suspected confounding factors: age, parity, smoking and marital status
because its inclusion might have biased the estimated association
(Table II). Additional control for education and alcohol consumption
between PA and fertility status. However, to assess whether inclusion of
did not meaningfully affect the ORs. In particular, women who were
BMI affected the estimated association with fertility status, we also con-
active on most days of the week were 3.2 times more likely to be
ducted additional (sensitivity) analyses that included this variable, treating
infertile than inactive women. Exercising to exhaustion was associated
it as categorical (World Health Organization categorization of under-,
with 2.3 times the odds of infertility compared with lower levels of
normal- and overweight, obesity and severe obesity) at baseline. Statistical
intensity. For duration, there was decreased risk of infertility in
those whose exercise was moderate (16 – 30 and 30 – 60 min) com-
The study was approved by the Norwegian Regional Committee for
pared with the shortest duration of less than 15 min. Women who,
Ethics in Medical Research and the Norwegian Data Inspectorate and
according to the index of PA, were highly active had an OR of 1.5
each subject gave written informed consent prior to participation.
for infertility compared women with low or moderate levels of activity.
The alternative analysis, designed to assess the sensitivity of results
to alternative models, in which we included BMI as a potential con-founding factor, did not result in notably different parameter esti-
Baseline characteristics and their bivariate relations to fertility status at
mates, consistent with the small differences in BMI across the
follow-up are shown in Table I. The mean baseline age was 27.2 years
(range 20 – 35) and mean BMI was 22.7 kg/m2 (range 14.5 – 44.1).
In a subgroup analysis restricted to women up to 30 years of age in
Overall, 90.1% of the women were classiﬁed as fertile, 5.1% as subfer-
whom subfertility was more common (6.3%) than in the older age
tile, 0.7% as involuntary childless and 3.9% as voluntary childless.
group (1.5%), we found that these younger women were more
A total of 62.4% of infertile women reported having visited a
likely than older women to exercise every day, exercise to exhaustion
medical doctor for fertility problems.
or exercise for more than 60 min. Among the young women, subfer-
The frequency of subfertility decreased with increased age, as did
tility was reported by 23.7% of those exercising to the level of exhaus-
voluntary childlessness. The relationships of BMI categories with
tion and by 11.1% of those who reported exercising almost every day.
analyses of this group, continued nulliparity at follow-up was associ-
Table II Unadjusted and adjusted ORs from logistic
ated with greater age and with lower or higher than normal BMI.
regression for fertility problems (subfertility or
Parity during follow-up was also related to education. However, it
should be noted that at baseline, 93.7% of women with ,10 yearseducation had already given birth (Table I). For women who had
not given birth at baseline, we found similar ORs of infertility as for
the whole sample but the 95% CIs were larger and included 1 in all
Women who reported the highest intensity of PA at baseline had
the lowest frequency of continuing nulliparity and highest frequency
of having three or more children during follow-up (P , 0.05). In the
unadjusted analysis, there was also a tendency towards less continuing
nulliparity and higher parity among those women with longer durationsof PA and higher scores on the PA index (Table III).
There was a signiﬁcant tendency towards higher parity among
women with the highest intensity of PA compared with moderate
intensity in an ordinal regression analysis, adjusting for age and edu-
cation. Women reporting low intensity of activity did not signiﬁcantly
differ from the high intensity group. The model was subsequently run
including BMI but no meaningful changes to the parameter estimates
were observed. In another subgroup analysis excluding those
women who were voluntarily childless in HUNT 2, although the direc-
tion of this association was the same, the relationship was no longer
Frequency of PA or tiredness from occupational activity was not
#Adjusted for age, parity, smoking and marital status.
In this study we examined the association between PA and fertility
*Reference category OR, odds ratio; CI, conﬁdence interval.
status and parity in healthy premenopausal Norwegian women. Wefound that women with the highest levels of frequency or intensity
In this subgroup there was a u-shaped relationship between duration
of PA had increased risk of infertility. We did not, however, see a
of exercise and subfertility. Among women who reported ,15 min or
trend for increased risk of infertility for the submaximal levels of inten-
.60 min duration, the frequency of subfertility was 12.6 and 12.4%,
sity or frequency of PA. Moreover, the highest level of our index of
respectively, while 3.9% of women exercising for 16 – 30 min and
PA, indicating the total load of exercise, was related to increased like-
7.0% of those exercising for 30 – 60 min reported subfertility.
lihood of infertility in adjusted analyses. There was decreased risk of
In this subgroup, adjusted logistic analysis showed similar but some-
infertility for women reporting moderate duration of activity in our
what stronger relationships than in the study group as a whole. The
adjusted analyses, compared with women reporting activity duration
OR for infertility was 3.5 (CI: 1.3 – 9.0) for women who exercised
of less than 15 min. Our estimate of 5.8% prevalence of infertility is
almost every day and women who exercised to the level of exhaustion
somewhat lower than earlier reports from the same study population
had 3.0 (CI: 1.3 – 6.0) times the risk of infertility compared with those
in which lifetime infertility prevalence has been estimated to be 6.6%
taking it easy. The trend between the calculated index of PA, occu-
(Rostad et al., 2006). The women in our sample were younger than in
pational activity or other variables did not differ meaningfully from
the Rostad et al. (2006) study and some may still experience infertility
patterns observed in the analyses that included all age groups.
in their remaining reproductive years.
Women who were nulliparous at baseline were signiﬁcantly younger
Although we showed that high intensity of PA was related to
than women who had given birth (24.0 versus 26.5; P , 0.001). They
increased risk of infertility, there was also a trend of having three or
had higher occurrence of frequent PA, higher intensity and longer dur-
more children during the follow-up period among women who
ation of activity (P , 0.001 for all). Nulliparous women were less likely
reported the highest level of PA intensity at baseline. A possible expla-
to feel tired from occupational activity. They had higher education at
nation may be that women who did high intensity training at baseline
baseline, and were more likely to be non-smokers and unmarried
and experienced infertility problems adjusted their activity level to a
than women who had given birth (P , 0.001 for all). The two
lower intensity level. Alternatively, although if intense PA had a nega-
groups did not differ on baseline BMI.
tive effect on the ﬁrst trial for conception, hormonal changes during
For women who were nulliparous at baseline, the mean number of
and following pregnancy may have counteracted that negative effect
children born between baseline and follow-up investigations was 1.5
in later pregnancies. The likelihood of fertility problems in our study
(range 0 – 4). A total of 179 women were still nulliparous at follow-up,
was, however, not notably different for nullipara women compared
of whom 152 (84.9%) reported voluntary childlessness. In descriptive
with those who had already given birth. This may suggest that the
Table III Baseline characteristics of women who were nulliparous at baseline (n 5 786)
PA, physical activity. Childbirths during follow-up reported in questionnaire. P-values indicate results of the Pearson’s chi-square test. n(%).
possible detrimental effects of high intensity exercise are reversible
did not try to identify different causes of infertility with biological
and not permanent. Unfortunately, our data did not allow further con-
testing but focused on the overall occurrence in the population and
this may have biased our results. Green et al. (1986) found that
There have only been a few studies regarding the effect of PA on fer-
among infertile women who had never conceived, high intensity exer-
tility status and parity in the general female population. Our results are
cise for more than 60 min/day was associated with even higher risk
somewhat comparable to results of a retrospective case – control study
of infertility after excluding cases with evidence of tubal dysfunction.
which found a 6.2-fold increased risk of infertility in women who exer-
Differences in measurement methods for infertility may contribute to
cised vigorously (deﬁned as aerobic activities with estimated energy
differences in results, to some extent. Also, not all women who partici-
requirements of 6 kcal/min) for at least 60 min per day (Green et al.,
pated in the baseline survey did so in the follow-up survey. If non-
1986). There are also reports of lutheal phase alterations, which may
participation at follow-up was jointly associated with PA and infertility,
lead to infertility, among recreational runners with at least 2 h per
bias may have resulted. Many women did not answer questions regard-
week of exercise and running 32 km/week (De Souza et al., 2003).
ing fertility status and we cannot exclude the possibility that those
The recreational runners in the De Souze study had lower BMI and
women differed from women who did answer; this could have affected
fat mass than sedentary controls and the results were related to short-
our results. Our analysis found, however, that women who did not
term negative energy balance and hypometabolic state, similar to but
answer questions regarding fertility status in the follow-up study had
not as extensive as seen in amenorrheaic athletes.
somewhat lower intensity and duration of PA at baseline than women
Results from the Nurse’s Health Study indicate reduced risk of ovu-
who did answer those questions, although this difference was not stat-
latory infertility in women doing vigorous exercise for at least 30 min
istically signiﬁcant. Because data on fertility status were collected retro-
daily (Chavarro et al., 2007). An earlier report of the same cohort indi-
spectively, recall bias may have occurred. Nevertheless, since fertility
cated a 7% relative risk reduction for ovulatory infertility for each
status is of high personal interest, precise recollection of its aspects
additional hour of vigorous exercise per week (Rich-Edwards et al.,
may be likely. Other sources of bias may have been spousal infertility
2002). Thus, there is a considerable divergence in study results. The
and change of partner, which we could not control for in our data.
mean age in our study group was approximately 5 years lower than
PA may affect fertility merely temporary or more permanently in
in the women in the Nurse’s Health Study and this may explain the
the reproductive life, but our study did not obtain information at
difference in results to some extent, as we found younger women
the time when women tried to conceive. Although we excluded
more likely to report both highest intensity of PA and infertility than
women who reported infertility at baseline we cannot be certain
whether the level of PA reported at baseline was maintained until
PA can improve body composition, glucose homeostasis and insulin
sensitivity (Warburton et al., 2006). Exercise-induced weight loss has
In epidemiological studies, the intensity of PA can be assessed as
been shown to improve metabolic function and hormonal proﬁles, and
absolute or relative intensity. Metabolic equivalent task is frequently
often leads to signiﬁcant increase in fertility (Clark et al., 1995;
used to refer to absolute intensity indicating the energy cost of an
Norman et al., 2004). In our study, adjustment for BMI did not signiﬁ-
activity. PA-related energy expenditure has been associated with mor-
cantly alter the increased risk of infertility associated with the highest
bidity and mortality (Helmrich et al., 1994; Manini et al., 2006; Mora
intensity or frequency of activity. Thus, the effect of PA may largely act
et al., 2007). Besides being an important determinant of health, data
through a mechanism that does not involve BMI to a signiﬁcant extent.
on energy expenditure also facilitate study result comparisons and
It is possible that in normal weight women, high loads of exercise,
interpretation of ﬁndings into meaningful units, although misreporting
not coupled with increased energy intake, lead to negative energy
errors have been observed (Mahabir et al., 2006). Relative intensity
balance where the energy requirements of reproductive functions
can be described by use of the Borg scale (Borg, 1982) or other
cannot be met (Loucks et al., 1998). This may explain our ﬁnding of
measures of how intense the exercise is perceived by each subject.
increased infertility risk among women doing high intensity training
In the current study, information regarding PA was based on a self-
reported questionnaire that allowed for subjective interpretation of
Our study is population based, not depending on cases seeking help
PA variables. It may therefore suffer from misclassiﬁcation of activity
from a clinic, and can be viewed as a cohort study. The prospective
by factors such as age, social situations and seasonal variation
assessment of the effect of baseline PA on subsequent fertility problems
(Vanhees et al., 2005). Including a large number of study subjects
hopefully contributes to avoiding selection bias of cases and reporting
and classifying subjects into wide categories, e.g. low, moderate,
bias of activity levels. However, there were possible limitations. We
misclassiﬁcation (Shephard, 2003). The questions regarding intensity
Trøndelag County Council and The Norwegian Institute of Public
and duration of PA in the current study have previously been validated
in a small sample of women and men over a wide age range (Wisloffet al., 2006). The results indicate a positive association betweenreported intensity and the measures of oxygen uptake during exercise.
Taking it easy corresponded to an intensity of approximately 55% of
Augestad LB, Schei B, Forsmo S, Langhammer A, Flanders WD. The
the maximal oxygen uptake, independent of activity duration up to
association between physical activity and forearm bone mineral
90 min. Intensity descriptions of loosing breath and exercising to the
density in healthy premenopausal women. J Womens Health 2004;
level of exhaustion corresponded to 71 – 75 and 76 – 90%, respect-
ively, of the maximum oxygen uptake. Relative intensity decreased
Augestad LB, Schei B, Forsmo S, Langhammer A, Flanders WD. Healthy
when duration of exercise was increased. A study of reliability and val-
postmenopausal women—physical activity and forearm bone mineral
idity of the questions including only men found good reliability and
density: the Nord-Trondelag health survey. J Women Aging 2006;
kappa statistics of 0.69 – 0.82 between test and retest (Kurtze et al.,
2008), although these results are not necessarily valid for women. It
Augood C, Duckitt K, Templeton AA. Smoking and female infertility: a
seems likely that any measurement error in the assessment of PA
systematic review and meta-analysis. Hum Reprod 1998;13:1532 – 1539.
should not depend on fertility status and would tend to bias results
Boivin J. A review of psychosocial interventions in infertility. Soc Sci Med
toward the null. In our study, there were many missing responses
Boivin J, Bunting L, Collins JA, Nygren KG. International estimates of
on both PA and fertility. For instance, our PA index could not be cal-
infertility prevalence and treatment-seeking: potential need and
culated for just over 1000 women in our study sample. In summary,
demand for infertility medical care. Hum Reprod 2007;22:1506 – 1512.
we found that women who were active on most days of the week
Borg GA. Psychophysical bases of perceived exertion. Med Sci Sports Exerc
had a threefold risk of infertility compared with inactive women.
The increased risk of infertility was only found for the most extreme
Bouchard C, Blair SN. Introductory comments for the consensus on
intensity and frequency of PA, perhaps representing women who
physical activity and obesity. Med Sci Sports Exerc 1999;31:S498 – S501.
exercised at levels comparable to elite athletes. While we did not
Chavarro JE, Rich-Edwards JW, Rosner BA, Willett WC. Diet and lifestyle
ﬁnd any associations between lower levels of activity and fertility
in the prevention of ovulatory disorder infertility. Obstet Gynecol 2007;
status or parity, other studies have found regular moderate PA to
improve reproductive function. It could therefore be suggested that
Clark AM, Ledger W, Galletly C, Tomlinson L, Blaney F, Wang X,
moderate PA has little or perhaps even a positive direct effect on fer-
Norman RJ. Weight loss results in signiﬁcant improvement inpregnancy and ovulation rates in anovulatory obese women. Hum
tility. This should undoubtedly be conﬁrmed, as it may possibly affect
fertility by other means than biological, for example via positive effects
De Souza MJ, Van Heest J, Demers LM, Lasley BL. Luteal phase deﬁciency
on self-esteem and improved mental health. Psychosocial interven-
in recreational runners: evidence for a hypometabolic state. J Clin
tions, particularly those emphasizing skills training, such as relaxation
Endocrinol Metab 2003;88:337 – 346.
training, have been found effective in infertility treatment (Boivin
Eggert J, Theobald H, Engfeldt P. Effects of alcohol consumption on female
2003) and the positive psychological effects of regular PA on mental
fertility during an 18-year period. Fertil Steril 2004;81:379 – 383.
well-being have been reported in numerous studies (Scully et al.,
Gill J. The effects of moderate alchohol consumption on female hormone
1998; Kull 2002). Results on the association between high levels of
levels and reproductive function. Alcohol Alcohol 2000;35:417 – 423.
PA and fertility status are conﬂicting. We found increased risk of infer-
Green BB, Daling JR, Weiss NS, Liff JM, Koepsell T. Exercise as a risk factor
tility in women who reported doing PA almost every day and in those
for infertility with ovulatory dysfunction. Am J Public Health 1986;
reporting exercising to the level of exhaustion. Although it cannot be
Greil AL. Infertility and psychological distress: a critical review of the
concluded from this study, it seems likely that fertility may be positively
literature. Soc Sci Med 1997;45:1679 – 1704.
affected by PA up to a certain level after which the energy require-
Grundy E, Kravdal O. Reproductive history and mortality in late middle age
ments of the activity outweigh the positive effects, resulting in inferti-
lity. It is also possible that this threshold level applies merely at an
individual level, perhaps in combination with energy availability, and
Hassan MA, Killick SR. Negative lifestyle is associated with a signiﬁcant
cannot be identiﬁed at the population level.
reduction in fecundity. Fertil Steril 2004;81:384 – 392.
The potential role of regular PA in the prevention and treatment of
infertility needs further investigation. However, awareness of the poss-
non-insulin-dependent diabetes mellitus with physical activity. Med Sci
ible risks of infertility should be highlighted among non-athletic women
Hjollund NH, Jensen TK, Bonde JP, Henriksen TB, Andersson AM,
Kolstad HA, Ernst E, Giwercman A, Skakkebaek NE, Olsen J et al.
Distress and reduced fertility: a follow-up study of ﬁrst-pregnancy
planners. Fertil Steril 1999;72:47 – 53.
Hughes EG, Brennan BG. Does cigarette smoking impair natural or
The faculty of Social Sciences and Technology Management, The
assisted fecundity? Fertil Steril 1996;66:679 – 689.
Norwegian University of Science and Technology (NTNU) supported
Hull MG, North K, Taylor H, Farrow A, Ford WC, Hull MG, North K,
this study by a Doctoral Research Fellowship. The HUNT Study is a
Taylor H, Farrow A, Ford WC. Delayed conception and active and
collaboration between HUNT Research Centre (Faculty of Medicine,
passive smoking. The Avon Longitudinal Study of Pregnancy and
Norwegian University of Science and Technology NTNU), Nord-
Childhood Study Team. Fertil Steril 2000;74:725 – 733.
Jain T, Hornstein MD. To pay or not to pay. Fertil Steril 2003;80:
Norman RJ, Noakes M, Wu R, Davies MJ, Moran L, Wang JX. Improving
Jensen A, Sharif H, Olsen JH, Kjaer SK. Risk of breast cancer and
gynecologic cancers in a large population of nearly 50,000 infertile
Danish women. Am J Epidemiol 2008;168:49 – 57.
Otis CL, Drinkwater B, Johnson M, Loucks A, Wilmore J. American
Kohrt WM, Bloomﬁeld SA, Little KD, Nelson ME, Yingling VR. American
College of Sports Medicine position stand. The Female Athlete Triad.
College of Sports Medicine Position Stand: physical activity and bone
Med Sci Sports Exerc 1997;29:i – ix.
health. Med Sci Sports Exerc 2004;36:1985 – 1996.
Redman L. Symposium: diet, nutrition and exercise in reproduction.
Kull M. The relationships between physical activity, health status and
Physical activity and its effects on reproduction. Reprod BioMed Online
psychological well-being of fertility-aged women. Scand J Med Sci
Rich-Edwards JW, Spiegelman D, Garland M, Hertzmark E, Hunter DJ,
Kumar A, Ghadir S, Eskandari N, DeCherney AH. Infertility. In
Colditz GA, Willett WC, Wand H, Manson JE. Physical activity, body
DeCherney AH, Nathan L, Goodwin TM, Laufer N (eds)., Current
mass index, and ovulatory disorder infertility. Epidemiology 2002;
Diagnosis and Treatment. Obstetrics & Gynecology. New York: McGraw
Rostad B, Schei B, Sundby J. Fertility in Norwegian women: results from a
Kurtze N, Rangul V, Hustvedt BE, Flanders WD. Reliability and validity of
population-based health survey. Scand J Public Health 2006;34:5 – 10.
self-reported physical activity in the Nord-Trondelag Health Study:
Russell JB, Mitchell D, Musey PI, Collins DC. The relationship of exercise
HUNT 1. Scand J Public Health 2008;36:52 – 61.
to anovulatory cycles in female athletes: hormonal and physical
Loucks AB, Verdun M, Heath EM. Low energy availability, not stress of
characteristics. Obstet Gynecol 1984;63:452 – 456.
exercise, alters LH pulsatility in exercising women. J Appl Physiol 1998;
Scully D, Kremer J, Meade MM, Graham R, Dudgeon K. Physical exercise
and psychological well being: a critical review. Br J Sports Med 1998;
Mahabir S, Baer DJ, Giffen C, Clevidence BA, Campbell WS, Taylor PR,
Hartman TJ. Comparison of energy expenditure estimates from
Shephard RJ. Limits to the measurement of habitual physical activity by
4 physical activity questionnaires with doubly labeled water estimates
questionnaires. Br J Sports Med 2003;37:197 – 206.
in postmenopausal women. Am J Clin Nutr 2006;84:230 – 236.
Sternfeld B, Ainsworth BE, Quesenberry CP. Physical activity patterns in a
Manini TM, Everhart JE, Patel KV, Schoeller DA, Colbert LH, Visser M,
diverse population of women. Prev Med 1999;28:313 – 323.
Tylavsky F, Bauer DC, Goodpaster BH, Harris TB. Daily activity
Vanhees L, Lefevre J, Philippaerts R, Martens M, Huygens W, Troosters T,
energy expenditure and mortality among older adults. JAMA 2006;
Beunen G. How to assess physical activity? How to assess physical
ﬁtness? Eur J Cardiovasc Prev Rehabil 2005;12:102 – 114.
Meacham RB, Joyce GF, Wise M, Kparker A, Niederberger C. Male
Warburton DE, Nicol CW, Bredin SS, Warburton DER, Nicol CW,
infertility. J Urol 2007;177:2058 – 2066.
Bredin SSD. Health beneﬁts of physical activity: the evidence. CMAJ
Miller YD, Trost SG, Brown WJ. Mediators of physical activity behavior
change among women with young children. Am J Prev Med 2002;
Warren MP, Perlroth NE, Warren MP, Perlroth NE. The effects of intense
exercise on the female reproductive system. J Endocrinol 2001;
Mora S, Cook N, Buring JE, Ridker PM, Lee IM. Physical activity and
Wilcox A, Weinberg C, Baird D. Caffeinated beverages and decreased
mechanisms. Circulation 2007;116:2110 – 2118.
fertility. Lancet 1988;2:1453 – 1456.
Morris SN, Missmer SA, Cramer DW, Powers RD, McShane PM,
Wisloff U, Nilsen TI, Droyvold WB, Morkved S, Slordahl SA, Vatten LJ. A
Hornstein MD. Effects of lifetime exercise on the outcome of in vitro
single weekly bout of exercise may reduce cardiovascular mortality: how
fertilization. Obstet Gynecol 2006;108:938 – 945.
little pain for cardiac gain? ‘The HUNT study, Norway’. Eur J Cardiovasc
Nattiv A, Loucks AB, Manore MM, Sanborn CF, Sundgot-Borgen J,
Warren MP. American College of Sports Medicine position stand.
Submitted on March 13, 2009; resubmitted on August 21, 2009; accepted onSeptember 1, 2009
scientific interest may be partly attributable to theprevious consensus that the female orgasm has noclear role in reproduction. This view was challengedby research showing that the orgasm helped facilitatesperm retention (Further evidence for the orgasm’s reproductive rolecomes from studies linking it with the menstrual cycle(More recently, this was corroborated in other studiesshowing tha
Chronic stress, glucocorticoid receptor resistance,inﬂammation, and disease riskSheldon Cohena,1, Denise Janicki-Devertsa, William J. Doyleb, Gregory E. Millerc, Ellen Frankd, Bruce S. Rabine,and Ronald B. TurnerfaDepartment of Psychology, Carnegie Mellon University, Pittsburgh, PA 15213; bDepartment of Otolaryngology/ENT, Children’s Hospital of Pittsburgh,Pittsburgh, PA 15224; cDepartment