Animal Reproduction Science 60–61 Ž2000.

743–752
www.elsevier.comrlocateranireprosci

What is stress, and how does it affect reproduction?

Hilary Dobson ) , R.F. Smith
Department of Veterinary Clinical Science and Animal Husbandry, UniÕersity of LiÕerpool, Leahurst, Neston,
Wirral, CH64 7TE, UK

Abstract

Stress is revealed by the inability of an animal to cope with its environment, a phenomenon
that is often reflected in a failure to achieve genetic potential. Field data from dairy cows show
that stressors such as milk fever or lameness increase the calving to conception interval by 13–14
days, and an extra 0.5 inseminations are required per conception. We suggest that a variety of
endocrine regulatory points exist whereby stress limits the efficiency of reproduction. Transport
produces an immediate constant increase in arginine vasopressin ŽAVP. and corticotrophin-releas-
ing hormone ŽCRH. secretion in ewes, but adrenocorticotrophic hormone ŽACTH. reaches a
maximum in the first hour while cortisol is highest during the second hour. In contrast, after an
insulin injection, the hypothalamo–pituitary–adrenal ŽHPA. response is delayed occurring only
after glucose decreases below a threshold. Changes in AVP, CRH and ACTH each follow a
similar time course, but eventually the secretion of AVP and CRH decreases while glucose is still
at a nadir. Negative feedback effects appear to operate mainly at the pituitary level during
transport but at the hypothalamus during hypoglycaemia.
We also have endocrine evidence to show that stressors interfere with precise timings of
reproductive hormone release within the follicular phase. Transport, or insulin, reduce the
frequency and amplitude of gonadotrophin-releasing hormone and LH pulses, suggesting that
these stressors exert effects at the hypothalamus or higher centres in the brain. Both stressors also
delay the onset of the luteinising hormone ŽLH. surge. Preliminary results suggest that opioids
mediate these effects but progesteronerglucocorticoid receptors are not involved because the
antagonist, RU486, is unable to reverse insulin-induced delays in the LH surge. There is also
evidence to support effects at pituitary level because exogenous ACTH, or transport, reduce the
amount of LH released by challenges with GnRH. The reduction in endogenous GnRHrLH
secretion ultimately deprives the ovarian follicle of adequate gonadotrophin support leading to
reduced oestradiol production by slower growing follicles. Thus, there is a level of interference by
stressors at the ovary. Reproduction is such an important physiological system that animals have

)
Corresponding author.

0378-4320r00r$ - see front matter q 2000 Elsevier Science B.V. All rights reserved.
PII: S 0 3 7 8 - 4 3 2 0 Ž 0 0 . 0 0 0 8 0 - 4
744 H. Dobson, R.F. Smith r Animal Reproduction Science 60–61 (2000) 743–752

to ensure that they can respond to their surroundings; thus, it is advantageous to have several
protein mechanisms, i.e. at higher brain, hypothalamus, pituitary and target gland levels. However,
when pushed too far, subfertility occurs. q 2000 Elsevier Science B.V. All rights reserved.

Keywords: Stress; Reproduction; Subfertility

1. Introduction

‘‘Stress’’ is responsible for many things, including subfertility. Many agricultural

advisers and veterinarians are very familiar with those intangible factors that reduce
fertility on farms but often they are unable to pinpoint precise contributory causes —
and hence blame ‘‘stress’’.
This, in itself, provides a definition of ‘‘stress’’, that is, the inability of an animal to
cope with its environment, a phenomenon that is revealed by a failure to achieve genetic
potential, e.g. for growth rate, milk yield, disease resistance, or fertility.

2. Field observations

Strong evidence that stressors affect reproductive efficiency in dairy cattle has been
gained by comparing fertility data of normal cows and herd-mates suffering from
various stressful clinical conditions ŽTable 1..
Furthermore, evidence of a social stressor affecting fertility has been provided by a
behavioural study, which identified cows that changed social position in the herd
hierarchy within the breeding period. Those cows that increased social status were more
fertile and had better milk production figures than those with a lowering of social status
ŽTable 2; Dobson, unpublished data.. They also had a different lameness score on a 0–5
point scale, with 5.0 representing a very lame animal.
Unfortunately, the clinical conditions in Table 1 and social interactions exemplified
in Table 2 are very common in the dairy industry. Worse than that, these factors are

Table 1
Summary of fertility parameters of dairy cows with clinical disease conditions diagnosed and treated in the
postpartum period Ždata selected from Borsberry and Dobson, 1989; Collick et al., 1989.. Each diseased cow
was compared with a similar untreated healthy herd-mate
Pairs Ca — 1st service Ždays. Ca — conception Ždays. Inseminations per conception
Control Diseased Control Diseased Control Diseased
) )
Milk fever 95 61 67 75 88 1.2 1.7 )
Ovarian cyst 73 63 78 ) ) 77 142) ) 1.3 3.2 ) )
Sick 9 72 69 76 157) 1.3 3.1) )
Lameness 427 68 72 ) 86 100) ) 1.7 2.1) ) )
)
P - 0.05.
))
P - 0.01.
)))
P - 0.001.
 H. Dobson, R.F. Smith r Animal Reproduction Science 60–61 (2000) 743–752 745

Table 2
Summary of fertility and milk production figures for 45 pairs of cows that displayed increasing or decreasing
social status Žbased on dominance and submissiveness. during the breeding period in three commercial dairy
herds
Change in social status
Increase Decrease
Calving to conception Ždays. 97 143 )
Inseminations per conception 1.6 2.2
Milk yield Žkgrday. q0.58 y1.03 )
Somatic cell counts Ž’000rml. y18 q371)
Difference in lameness score y0.21 q0.54 )
)
P - 0.05.

clearly hindering the genetic progress of one of the major domesticated species in the
world. No doubt similar data can be compiled for other commercially important species.
If we are to avoid paying this price for domestication of any species, it is necessary to
learn more about how animals respond to stressors, and how this affects the mechanisms
controlling reproductive efficiency.

3. Examination of control mechanisms

Studying the effects of stress on reproduction is beset with difficulties. The complex
nature of some stressors in the modern farm environment simultaneously exposes
animals to several different stimuli. Furthermore, there is considerable variability
between individuals in response to a given stimulus. Added to this, is the overriding
importance of the reproductive system to pass genes on to the next generation. This last
issue means that animals have developed several strategies to cope with environmental
problems including alternative responses to compensate for failure of any part of the
protection mechanism.
In brief, we hypothesise that there are several regulatory by which stressors regulate
reproductive mechanisms. Endocrine systems appear to be an ideal way of coordinating
this regulation throughout the whole body. In order to unravel the complexity of
stress-induced subfertility, it is necessary to study the reaction to stressors of repeatable
severity, firstly by examining responses to clearly defined stimuli, and then by investi-
gating the influence on reproductive mechanisms.
The sheep is often used as a model for mammalian reproduction and this species
provides a more manageable, less expensive, experimental animal than the cow. In
addition, mechanisms controlling normal reproductive endocrinology of the sheep are
well characterised and are similar to those in the cow. Another species that might be
used for these studies, the rat, is too small for the simultaneous studies proposed.
Furthermore, gonadotrophin secretion in the rat is controlled by a species-specific
reliance on circadian adrenal function that limits interpretation of observations concern-
ing the effects of stress-induced adrenal activity on reproductive function.
746 H. Dobson, R.F. Smith r Animal Reproduction Science 60–61 (2000) 743–752

4. Responses to specific stressful stimuli

The combined physical and psychological stimulus of 2 h transport in a vehicle

produces an immediate and constant increase in both arginine vasopressin ŽAVP. and
corticotrophin-releasing hormone ŽCRH. concentrations in hypophyseal-portal blood of
ewes, but the adrenocorticotrophic hormone ŽACTH. response reaches a maximum in
the first hour while cortisol concentrations are highest during the second hour ŽFig. 1;
Smith et al., 1997.. This suggests that the input into the hypothalamus is constant during
this stimulus, even though the final output Žmeasured as plasma cortisol. changes during
the stimulus. Insulin-induced hypoglycaemia, a physiological stimulus that does not
require cognitive processing, produces a different response. After an insulin injection,
the hypothalamo–pituitary–adrenal ŽHPA. response only occurs after the blood glucose
concentration decreases below a threshold Žapproximately 3 mmolrl.. Changes in AVP,
CRH and ACTH each follow a similar time course, but eventually the secretion of AVP
and CRH decreases while plasma glucose is still at a nadir ŽFig. 2.. Plasma cortisol

Fig. 1. HPA responses to 2 h transport. Mean and S.E.M. of data obtained from four ewes transported on four
occasions at two week intervals.
 H. Dobson, R.F. Smith r Animal Reproduction Science 60–61 (2000) 743–752 747

Fig. 2. HPA responses to insulin-induced hypoglycaemia. Mean and S.E.M. of data obtained from three ewes
treated on four occasions at two week intervals.

concentrations remain elevated due to the long half-life of this steroid. In this case,
during the prolonged presence of a stressful input Žlow glucose concentration., there is a
decrease in hypothalamic AVPrCRH.
To limit over-stimulation of the stress axis and its deleterious effects, there are
mechanisms to control the HPA including cortisol negative feedback effects at hypotha-
lamic andror pituitary levels to restrict on-going responses to a stimulus. However, an
opposite balancing mechanism Žfacilitation. also occurs within the HPA so that the
responses are not totally inhibited ŽDallman et al., 1992..
From the above evidence comparing responses to transport and insulin, we suggest
that negative feedback effects operate mainly at the pituitary level during transport and
at the hypothalamus during hypoglycaemia. However, these differences could be
inherent in the pituitary responses to different concentrations and ratios of AVPrCRH
secreted in response to different stimuli.
Increasing the duration of transport or hypoglycaemia does not prolong the cortisol
response at the same magnitude. Furthermore, if transport is repeated every week or at
748 H. Dobson, R.F. Smith r Animal Reproduction Science 60–61 (2000) 743–752

longer intervals, there is no reduction in total cortisol response. However, if transport is

repeated daily, some ewes have reduced responses after four trips ŽSmith and Dobson,
unpublished observation.. Individual differences in response may be due to prenatal or
early life experiences ŽLay et al., 1997; Liu et al., 1997. or genetic background
ŽRomeyer and Bouissou, 1992..
Our working hypothesis is that the time-course of a response varies at each level of
the system and depends on the nature of the stimulus. The corresponding time-course of
the stress effect on the reproductive system may indicate which precise components of
the HPA are important in the interaction with the hypothalamo–pituitary–ovarian ŽHPO.
axis.

5. How do stressful stimuli affect reproduction?

In the follicular phase of a normal oestrous cycle, the correct pattern of go-
nadotrophin-releasing hormone ŽGnRH. secretion from the hypothalamus leads to
increased pulsatile release of luteinising hormone ŽLH. from the pituitary gland ŽMoenter
et al., 1990.. In concert with follicle stimulating hormone, this dictates the rate of
follicular growth and oestradiol production, ultimately leading to a preovulatory LH
surge and ovulation ŽMcNeilly et al., 1991..
In order to achieve a perfectly timed LH surge, a series of closely controlled events
must occur within the hypothalamus and pituitary gland. After removal of the suppres-
sive effects of progesterone during luteolysis, GnRH Žand thus LH. pulses are secreted
with increasing frequency, to culminate eventually in continuous secretion at the onset
of the LH surge in response to the positive-feedback effects of oestradiol ŽEvans et al.,
1995..
In view of the complications incurred with repeatability, habituation and duration of
stressors as already high-lighted, these aspects have to be standardised as much as
possible when examining the influence of stress responses on physiological mechanisms
such as reproduction. Furthermore, the effects of more than one stressor must be
investigated in order to avoid the dangers inherent with stressor-specific artefacts.
However, in spite of some differences between stress responses discussed above, there
are some quite surprisingly consistent effects on reproductive endocrinology.
From a series of experiments conducted over the past 5 years, we suggest that
stressors reduce fertility by interfering with the mechanisms that regulate the precise
timings of events within the follicular phase. Acute stressors Žeither transport or
hypoglycaemia. imposed at precisely defined times have been investigated for effects on
different parts of the reproductive control mechanism.
Transport for 4 or 8 h reduces the frequency and amplitude of LH pulses especially
within the first few hours in ovariectomised ewes or intact animals in the late follicular
phase ŽDobson et al., 1999b; Phogat et al., 1999b.. Similar effects have been observed
during insulin-induced hypoglycaemia ŽFig. 3. — even though glucose concentrations
decrease after insulin but increase during transport. The reduction in LH pulse frequency
suggests an effect of both these stressors on GnRH pulsatile secretion mediated through
effects at the hypothalamus or higher centres in the brain; whereas effects on LH pulse
 H. Dobson, R.F. Smith r Animal Reproduction Science 60–61 (2000) 743–752 749

Fig. 3. Mean LH concentration, LH pulse frequency and amplitude in 10 ovariectomised ewes for 4-h periods
before Žopen bars., during Žfilled bars. and after Žhatched bars. insulin Ž2 IUrkg; top panels. or transport Ž4 h;
lower panels.. ) denotes significantly different Ž P - 0.05. from pretreatment value.

amplitude could either be mediated by the hypothalamus, or at pituitary level. Direct

proof of the suppressive effects of an acute stressor on GnRH secretion has been
provided by Battaglia et al. Ž1997. after endotoxin administration.
In addition, there is evidence from both in vitro perifusions and in vivo experiments
to show that exogenously increased ACTH concentrations or transport reduce the
amount of LH released by challenges with small doses of GnRH ŽPhogat et al., 1997,
1999a,b.. This provides support for additional effects at pituitary level.
Clearly, activation of the hypothalamus–pituitary–adrenal axis by stressors reduces
the pulsatility of GnRHrLH by actions at both the hypothalamus and pituitary gland,
ultimately depriving the ovarian follicle of adequate LH support. This will lead to
reduced oestradiol production by slower growing follicles. Such a hypothesis is sup-
ported by the marked decrease in oestradiol secretion observed after reducing the
frequency of exogenous LH pulses driving follicular growth in an ovarian autotransplant
model ŽDobson et al., 1999a..
A combination of the above effects on LH pulsatility at hypothalamic and pituitary
levels no doubt contributes to the delay and reduced magnitude of the LH surge
observed after transport or insulin administration in the follicular phase just prior to the
expected LH surge ŽDobson et al., 1999c; Table 3.. This effect on the LH surge control
mechanism could be exerted directly via an influence of GnRH on production of its own
receptors, or indirectly by the induced reduction in oestradiol which, in turn, will alter
the balance of systems controlling LH surge release. Thus, another level of interference
Žat the ovary. has been revealed to play a part in the multi-centred effects of stress on
reproductive control mechanisms.
One of the key sites for the changes exerted by stressful stimuli appears to be the
neuronal control of GnRH secretion. We have recently carried out studies on the effects
750 H. Dobson, R.F. Smith r Animal Reproduction Science 60–61 (2000) 743–752

Table 3
Time of onset of the LH surge relative to progesterone withdrawal in intact control ewes and those treated with
insulin, with or without additional naloxone or RU486
Treatment Onset of LH surge Žhours after P withdrawal.
Control Ž ns9. 62"7 a
Insulin alone Ž5 IUrkg at 38 and 40 h; ns13. 76"1b
Insulinqnaloxone infusion
Ž1 mg naloxonerkgrh for 12 h starting at 37 h; ns 5. 61"5a
InsulinqRU486 Ž1 injection of 100 mg at 37 h; ns 4. 78"4b
a
Significantly different from insulin alone group Ž P - 0.05..
b
Significantly different from control group Ž P - 0.05..

of two neurotransmitter regulators, and the preliminary results are very interesting.
Infusion of the opioid antagonist, naloxone just before insulin administration prevented
the delay in the onset of the LH surge observed after insulin alone ŽTable 3.. This
clearly implicates opioids in the mediation of stress-induced changes in LH secretion.
Furthermore, it would appear that progesteronerglucocorticoid receptors are not in-
volved in the interaction between the stress and reproductive axes because the antagonist
RU486 was unable to reverse the insulin-induced delays in the LH surge ŽTable 3..

6. The link between stress-induced low LH pulse frequency and cases of subfertility
Within the growing follicle, the oocyte maintains direct contact with granulosa cells
by means of cellular projections through the zona pellucida ŽMoor et al., 1980.. Thus,
events influencing the integrity of follicular function can have direct effects on oocyte
viability. These effects are not always immediately obvious, for example, it is known
that mRNAs are laid down in the oocyte nucleus but not translated until the 8-cell stage
of conceptus development ŽStaigmiller and Moor, 1984.. Consequently, any event that
changes granulosa cell activity may influence pregnancy rates. Indeed, Mihm et al.
Ž1994. have provided evidence for reduced pregnancy rates after prolonging the duration
of the follicular phase by artificially delaying the onset of the LH surge.
It is envisaged that in some situations, such as during the chronic stress of more
severe lameness or fever, the pulse GnRHrLH frequency will be so slow that initial
follicular growth will occur but will be unable to continue in to the later stages that
depend on faster pulse frequencies. Thus, the animal fails to maintain oestrous cycles
and the consequent anoestrus is easily recognised by veterinarians.
In slightly less stressful situations, GnRHrLH pulse frequency may be just fast
enough to support follicular growth, but because it is on a knife-edge, it will be
susceptible to interruption or variation by otherwise innocuous stimuli. In this case, the
integrity of granulosa cells and thus the oocyte may be compromised, and although
oestrus and fertilisation may occur, the conceptus will fail to develop into a pregnancy.
This is reflected in the insidious idiopathic subfertility recognised by agricultural
advisers and veterinarians.
A third scenario could exist in which pulse frequency is sufficient to get a follicle
into the later stages of development but is not quite fast enough to provide correct
 H. Dobson, R.F. Smith r Animal Reproduction Science 60–61 (2000) 743–752 751

GnRH priming of the pituitary andror adequate oestradiol production. Hence, an

inappropriate LH surge is generated and, as it is unable to cause ovulation and
luteinisation, the follicle persists to produce the clinically recognised cystic ovarian
syndrome.

7. Conclusion

In evolutionary terms, progression from single cell organisms to the complexity of

mammals has required the development of communication systems throughout the body
via which regulatory mechanisms are exerted. Intracellular mechanisms provide a basis
for contact within individual cells, an endocrine system working through the blood
circulatory system provides another level of control in more complex animals, and a
further level of control exists via the nervous system, masterminded by centres in the
brain.
Reproduction is a very important physiological system for the furtherance of a
species, and this has to succeed despite the imposition of sometimes detrimental
environmental stimuli. To ensure that an animal can respond to its surroundings, it is
advantageous to have several lines of defence as exemplified above by the different
levels of stress response, i.e. higher brain, hypothalamus, pituitary and adrenal glands.
Likewise each of these responses has influence on the different levels of the reproduc-
tive organisation, i.e. higher brain, hypothalamus, pituitary and gonads. Such diverse
methods of control to ensure the success of a species might also have the advantage of
energy conservation, so that less detrimental stressors may be dealt with at one level,
while more severe circumstances Žor several additive situations. may require responses
at many levels.
The continued genetic development of several species, carried out to meet Man’s
needs, especially for food production, is now meeting a stumbling block in that fertility
appears to be reducing. For example, conception rates in high-producing dairy cattle are
thought to be declining. Similarly, in some human situations, increased levels of
‘‘stress’’ lead to an inability to reproduce. Are all these animals trying to tell us
something?

Acknowledgements

The authors are grateful for the collaboration of many colleagues who have worked in
the Reproduction-Stress Research Group in the University of Liverpool Faculty of
Veterinary Science. We are also grateful for discussions with Gerald Lincoln from
Edinburgh.

References
Battaglia, D.F., Bowen, J.M., Krasa, H.B., Thrun, L.A., Viguie, C., Karsch, F.J., 1997. Endotoxin inhibits the
reproductive neuroendocrine axis while stimulating adrenal steroids: a simultaneous view from hypophy-
seal portal and peripheral blood. Endocrinology 138, 4273–4281.
752 H. Dobson, R.F. Smith r Animal Reproduction Science 60–61 (2000) 743–752

Borsberry, S., Dobson, H., 1989. Periparturient diseases and their effect on reproductive performance in five
dairy herds. Vet. Rec. 124, 217–219.
Collick, D.W., Ward, W.R., Dobson, H., 1989. Associations between types of lameness and fertility. Vet. Rec.
125, 103–106.
Dallman, M.F., Akana, S.F., Scribner, K.A., Bradbury, M.J., Walker, C.-D., Strack, A.M., Cascio, C.S., 1992.
Stress, feedback and facilitation in the hypothalamo–pituitary–adrenal axis. J. Neuroendocrinol. 4,
517–526.
Dobson, H., Campbell, B.K., Scaramuzzi, R.J., Baird, D.T., 1999a. Effect of reducing LH pulse frequency and
amplitude on ovarian oestradiol production in the ewe. In: 5th International Symposium on Reproduction
in Domestic Ruminants. J. Reprod. Fertil., Suppl. 54, Žin press..
Dobson, H., Tebble, J.E., Ozturk, M., Smith, R.F., 1999b. J. Reprod. Fertil., Žin press..
Dobson, H., Tebble, J.E., Phogat, J.B., Smith, R.F., 1999c. Effect of transport on pulsatile and surge secretion
of LH in ewes in the breeding season. J. Reprod. Fertil. 116, 1–8.
Evans, N.P., Dahl, G.E., Mauger, D., Karsch, F.J., 1995. Estradiol induces both qualitative and quantitative
changes in the pattern of gonadotropin-releasing hormone secretion during the presurge period in the ewe.
Endocrinology 136, 1603–1609.
Lay, D.C., Randel, R.D., Friend, T.H., Carroll, J.A., Welsh, T.H., Jenkins, O.C., Neuendorff, D.A., Bushong,
D.M., 1997. Effects of prenatal stress on the fetal calf. Domest. Anim. Endocrinol. 14, 73–80.
Liu, D., Diorio, J., Tannenbaum, B., Caldji, C., Francis, D., Freedman, A., Sharma, S., Pearson, D., 1997.
Maternal care, hippocampal glucocorticoid receptors, and hypothalamic–pituitary–adrenal responses to
stress. Science 277, 1659–1662.
McNeilly, A.S., Picton, H.M., Campbell, B.K., Baird, D.T., 1991. Gonadotrophic control of follicle growth in
the ewe. J. Reprod. Fertil. 92, 177–186.
Mihm, M., Baguisi, A., Boland, M.P., Roche, J.F., 1994. Association between the duration of dominance of
the ovulatory follicle and pregnancy rate in beef heifers. J. Reprod. Fertil. 102, 123–130.
Moenter, S.M., Caraty, A., Karsch, F.J., 1990. The estradiol-induced surge of gonadotropin-releasing hormone
in the ewe. Endocrinology 127, 1375–1384.
Moor, R.M., Smith, M.W., Dawson, R.M.C., 1980. Measurement of intercellular coupling between oocytes
and cumulus cells using intracellular markers. Exp. Cell Res. 126, 15–29.
Phogat, J.B., Smith, R.F., Dobson, H., 1997. Effect of adrenocorticotrophic hormone on gonadotrophin-releas-
ing hormone-induced luteinizing hormone secretion in vitro. Anim. Reprod. Sci. 48, 53–65.
Phogat, J.B., Smith, R.F., Dobson, H., 1999a. Effect of adrenocorticotrophic hormone ŽACTHŽ1–24.. on
ovine pituitary gland responsiveness to exogenous pulsatile GnRH and oestradiol-induced LH release in
vivo. Anim. Reprod. Sci. 55, 193–203.
Phogat, J.B., Smith, R.F., Dobson, H., 1999b. Effect of transport on pituitary responsiveness to exogenous
pulsatile GnRH and oestradiol-induced LH release in intact ewes. J. Reprod. Fertil. 116, 9–18.
Romeyer, A., Bouissou, M.F., 1992. Assessment of fear reactions in domestic sheep, and influence of breed
and rearing conditions. Appl. Anim. Behav. Sci. 34, 1–2.
Smith, R.F., Gore, S.W., Phogat, P.B., Dobson, H., 1997. The psychological stress of transport stimulates both
CRF and AVP secretion into hypophysial portal blood. J. Endocrinol. 152, P172, Suppl.
Staigmiller, R.B., Moor, R.M., 1984. Effect of follicle cells on the maturation and developmental competence
of ovine oocytes matured outside the follicle. Gamete Res. 9, 221–229.