Relationship between hair and salivary cortisol and pregnancy in women undergoing IVF
Introduction
The role that stress may exert on reproductive functioning has remained a topic of debate for decades (Boivin and Venetis, 2011, Matthiesen et al., 2011). This debate has been fuelled not least by the complexities of delineating plausible psychobiological pathways by which stress might influence reproductive outcomes (Whirledge and Cidlowski, 2010, Toufexis et al., 2014). Two main pathways have been considered: the hypothalamic pituitary adrenal (HPA) − hypothalamic pituitary gonadal (HPG) axis and the HPA − immune axis. The glucocorticoid cortisol, the main effector of the HPA axis, plays a central role in both pathways. With regard to the HPA-HPG pathway, elevated cortisol has been shown to inhibit sexual functions at all three levels of the HPG axis. At the level of the hypothalamus via inhibition of gonadotrophin releasing hormone (GnRh) secretion (Dubey and Plant, 1985, Kamel and Kubajak, 1987), at the level of the pituitary by interfering with GnRh induced luteinising hormone release (Briski and Sylvester, 1991), and at the level of the gonads by altering the stimulatory effect of gonadotrophins on sex steroid secretion (Bambino and Hsueh, 1981, Hsueh and Erikson, 1978).
With regard to the HPA-immune pathway it is well established that cortisol can regulate the activity of the immune system (Tsigos and Chrousos, 2002), including effects on the production of proinflammatory cytokines, tumour necrosis factor (TNF) and natural killer (NK) cell activity (Gatti et al., 1987, Mavoungou, 2006) all of which have been implicated in reproductive outcomes (Daher et al., 1999). However, the precise role of the immune system in pregnancy continues to be an area of debate. Taking the early phases of pregnancy as an example, it has been suggested that because the fetus is semi-allogenic to the maternal host, that a degree of immune suppression may favour conception (Thornton, 2010, Nepomnaschy et al., 2007). In contrast, it has been argued that, rather than immune suppression, a strong inflammatory response is required for conception to occur (Mor, 2011). These competing views are likely to be due, in part, to the fact that as observed by Mor, (2011), pregnancy has often been conceptualised as a single immunological event. However, there is evidence that it consists of three distinct phases (characterised by a pro-inflammatory (first trimester), anti-inflammatory (second trimester) and then finally a return to a pro-inflammatory state (third trimester). Furthermore, as noted by Christian (2015), even though the immune system is highly regulated by the neuroendocrine system, to date, there has been little integration of these literatures in the context of fertility and pregnancy: further impeding attempts to delineate the significance of the HPA-immune pathway.
These contrasting views regarding the role of the immune system during conception are further complicated by the presence or absence of glucocorticoid resistance (GR). GR refers to the phenomenon whereby immune function fails to downregulate despite the presence of chronically elevated cortisol (Cohen et al., 2012, Barnes and Adcock, 2009). Thus, two scenarios are possible. In the first, elevated cortisol results in immunosuppression (in the absence of GR). In the second, elevated cortisol (in the presence of GR) fails to down regulate inflammatory responses (Nepomnaschy et al., 2007, Thornton, 2010). These opposing effects of cortisol on immunity have fuelled the debate about the role of the immune system in fertility and specifically whether and when immune suppression, normal immune function or indeed heightened immune activity promotes or hinders reproductive outcomes.
Given what is understood about the direct effects of the HPA-HPG axis on reproduction, and the proposed role that immunosuppression may or may not play, divergent effects on pregnancy outcomes could be expected in women undergoing infertility treatment. Considering, for example, just the period prior to commencing treatment, increased cortisol, from the perspective of the HPA-HPG (and potentially also the HPA-immune axis in the presence of GR and excessive inflammation) could be expected to be associated with decreased likelihood of pregnancy. In contrast, from the perspective of an HPA-immune pathway (in the absence of GR), advocates of the view that immune suppression favours conception (e.g., Thornton, 2010, Nepomnaschy et al., 2007) may expect greater levels of cortisol to be associated with an increased likelihood of pregnancy. These contrasting outcomes were evident in a recent systematic review which reported that both high and low cortisol levels were associated with an increased likelihood of pregnancy in women undergoing IVF treatment (Massey et al., 2014). Twelve studies were identified that explored associations between cortisol (measured in blood, urine, saliva and follicular fluid) and establishment of clinical pregnancy. Three studies reported significant associations between elevated cortisol in follicular fluid and increased likelihood of clinical pregnancy (Andersen and Hornnes, 1994, Keay et al., 2002, Thurston et al., 2003). Four studies showed the opposite relationship albeit at different stages of treatment (Demyttenaere et al., 1992, Micheal et al., 1999, An et al., 2011, An et al., 2013). The remaining five studies failed to find any significant associations between cortisol and pregnancy outcomes (Lovely et al., 2003, Lewicka et al., 2003, Smeenk et al., 2005, Nouri et al., 2011, Csemiczky et al., 2000). These equivocal findings are in contrast to evidence from animal studies which have shown a more consistent relationship between elevated cortisol levels and impaired reproductive outcomes across a range of species (Dobson and Smith, 1995, Alejandro et al., 2014).
A number of methodological factors may have contributed to the findings in this review. First, the majority of studies failed to control for extraneous variables known to influence cortisol including time of day, caffeine consumption and body mass index (BMI) (Massey et al., 2014). Second, studies measured cortisol at different stages of treatment. Whilst IVF is a useful clinical model for exploring the effects of stress on reproductive outcomes, over half of the studies (7/12) measured cortisol following gonadotrophin administration (Andersen and Hornnes, 1994, Keay et al., 2002, Thurston et al., 2003, Micheal et al., 1999, Lovely et al., 2003, Lewicka et al., 2003, Nouri et al., 2011). Gonadotrophins have profound effects on the HPA axis and, therefore, are likely to have confounded any observed associations between HPA function and pregnancy. Third, studies relied exclusively on acute (saliva, blood) or short term (urine, follicular fluid) measures of cortisol. However, these measures are only able to provide information about cortisol levels over minutes (saliva/blood) to hours (urine/follicular fluid). At best, such measures may not accurately represent basal function or, at worst, may only capture transient levels of the hormone which are unlikely to influence health (Saxbe, 2008).
A recent methodological development that may prove promising for understanding HPA function is hair sampling which provides a retrospective index of cumulative cortisol exposure over extended periods of 3–6 months (Gow et al., 2010). The validity and reliability of hair cortisol sampling has been supported in both animal (Davenport et al., 2006) and human studies (Sauve et al., 2007, Stalder and Kirschbaum, 2012). Furthermore, the utility of hair sampling as a marker of long term HPA function has been demonstrated in a range of health contexts (Yamada et al., 2007, Kalra et al., 2007, Dettenborn et al., 2011, Steudte et al., 2011, Pereg et al., 2011, Braig et al., 2015).
The aim of this study was, therefore, to build upon previous work by exploring the relationship between acute (saliva) and chronic (hair) cortisol measured in the 1–2 weeks prior to commencing IVF treatment and treatment outcomes, as measured by the absence/presence of a clinical pregnancy. In view of the continued uncertainty regarding the precise role of the HPA-immune pathway in reproductive outcomes, we hypothesised that, in support of the HPA-HPG pathway elevated levels of acute and chronic cortisol would be associated with a reduced likelihood of pregnancy.
Section snippets
Recruitment
Recruitment took place at a single English fertility clinic between December 2012 to April 2014. Non-smoking women, with a body mass index (BMI) of 19–35, undergoing a long GnRH agonist protocol were eligible for participation. Study exclusion criteria included subjects with any recognised endocrinological health conditions (Kudielka et al., 2009). Patients taking concurrent corticosteroids were excluded from recruitment. Eligible patients received information regarding the study along with
Comparing participants and non-participants on demographic and clinical variables
Independent sample t-tests were conducted to compare patients providing salivary and or hair cortisol data (n = 135) and the original sample who were recruited (n = 190). For all measures, no statistically significant differences were evident between these groups on age, number of years infertile, number of previous attempts, fertilisation rate (%), number of oocytes or BMI (all p’s > 0.05: data not shown).
Patient/treatment demographics and levels of cortisol
The mean age of the participants was 34 years (mean 34.5 SD 4.8). The study sample comprised of
Discussion
To our knowledge, this is the first study to explore associations between both acute and chronic measures of cortisol exposure on the outcome of IVF. We examined whether levels of cortisol prior to IVF treatment were associated with the likelihood of clinical pregnancy. Measures of salivary cortisol (CAR, AUCg, AUCi and diurnal slope) did not predict clinical pregnancy. However, hair cortisol concentrations did, with greater levels associated with a reduced likelihood of conception. Our
Conflicts of interest
None.
Contributions
All authors contributed to the design of the study. Adam Massey collected the data. Adam Massey and Jane Perry analysed the data. Catherine Pinicott-Allen and Adam Massey analysed hair samples. Adam Massey, Bruce Campbell, Nick Raine-Fenning, Jane Perry and Kavita Vedhara wrote the paper. All authors have approved the final article.
Role of the funding source
NURTURE fertility funded the project but did not play a role in the study design; collection, analysis or interpretation of data, in the writing of the report or in the decision to submit the article for publication.
Acknowledgements
None.
References (64)
- et al.
Pituitary-adrenal and sympathetic nervous system responses to psychiatric disorders in women undergoing in vitro fertilization treatment
Fertil. Steril.
(2011) - et al.
Glucocorticoid resistance in inflammatory diseases
Lancet
(2009) - et al.
Determinants of maternal hair cortisol concentrations at delivery reflecting the last trimester of pregnancy
Psychoneuroendocrinology
(2015) - et al.
Tumor necrosis factor during pregnancy and at the onset of labor and spontaneous abortion
Eur. J. Obst. Gynaecol. Reprod. Biol.
(1999) - et al.
Analysis of endogenous cortisol concentrations in the hair of rhesus macaques
Gen. Comp. Endocrinol.
(2006) - et al.
Coping and the ineffectiveness of coping influence the outcome of in vitro fertilisation through stress responses
Pychoneuroendocrinology
(1992) - et al.
Inhibition by cortisol of human natural killer (NK) cell activity
J. Steriod Biochem.
(1987) - et al.
Hair as a retrospective calendar of cortisol production–increased cortisol incorporation into hair in the third trimester of pregnancy
Psychoneuroendocrinology
(2009) - et al.
Why do we respond so differently? Determinants of human salivary cortisol responses to challenge
Psychoneuroendocrinology
(2009) - et al.
Transformation techniques for cross sectional and longitudinal endocrine data: application to salivary cortisol concentrations
Psychoneuroendocrinology
(2013)