Chronic psychological stressors during pregnancy have long been associated with adverse pregnancy outcomes including preterm birth,1–3 low birth weight,4 maternal mood disorders,5 and adverse child neurodevelopment,6–8 yet the mechanism by which stress influences these outcomes is unclear. Cortisol, the hormonal endpoint of physiologic stress-related activation of the hypothalamic–pituitary–adrenal axis, has been associated with preterm birth9 and is often cited as a potential mediator between stress and pregnancy outcomes.10 Cortisol fluctuates rapidly in response to even subtle changes in the environment and single serum and saliva measurements do not reflect chronic levels. Multiple salivary samples collected across days, as representative of the varied intraindividual patterns of cortisol, have been utilized to improve stress characterization.11–13 Although an improvement over single samples, these approaches also have limitations.12 These limitations likely contribute to the lack of clarity concerning the status of cortisol physiology in mediating relationships between stress reflected as hypothalamic–pituitary–adrenal activity and adverse pregnancy outcomes such as preterm birth.
Cortisol is deposited in hair14 and normally increases as gestation progresses.15 Assessment of the hair cortisol concentration provides a reliable and noninvasive measure of chronic hypothalamic–pituitary–adrenal activity.16,17 Differences have been noted in hair cortisol concentration in relation to chemical hair processing,18 season of delivery, and hair color at the end of the third trimester in term pregnancies.16 Taking these data into consideration, the objective of this study was to determine the relationships between psychological and physiologic measures of stress, mood, gestational age at delivery, and preterm birth.
MATERIALS AND METHODS
Our prospective cohort study population consisted of women recruited during routine scheduled prenatal care visits during the first trimester of pregnancy. The patients were between 18 and 45 years of age and planned to deliver at Denver Health and Hospital Authority, Denver, Colorado. Women were excluded if they had a multiple gestation; reported daily illicit drug or alcohol abuse; had medical conditions requiring chronic corticosteroid use (topical, inhaled, or oral); had active infections such as human immunodeficiency virus, hepatitis, influenza, or other clinically significant viral illnesses during pregnancy19; or used bleach or peroxide products on their hair because these products are known to affect measured hair cortisol.18 Gestational age was a best clinical estimate based on a combination of participant-reported last menstrual period and first dating ultrasonogram. Hollingshead index was used to characterize socioeconomic status and is based on marital status, employment, level of education attained, and occupational prestige.20 Preterm birth was defined as delivery at less than 37 weeks of gestation and only preterm birth resulting from spontaneous preterm labor or preterm premature rupture of membranes was included. This project was approved by the Colorado multiple institutional review board and all participants provided informed consent.
Maternal hair was cut three times during pregnancy. The first haircut occurred between 16 and 18 weeks (16 weeks), the second between 28 and 30 weeks (28 weeks), and the third between 38 and 42 weeks (40 weeks) of gestation. Because hair provides a retrospective cortisol history and grows at 0.8–1 cm per month,21 each 3-cm long segment was used to represent 3 preceding months of cumulative physiologic hypothalamic–pituitary–adrenal activity, approximating first-, second-, and third-trimester cumulative cortisol concentrations.
Hair was cut using stylists' scissors as close to the scalp as possible along a 4-cm-long×1- to 2-mm-wide strip at the posterior vertex of the scalp in the same area each time, as previously described.15 The hair samples were taped with painter's tape to aluminum foil for protection and storage22 until all of a participant's hair samples were available and could be processed and analyzed within the same assay to limit assay variability. New growth that emerged between hair collections allowed us to confirm that each participant's hair growth was consistent with published rates.21 If a participant delivered at less than 34 weeks of gestation, her hair was cut on postpartum day 0–2, but these cortisol values were not included in third-trimester hair cortisol concentration analyses.
After the hair segments were measured and the proximal 3 cm from the scalp cut, the hair was placed in a preweighed 2-mL cryovial and washed three times in 100% isopropanol and dried as previously described.15 After washing, drying, and reweighing, the hair and tubes were frozen in liquid nitrogen and ground for 4–5 minutes in the same tube using a ball mill with a stainless steel ball bearing added to the tube. Specially milled aluminum cassettes were designed to hold three of these cryovials. The powdered hair (5–20 mg) was extracted at room temperature in the same cryovial in 1,000 microliters high-performance liquid chromatography-grade methanol overnight. After methanol extraction, the cryovial was centrifuged and the supernatant was removed, placed into a second microcentrifuge tube, and dried under a stream of nitrogen. The extracts were reconstituted with assay diluent based on hair weight. One hair cortisol concentration per trimester per participant was determined using a commercial high-sensitivity enzyme immunoassay kit per manufacturer's instructions. A pooled control of previously ground hair was extracted as described previously and included on each enzyme immunoassay plate in duplicate for determination of interassay coefficients of variation. Interassay coefficient of variation for the control hair pool was 9.2% and intraassay coefficient of variation for duplicates was 2.8%. Hair cortisol concentrations are reported as pg/mg.
At the five study time points, approximately 16, 22, 28, 34, and 40 weeks of gestation, pregnant women completed psychological self-report measures that included: a Perceived Stress Scale,23 a Center for Epidemiologic Studies-Depression Scale,24 and the State-Trait-Anxiety Inventory, State version.25 The Perceived Stress Scale is a 14-question survey with possible scores ranging from 0 to 56. The higher the Perceived Stress Scale score, the higher the number and frequency of stressors, irritations, poor coping, anger, and difficulties in the preceding 4 weeks.23 The Center for Epidemiologic Studies-Depression Scale is a 20-question survey with possible scores ranging from 0 to 60 on which respondents rate the frequency of 20 depressive symptoms within the last week. It was designed to screen for major or clinical depression in adolescents and adults.24 The Spielberger State-Trait Anxiety Inventory, State version is a 20-question self-report that measures state (short-term, situational) anxiety with scores ranging from 20 to 8025 and questions are anchored to the past month. All self-report measures have been validated in multiple languages and multicultural populations including pregnancy or the postpartum period.26–28 For all self-report measures, the higher the score, the greater the symptomatology.
Participant characteristics including gravidity, parity, maternal body mass index, gestational age, birth weight, mode of delivery, antepartum or intrapartum complications, medications and illicit substances (assessed throughout pregnancy), and major life events were all recorded.
Analyses of variance and χ2 analyses were used to compare groups on demographic and study characteristics including: race–ethnicity; socioeconomic status29; level of education; marital status; tobacco use; hair cortisol concentration in the first, second, and third trimesters; gestational age at delivery; birth weight; and Perceived Stress Scale, Center for Epidemiologic Studies-Depression Scale, and Spielberger State-Trait Anxiety Inventory, State version scores. Hair cortisol concentrations were natural log (LN, base e) -transformed to normalize values. Pearson correlation analyses were used to examine cross-sectional associations among hair cortisol concentrations, Perceived Stress Scale scores, Center for Epidemiologic Studies-Depression Scale scores, Spielberger State-Trait Anxiety Inventory, State version scores, gestational age at delivery, and birth weight. Comparisons of hair cortisol concentration across trimesters were compared for the same participants using repeated-measures analysis of variance and t tests. Regression modeling was used to assess the relationship between gestational age at delivery and hair cortisol concentration in each trimester, controlling for race–ethnicity and any tobacco use.
To test our hypothesis that psychological stress affects physiologic stress and that both are associated with gestational age at delivery, a mediation analysis using the difference in coefficients approach was used.30 Mediation analyses clarify the relationships among an independent variable, a dependent variable, and a third explanatory variable, called the mediator. This approach compared the coefficient of the psychological measure (perceived stress) on gestational age at delivery both before and after controlling for the coefficient of the potential mediator, the physiologic measure (hair cortisol concentration). The standard error for the difference in coefficients was estimated using 5,000 bootstrap samples and used to estimate the conditional indirect effect. The association between gestational age at delivery and hair cortisol concentration was modeled using a third-degree polynomial. Race–ethnicity and tobacco use were included as covariates. Based on previously published studies of hair cortisol concentrations during pregnancy and a preterm birth rate of 12%, we determined that a sample size of 90 would have 80% power to detect a difference of 30% in hair cortisol concentration between the preterm- and term-delivered groups. All data were analyzed using IBM SPSS 22.0 and SAS 9.4 was used for the mediation analysis.
A total of 92 women were enrolled. Two outliers were inspected to determine characteristics that may have contributed to the value. One case of maternal peroxide use (hair color) and one participant with topical corticosteroid cream were discovered, and those participants' hair cortisol concentration values were excluded. Both of these participants delivered at term. Eleven (12%) births resulting from spontaneous preterm labor or preterm premature rupture of membranes occurred in the 90 pregnancies remaining. Women who delivered at term and preterm are presented in Table 1. Overall, participants were predominantly multiparous, white Hispanic, nonsmokers with a high school education or less, and of lower socioeconomic status. Women who delivered preterm were older, white non-Hispanic and African American, more likely to have used tobacco before pregnancy, and less likely to be married or living with a partner. Self-report measures of stress and mood and mean and standard deviation hair cortisol concentrations in the first, second, and third trimesters are also presented in Table 1. Women who delivered preterm (n=11) had higher mean second-trimester hair cortisol concentration compared with those who delivered at term (n=79) (Table 1). This difference was not seen between categories of term and preterm for first- or third-trimester hair cortisol concentrations.
Associations between other demographic factors and hair cortisol concentration were analyzed including level of education, marital status, socioeconomic status, mode of delivery, season of hair collection, and hair color, and the only significant finding was that hair cortisol concentration in African American participants was higher in the first and second trimesters (analysis of variance: first trimester P<.01, second trimester P=.01, third trimester P=.07) than white non-Hispanic and Hispanic participants.
Over the course of pregnancy, hair cortisol concentration rose as expected15 and was significantly higher in the third trimester compared with the first and second trimesters (Fig. 1). Maternal hair cortisol concentration in the first trimester correlated with both the second- and third-trimester hair cortisol concentration measurements: first compared with second trimester r=0.42, P<.01 and first compared with third trimester r=0.29, P=.01.
Average Perceived Stress Scale, State-Trait-Anxiety Inventory, State version, and Center for Epidemiologic Studies-Depression Scale scores at 16, 22, 28, 34, and 40 weeks of gestation are presented in Table 1 from both term- and preterm-delivered participants. There was a significant difference between Perceived Stress Scale score in the term- and preterm-delivered groups at 16 weeks of gestation (P=.01). Otherwise, Perceived Stress Scale, Center for Epidemiologic Studies-Depression Scale, and State-Trait-Anxiety Inventory, State version scores did not differ significantly over time or between groups (all P values >.05). Scores on the Perceived Stress Scale, Center for Epidemiologic Studies-Depression Scale, and State-Trait-Anxiety Inventory, State version were highly correlated across gestation, both within and across measures (all P values <.01). Of correlation analyses for all 15 self-report measures (three measures at five study time points) and gestational age at delivery, only perceived stress at 16 weeks of gestation was correlated with gestational age at delivery (Fig. 2; r=−0.30, P<.01).
Correlation analyses were done for all relationships at all five study time points between Perceived Stress Scale score, Center for Epidemiologic Studies-Depression Scale score, and State-Trait-Anxiety Inventory, State version score and hair cortisol concentration in the first, second, and third trimesters (Table 2). The only significant relationships with first-trimester hair cortisol concentration were Perceived Stress Scale and Center for Epidemiologic Studies-Depression Scale at the 40-week study visit. In the second trimester, hair cortisol concentration was correlated with all three psychological measures at 16 weeks of gestation and Center for Epidemiologic Studies-Depression Scale and State-Trait-Anxiety Inventory, State version at 28 weeks of gestation. Similar associations were noted between third-trimester hair cortisol concentration and 16- and 28-week Center for Epidemiologic Studies-Depression Scale and State-Trait-Anxiety Inventory, State version.
Correlation analyses assessing the relationships between hair cortisol concentration and gestational age at delivery were not significant in the first (r=−0.16, P=.15) or third (r=0.13, P=.25) trimesters. However, second-trimester hair cortisol concentration was significantly correlated with gestational age at delivery (Fig. 3; r=−0.25, P=.02). All women who had a preterm birth at less than 37 weeks of gestation had a second-trimester hair cortisol concentration above the median, 2.3 pg/mg.
After taking into consideration the preceding analyses, which included a significant association between 16-week Perceived Stress Scale and gestational age at delivery (Fig. 2), a significant association between 16-week Perceived Stress Scale and second-trimester hair cortisol concentration (r=0.28, P=.007), and a significant association between the hair cortisol concentration and gestational age at delivery (Fig. 3), we used mediation analyses to assess whether gestational age at delivery was more strongly associated with the physiologic measure (second-trimester hair cortisol concentration) or the psychological measure (Perceived Stress Scale at 16 weeks). Race–ethnicity and tobacco use were included as covariates in the analyses. Gestational age at delivery was more strongly associated with second-trimester hair cortisol concentration than 16-week perceived stress and the correlation between perceived stress and gestational age at delivery was no longer significant after including the mediator (hair cortisol concentration) using the difference in coefficients (indirect effect and bootstrapped standard error of −0.05 [0.02], P=.04). This indicates that hair cortisol concentration meets criteria for a mediator of the relationship between 16-week perceived stress and gestational age at delivery.
The prematurity rate in this cohort was consistent with state and national norms31 and hair cortisol levels rose across pregnancy similar to what has been published previously.15 Women who delivered preterm had a higher mean second-trimester hair cortisol concentration than those who delivered at term. Psychological measures of stressors and mood were highly correlated both across measures and across gestation. Our findings of a relationship between second-, but not third-, trimester psychological and physiologic stress and gestational age at delivery are congruent with previous reports indicating a relationship between second-trimester plasma cortisol and gestational age at delivery.4,9 However, this is the first report of a marker of chronic hypothalamic–pituitary–adrenal physiology replicating these data and also of testing hair cortisol concentration as a statistical mediator of these relationships.
In the second trimester, but not the first or third, hair cortisol mediates the relationship between perceived stress and gestational age at delivery. Early in the first trimester, the embryo may be relatively protected from the maternal psychobiologic environment, including the influence of the maternal hypothalamic–pituitary–adrenal axis, as a result of a small placental volume, low maternal blood flow to the placenta, and low oxygen tension within the fetoplacental unit.32 Conversely, in the third trimester, maternal cortisol levels are significantly increased but levels of the cortisol metabolizing enzyme, 11β-hydroxysteroid dehydrogenase type 2, are concomitantly increased and there is an overall blunted maternal response to late pregnancy cortisol in preparation for delivery.15,33,34
Sandman et al9 reported higher plasma cortisol levels at 15 weeks of gestation in women who delivered preterm compared with those who delivered at term. The elevation seen in plasma cortisol at 15 weeks of gestation predicted elevated placental corticotrophin-releasing hormone in the third trimester, suggesting a stressor signaling pattern initiated during the second trimester that might influence earlier delivery.9 This cortisol pattern is also observed in second-trimester hair cortisol concentration in our present study.
What has been proposed but has never been directly tested is how physiologic stress relates to psychological stress and, in combination, how these two measures correlate with gestational age at delivery and preterm birth. This study is unique in that a longitudinal study design was used providing repeated measures at multiple time points, allowing sufficient data to perform a mediation analysis. The strengths of this study are prospectively collected psychological and physiologic data, multiple time points of assessment over the course of pregnancy, and a noninvasive biomarker of long-term retrospective hypothalamic–pituitary–adrenal status collected to represent a single trimester of maternal hypothalamic–pituitary–adrenal activity. Limitations of this study include a relative lack of racial diversity, collection from a single study site, and, as just recently published, an inability to account for the effect of childhood trauma on maternal hair cortisol concentrations.35
In summary, early pregnancy physiologic stress is associated with psychological stress and both are correlated with gestational age at delivery. These findings have biological plausibility and are consistent with health disparities data that relate chronic stress, dysregulated hypothalamic–pituitary–adrenal function, and adverse mental and physical health outcomes including preterm birth.36–38 This information advances our understanding of hypothalamic–pituitary–adrenal function in the presence of second-trimester perceived stress, elucidating one plausible contributor to preterm birth. Because stress remains a significant risk factor for preterm birth, its recognition and amelioration during early pregnancy deserve further attention and investigation.
1. Manuck TA, Esplin MS, Biggio J, Bukowski R, Parry S, Zhang H, et al.. The phenotype of spontaneous preterm birth: application of a clinical phenotyping tool. Am J Obstet Gynecol 2015;212:487.e1–487.e11.
2. Coussons-Read ME, Lobel M, Carey JC, Kreither MO, D'Anna K, Argys L, et al.. The occurrence of preterm delivery is linked to pregnancy-specific distress and elevated inflammatory markers across gestation. Brain Behav Immun 2012;26:650–9.
3. Wadhwa PD, Entringer S, Buss C, Lu MC. The contribution of maternal stress to preterm birth: issues and considerations. Clin Perinatol 2011;38:351–84.
4. Diego MA, Jones NA, Field T, Hernandez-Reif M, Schanberg S, Kuhn C, et al.. Maternal psychological distress, prenatal cortisol, and fetal weight. Psychosom Med 2006;68:747–53.
5. Glynn LM, Davis EP, Sandman CA. New insights into the role of perinatal HPA-axis dysregulation in postpartum depression. Neuropeptides 2013;47:363–70.
6. Davis EP, Sandman CA. Prenatal psychobiological predictors of anxiety risk in preadolescent children. Psychoneuroendocrinology 2012;37:1224–33.
7. Davis EP, Sandman CA. The timing of prenatal exposure to maternal cortisol and psychosocial stress is associated with human infant cognitive development. Child Dev 2010;81:131–48.
8. Glover V. Prenatal stress and its effects on the fetus and the child: possible underlying biological mechanisms. Adv Neurobiol 2015;10:269–83.
9. Sandman CA, Glynn L, Schetter CD, Wadhwa P, Garite T, Chicz-DeMet A, et al.. Elevated maternal cortisol early in pregnancy predicts third trimester levels of placental corticotropin releasing hormone (CRH): priming the placental clock. Peptides 2006;27:1457–63.
10. Diego MA, Field T, Hernandez-Reif M, Schanberg S, Kuhn C, Gonzalez-Quintero VH. Prenatal depression restricts fetal growth. Early Hum Dev 2009;85:65–70.
11. Gozansky WS, Lynn JS, Laudenslager ML, Kohrt WM. Salivary cortisol determined by enzyme immunoassay is preferable to serum total cortisol for assessment of dynamic hypothalamic–pituitary–adrenal axis activity. Clin Endocrinol (Oxf) 2005;63:336–41.
12. Vanaelst B, Huybrechts I, Bammann K, Michels N, de Vriendt T, Vyncke K, et al.. Intercorrelations between serum, salivary, and hair cortisol and child-reported estimates of stress in elementary school girls. Psychophysiology 2012;49:1072–81.
13. Yokokawa A, Takasaka T, Shibasaki H, Kasuya Y, Kawashima S, Yamada A, et al.. The effect of water loading on the urinary ratio of cortisone to cortisol in healthy subjects and a new approach to the evaluation of the ratio as an index for in vivo human 11beta-hydroxysteroid dehydrogenase 2 activity. Steroids 2012;77:1291–7.
14. Meyer JS, Novak MA. Minireview: Hair cortisol: a novel biomarker of hypothalamic-pituitary-adrenocortical activity. Endocrinology 2012;153:4120–7.
15. D'Anna-Hernandez KL, Ross RG, Natvig CL, Laudenslager ML. Hair cortisol levels as a retrospective marker of hypothalamic-pituitary axis activity throughout pregnancy: comparison to salivary cortisol. Physiol Behav 2011;104:348–53.
16. Braig S, Grabher F, Ntomchukwu C, Reister F, Stalder T, Kirschbaum C, et al.. Determinants of maternal hair cortisol concentrations at delivery reflecting the last trimester of pregnancy. Psychoneuroendocrinology 2015;52:289–96.
17. Yamada J, Stevens B, de Silva N, Gibbins S, Beyene J, Taddio A, et al.. Hair cortisol as a potential biologic marker of chronic stress in hospitalized neonates. Neonatology 2007;92:42–9.
18. Hoffman MC, Karban LV, Benitez P, Goodteacher A, Laudenslager ML. Chemical processing and shampooing impact cortisol measured in human hair. Clin Invest Med 2014;37:E252–7.
19. Robinson DP, Klein SL. Pregnancy and pregnancy-associated hormones alter immune responses and disease pathogenesis. Horm Behav 2012;62:263–71.
20. Pols H. August Hollingshead and Frederick Redlich: poverty, socioeconomic status, and mental illness. Am J Public Health 2007;97:1755.
21. Wosu AC, Gelaye B, Valdimarsdóttir U, Kirschbaum C, Stalder T, Shields AE, et al.. Hair cortisol in relation to sociodemographic and lifestyle characteristics in a multiethnic US sample. Ann Epidemiol 2015;25:90–5, 95.e1–2.
22. Wennig R. Potential problems with the interpretation of hair analysis results. Forensic Sci Int 2000;107:5–12.
23. Cohen S, Kamarck T, Mermelstein R. A global measure of perceived stress. J Health Soc Behav 1983;24:385–96.
24. Radloff LS. The use of the Center for Epidemiologic Studies Depression Scale in adolescents and young adults. J Youth Adolesc 1991;20:149–66.
25. Spielberger CD, Gorssuch RL, Lushene PR, Vagg PR, Jacobs GA. Manual for the state-trait anxiety inventory. Palo Alto (CA): Consulting Psychologists Press; 1983.
26. Kalra S, Einarson A, Karaskov T, Van Uum S, Koren G. The relationship between stress and hair cortisol in healthy pregnant women. Clin Invest Med 2007;30:E103–7.
27. Nicholson WK, Setse R, Hill-Briggs F, Cooper LA, Strobino D, Powe NR. Depressive symptoms and health-related quality of life in early pregnancy. Obstet Gynecol 2006;107:798–806.
28. Tamaki R, Murata M, Okano T. Risk factors for postpartum depression in Japan. Psychiatry Clin Neurosci 1997;51:93–8.
29. Hollingshead AB, Redlich FC. Social class and mental illness: a community study. 1958. Am J Public Health 2007;97:1756–7.
30. MacKinnon DP, Dwyer JH. Estimating mediated effects in prevention studies. Eval Rev 1993;17:144–58.
32. Burton GJ, Jaunaiux E. Maternal vascularisation of the human placenta: does the embryo develop in a hypoxic environment? Gynecol Obstet Fertil 2001;29:503–8.
33. Mastorakos G, Ilias I. Maternal and fetal hypothalamic-pituitary-adrenal axes during pregnancy and postpartum. Ann N Y Acad Sci 2003;997:136–49.
34. Murphy VE, Clifton VL. Alterations in human placental 11beta-hydroxysteroid dehydrogenase type 1 and 2 with gestational age and labour. Placenta 2003;24:739–44.
35. Schreier HM, Enlow MB, Ritz T, Gennings C, Wright RJ. Childhood abuse is associated with increased hair cortisol levels among urban pregnant women. J Epidemiol Community Health 2015;69:1169–74.
36. Blumenshine P, Egerter S, Barclay CJ, Cubbin C, Braveman PA. Socioeconomic disparities in adverse birth outcomes: a systematic review. Am J Prev Med 2010;39:263–72.
37. Jackson JS, Knight KM, Rafferty JA. Race and unhealthy behaviors: chronic stress, the HPA axis, and physical and mental health disparities over the life course. Am J Public Health 2010;100:933–9.
© 2016 by The American College of Obstetricians and Gynecologists. Published by Wolters Kluwer Health, Inc. All rights reserved.
38. Berger M, Sarnyai Z. “More than skin deep”: stress neurobiology and mental health consequences of racial discrimination. Stress 2015;18:1–10.