- Email: [email protected]
Association between maternal and amniotic fluid cortisol is moderated by maternal anxiety
Psychoneuroendocrinology (2009) 34, 430—435
a v a i l a b l e a t w w w. s c i e n c e d i r e c t . c o m
j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / p s y n e u e n
Association between maternal and amniotic fluid cortisol is moderated by maternal anxiety Vivette Glover a,*, Kristin Bergman a, Pampa Sarkar a,b,c, Thomas G. O’Connor d a
Institute of Reproductive and Developmental Biology, Imperial College London, Hammersmith Campus, Du Cane Road, London W12 0NN, UK b Department of Obstetrics and Gynaecology, Wexham Park Hospital, Slough SL2 4HL,UK c Centre for Fetal Care, Queen Charlotte’s and Chelsea Hospital, Du Cane Road, London W12 0HS, UK d Department of Psychiatry, University of Rochester Medical Center, 300 Crittenden Boulevard, Rochester, NY 14642, USA Received 15 January 2008; received in revised form 3 October 2008; accepted 9 October 2008
KEYWORDS Cortisol; Anxiety; Stress; Amniotic fluid; Pregnancy; 11b-HSD2
Summary Maternal stress or anxiety during pregnancy can lead to neurodevelopmental and other problems in the child, and cortisol is one possible mediator. Animal models show that maternal prenatal stress can affect placental function, including regulation of placental 11bHSD2, the main barrier to the placental passage of cortisol. It is not known whether a parallel process exists in humans. The aim of the current study was to determine whether maternal anxiety increases the association between maternal plasma cortisol and amniotic fluid cortisol. The sample consisted of 262 women having amniocentesis, with normal pregnancies, who completed Spielberger State and Trait anxiety scales, from whom a plasma sample and an aliquot of amniotic fluid was obtained. The correlation between maternal and amniotic fluid cortisol was strongly dependent on both State and Trait maternal anxiety; in the most anxious State quartile r(62) = .59, p < .001 and in the least r(60) = .05, ns, a significant difference ( p < .0015). The moderating effect of maternal anxiety on the association between maternal plasma and amniotic fluid cortisol remained when gestational age, maternal age, fetal sex, medication and time of collection were controlled for. There was no difference in amniotic fluid cortisol levels between the most and least anxious groups of mothers. However, the finding that there is a stronger correlation between maternal and fetal cortisol among more anxious pregnant women does suggests that the maternal emotional state can affect the function of the placenta. # 2008 Elsevier Ltd. All rights reserved.
1. Introduction
* Corresponding author. Tel.: +44 207 594 2136; fax: +44 207 594 2138. E-mail address: [email protected] (V. Glover).
There is good evidence from animal studies, and increasing evidence in humans, that prenatal stress can have long termeffects on the child, including on neurodevelopment (Van den Bergh et al., 2005; Talge et al., 2007). The increased risk for
0306-4530/$ — see front matter # 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.psyneuen.2008.10.005
Antenatal anxiety and maternal and amniotic fluid cortisol correlation child cognitive and behavioral problems is found with moderate elevations of anxiety and in non-selected samples e.g., (O’Connor et al., 2002; O’Connor et al., 2003), and is independent of postnatal anxiety and depression (O’Connor et al., 2002; Bergman et al., 2007). In rodent models it has been shown that the effects of prenatal stress on the offspring are at least partly mediated via the hypothalamic— pituitary—adrenal (HPA) axis, particularly exposure to glucocorticoids (Weinstock, 2001; Herbert et al., 2006; Seckl and Holmes, 2007). Non-human primate models also point to the mediating role of the HPA axis (Schneider et al., 2002). The potentially widespread role for exposure to increased cortisol in human fetal brain development has been strengthened by a study showing, by microarray analysis, that increased cortisol exposure affects the expression of over a thousand genes in fetal brain cells (Salaria et al., 2006). It is, however, less clear that cortisol, or the function of the maternal HPA axis, mediates the long-term effects of antenatal maternal stress on the fetus in humans. Only modest associations between maternal antenatal anxiety and maternal cortisol levels have been reported (Sarkar et al., 2006; Obel et al., 2005) and there is some evidence that maternal prenatal stress/anxiety predicts infant outcomes independent of maternal prenatal cortisol levels (Davis et al., 2007; Bergman, Sarkar, O’Connor and Glover unpublished observations). Also, the maternal HPA axis becomes hypo-responsive to stress as gestation increases (Schulte et al., 1990; Kammerer et al., 2002) and at least in some studies, this is the stage at which antenatal stress has the greatest effect on fetal development (O’Connor et al., 2002). Perturbations in the maternal environment must be transmitted across the placenta in order to affect the fetus, and it is possible that alterations in the function of the placenta may play an important part in regulating the exposure of the fetus to maternal hormone levels, independently of a direct increase in maternal cortisol (Jansson and Powell, 2007). Two recent rodent studies have shown that maternal prenatal stress can affect the function of the placenta, including the expression and activity of 11beta-hydroxysteroid dehydrogenase 2 (11b-HSD2) (Welberg et al., 2005; Mairesse et al., 2007). This is the enzyme that metabolises cortisol to the inactive cortisone in the placenta, and thus regulates maternal to fetal transfer of glucocorticoids. If the enzyme is downregulated by prenatal stress, as shown in rodents by Mairesse et al. (2007), one would expect more cortisol to pass from mother to fetus, and possibly contribute to fetal programming effects. In humans lower placental levels of 11bHSD2. (Kajantie et al., 2003; Dy et al., 2008), and mutations in the gene encoding it (Seckl and Holmes, 2007) have been found to be associated with intrauterine growth restriction, possibly due to increased fetal exposure to cortisol. In the current study we aimed to translate the rodent study of Mairesse et al. (2007) as part of a series of studies on prenatal anxiety/stress and child development (Sarkar et al., 2006, 2007a,b; Bergman et al., 2007). We have previously reported a significant association between maternal plasma and amniotic fluid cortisol, which increases with gestational age (Sarkar et al., 2007a). Here we have tested the hypothesis that the association between amniotic fluid and maternal plasma cortisol is increased with maternal anxiety and is greater among more anxious mothers. We also determined
431
whether amniotic fluid cortisol was raised in the more anxious mothers.
2. Subjects and methods This study was conducted at Queen Charlotte’s and Chelsea Hospital, London and involved obtaining maternal plasma and amniotic fluid samples from pregnant women undergoing amniocentesis for karyotyping to identify chromosome anomalies, the majority for Down’s syndrome. A total of 365 subjects with a gestational age range of 15—37 weeks were recruited between December 2001 and December 2004. All women gave written informed consent as approved by the institutional ethics committee. All subjects underwent preliminary ultrasound assessment. Gestational age was established to the nearest day by plotting ultrasound-derived fetal biometry. Fetal sex was determined on karyotype. Only samples from singleton pregnancies with a structurally normal fetus and appropriate fetal growth on ultrasound were included in the study. Exclusion criteria were fetal structural abnormalities, samples visually contaminated with blood, and those subsequently associated with fetal aneuploidy or fetal death. Of the 365 patients recruited, 48 were excluded as above. Of the remaining 317 cases, amniotic fluid cortisol data were available on n = 311 cases and maternal plasma data were available on n = 271 cases. Self-rating Spielberger questionnaires (Spielberger et al., 1970) were given to the women to complete in the 15 min prior to amniocentesis, to assess their anxiety at the time ‘‘how you feel right now’’ (State) and their habitual anxiety ‘‘how you generally feel’’ (Trait). The final study population comprised 262 cases with complete data on maternal and fetal cortisol and anxiety ratings. Information on key covariates, time of collection, gestational age, fetal sex, and maternal age were also recorded. Eighteen of the women were taking some type of medication (SSRI, 1; antihypertensive, 2; antiasthma, 10; antiepileptic 1; steroids, 1; other, 3). Maternal blood was drawn immediately before the procedure, centrifuged and supernatant plasma stored at 80 8C until batch assay. During amniocentesis an additional aliquot of up to 4 ml of amniotic fluid surplus to clinical requirement was drawn for the study and stored at 80 8C until assay. Time of collection was recorded to the nearest 15 min. Total cortisol in amniotic fluid was assayed by radio-immunoassay (Coat-A-Count, DPC, Los Angeles, CA), cortisol having been extracted by dichloromethane and reconstituted prior to assay (Sarkar et al., 2007a,b). Our intra- and inter-assay coefficients of variation for the cortisol assay in amniotic fluid were 4.4% and 6.5% respectively. Some of the cortisol data has been reported previously (Sarkar et al., 2006, 2007a,b).
2.1. Data analysis Normality of distribution of data was checked in a P-P plot. Non-normally distributed variables underwent natural log (ln) transformation, and normal distribution was attained for maternal and amniotic fluid cortisol but not for gestational age. Statistical analyses were performed using ln data, which are substantially identical to non-transformed data.
432
V. Glover et al.
Maternal age, fetal sex, gestational age, and time of collection were included as covariates in the analyses.
Table 1 Clinical characteristics of women undergoing amniocentesis. Data shown as mean (SD) [range] n = 262.
3. Results
Maternal age (years) Gestation (weeks) Time of collection
The characteristics of the sample are shown in Table 1. It may be seen that the cortisol levels in maternal plasma were about 28 fold higher than those in the amniotic fluid. There was a positive correlation between amniotic fluid cortisol with maternal plasma levels (r = .32, p < .001), in the group as a whole. Fig. 1 shows the correlations between maternal plasma and amniotic fluid cortisol in the mothers according to maternal State anxiety quartiles. In the most anxious group the correlation was r(62) = .59, p < .001; there was a substantial and significant decrease in the magnitude of the correlation in the less anxious quartiles, in stepped fashion, with the correlation falling to r(60) = .05, ns in the least anxious group (see figure legend for significant contrasts). Subsequent analyses that covaried gestational age, time of amniotic fluid collection, fetal sex, and maternal age and medication yielded substantively similar results. For example, the correlation between maternal plasma cortisol and amniotic fluid cortisol was .51 for the most anxious women and .01 for the least anxious women (the difference between correlations was significant at p < .01). A comparable change in the magnitude of the correlation between amniotic fluid and plasma cortisol was also found when we considered Trait anxiety (Fig. 2). Thus, the correlation between maternal plasma cortisol and amniotic fluid
35.3 (5.4) [18—47] 17.7 (3.6) [15—37] 11.30 (1.4) [9.05—17.30]
Ethnicity, number (%) Caucasian 189 (72.1) Indian/subcontinent 23 (8.8) Black 21 (8) Middle eastern 15 (5.7) Asian-far eastern 10 (1.5) Other 4 (1.5) Spielberger State Anxiety 49.7 (13.6) [20—77] Spielberger Trait Anxiety 36.3 (8.8) [21—63] Maternal cortisol (nmol/l) 579 (240) [157—1770] Amniotic fluid cortisol (nmol/l) 20.6 (12.5) [.66—100]
cortisol was r(67) = .49, p < .001 among women in the top quartile of Trait anxiety, whereas the correlation was r(64) = .13 among women in the least anxious quartile. These patterns also held after adjusting for gestational age, time of amniotic fluid collection, fetal sex, medication and maternal age. The correlation between State and Trait anxiety in the whole sample was moderate, r(262) = .35, p < .0001), implying that the stepped decrease in the link between maternal plasma and amniotic fluid cortisol according to maternal anxiety across the two measures, is noteworthy and not redundant.
Figure 1 Bivariate correlations between ln maternal plasma (nmol/l) and ln amniotic fluid (AF) cortisol (nmol/l) according to maternal State anxiety (quartiles). (a) Quartile 1. Speilberger State anxiety 20-38: r(60) = .05; (b) Quartile 2. Speilberger State anxiety 39—49: r(71) = .29; (c) Quartile 3 Speilberger State anxiety 50—61: r(69) = .40; (d) Quartile 4. Speilberger State anxiety 62—77: r(62) = .59. Correlations were significantly different (using Fisher r to z transformation) between Quartiles 1 and 3 ( p < .05) and Quartiles 1 and 4 ( p < .001) and Quartiles 2 and 4 ( p < .05).
Antenatal anxiety and maternal and amniotic fluid cortisol correlation
433
Figure 2 Bivariate correlations between ln maternal plasma (nmol/l) and ln amniotic fluid (AF) cortisol (nmol/l) according to maternal Trait anxiety (quartiles). (a) Quartile 1. Speilberger Trait anxiety 21—29: r(64) = .13; (b) Quartile 2. Speilberger Trait anxiety 30—34: r(64) = .37, p < .01; (c) Quartile 3. Speilberger Trait anxiety 35—41: r(67) = .40, p < .001; (d) Quartile 4. Speilberger Trait anxiety 42—63: r(67) = .49, p < .001. Correlations between Quartiles 1 and 4 were significantly different (using Fisher r to z transformation) at p < .05.
We next examined whether the level of amniotic fluid cortisol varied significantly as a function of maternal State anxiety. We found no evidence of a significant correlation (r(262) = .09, ns); Table 2 shows the raw (untransformed) means (SD) for the amniotic fluid cortisol for mothers in the least to most state anxiety quartiles, differences that fell short of statistical significance (F(3,258) = .46); the same pattern was observed with the transformed amniotic fluid cortisol data. In contrast, there was a significant difference in maternal plasma cortisol among women in the State anxiety quartiles (from least to most anxious quartiles): F(3,258) = 4.04, p < .01), although the bivariate correlation was modest r(262) = .20, p < .01. There were no significant differences in maternal plasma or amniotic fluid cortisol according to maternal Trait anxiety quartile. Table 2 Cortisol levels (nmol/l), mean (SD), in maternal plasma and amniotic fluid by State and Trait Anxiety quartiles. Maternal plasma
Amniotic fluid
State anxiety Least anxious (n = 60) 2nd quartile (n = 71) 3rd quartile (n = 69) Most anxious (n = 62)
505 556 612 640
(175) (212) (233) (306)
18.9 21.0 21.3 20.9
(12.6) (12.9) (9.3) (14.9)
Trait anxiety Least anxious (n = 64) 2nd quartile (n = 64) 3rd quartile (n = 67) Most anxious (n = 67)
541 570 608 596
(202) (213) (265) (269)
20.4 20.7 21.0 20.2
(12.4) (10.3) (10.6) (16.0)
3.1. Supplementary analysis Results so far indicated that maternal anxiety may be associated with altered placental function, indexed by the moderating effect of maternal anxiety on the association between maternal plasma and amniotic fluid cortisol, but not increased exposure of the fetus to cortisol. A set of supplemental analyses were carried out to examine further this lack of association between maternal anxiety and amniotic fluid cortisol and to examine differences according to fetal sex. Based on our previous findings that the association between maternal plasma cortisol and amniotic fluid cortisol was greater at later gestation age (Sarkar et al., 2007a,b), exploratory analyses examined if a significant link between prenatal anxiety and amniotic fluid cortisol was evident at later gestational age; it was not, and neither did time of collection appear to influence the link between prenatal anxiety and amniotic fluid cortisol. We next considered if the moderating effect of maternal anxiety on the link between maternal plasma and amniotic fluid cortisol was different in male and female fetuses. The correlation between maternal plasma and amniotic fluid cortisol was comparable in females (r(130) = .37, p < .001) and males (r(132) = .30, p < .001); there was no evidence that the moderating effect of maternal anxiety on the link between maternal plasma and amniotic fluid cortisol differed according to fetal sex.
4. Discussion This study is the first to examine the possible effects of maternal mood on placental function in humans. It supports
434 the hypothesis that if the mother is anxious while pregnant then there is a greater association between her cortisol level and that of her fetus, and suggests that the emotional state of the mother may have a direct effect on placental function. There was a similar stepped relationship between maternal/ amniotic fluid cortisol correlation and both State and Trait anxiety (Figs. 1 and 2), implying a robust effect of maternal anxiety on the correlation. Comparable results for Trait (how the individual generally feels) and State (how the person feels now) anxiety is noteworthy. We have previously reported that State anxiety is considerably elevated in these women awaiting amniocentesis, and this can cause a modest increase in their cortisol (Sarkar et al., 2006). However Trait anxiety is not raised in these women, and this together with the dose—response pattern shown in the figures, imply that this effect is not limited to clinical extremes. There is some evidence that placental glucocorticoid metabolism can be sexually dimorphic (Clifton, 2005). However, there was no significant difference in the moderating effect of maternal anxiety on maternal/amniotic fluid cortisol correlation dependent on the sex of the fetus. The results presented here are compatible with a down regulation of placental 11b-HSD2, as found in the prenatal stress rodent model of Mairesse et al. (2007). However, obviously our results provide only indirect evidence for such a down regulation. It should also be noted that Welberg et al. (2005), also using a rodent model, found that acute stress up regulated the placental enzyme, and that chronic stress reduced this up regulation, although there are several differences between the experimental designs of these two studies, in type of stress and timing, which may be relevant. The importance of placental 11b-HSD2 in determining outcome for the offspring, independent of maternal HPA axis function, has been shown by the experiments of Holmes et al. (2006). Using knock out mice for 11b-HSD2 they demonstrated the key role of fetoplacental 11b-HSD2 in glucocorticoid programming in the offspring, resulting in both lower birthweight and increased anxiety. The placenta clearly plays a crucial role in moderating fetal exposure to maternal factors, and presumably in preparing the fetus for the environment in which it is going to find itself (Gluckman et al., 2005). One notable aspect of our results was that there was no difference in the cortisol level in amniotic fluid from the most and least anxious women. That is, the change in maternal and amniotic fluid cortisol correlation, associated with maternal anxiety, was apparently not directly linked with increased fetal cortisol exposure. How that is reconciled with the strong moderation effects is not certain. It might be that the ‘‘signal strength’’, the correlation of r = .09 between maternal anxiety and amniotic fluid cortisol, is muted by other intermediating processes. It is notable that Mairesse et al. (2007) also found no difference in the corticosterone level in the rat fetus despite direct evidence for down regulation of both 11b-HSD2 expression and activity. However ACTH levels and adrenal size were significantly reduced in the fetuses from the prenatally stressed dams. These authors suggest that these changes in the fetal HPA axis are adjustments made by the fetus in response to prior increased exposure to maternal glucocorticoids. It is of interest that in the offspring of 11b-HSD2 knockout mice (Holmes et al., 2006) there was also a hypotrophy of the adrenal glands. Thus the contribution by the fetal adrenal
V. Glover et al. gland to the amniotic fluid cortisol may be reduced, despite an increased contribution from the mother, leaving the overall level unchanged. We were unable to examine either ACTH or fetal adrenal volume in our study. It will clearly be important to study further the effect of both acute and chronic maternal anxiety on placental 11b-HSD2, both by direct assay of the placental enzyme activity and expression, and by measuring cord blood or amniotic fluid cortisol/cortisone ratio. It is also important to note that our studies were carried out at a median gestational age of 17 weeks. It is possible that there is a stronger association between maternal anxiety and amniotic fluid cortisol later in gestation, when some studies suggest maternal anxiety has a stronger effect on fetal programming (e.g. O’Connor et al., 2002). The question arises of what may be the mediator between maternal anxiety and a putative down regulation of 11bHSD2, if not increased maternal cortisol. The peripheral sympathetic system is a plausible candidate; it is possible that the sympathetic system remains more responsive to anxiety or stress during pregnancy than the HPA axis. In a rat model it has been demonstrated that whereas the HPA axis becomes unreactive to an acute stressor as pregnancy progresses, the reactivity of the peripheral sympathetic system is unaffected (Douglas personal communication). Noradrenaline has been shown to down regulate 11b-HSD2 in isolated placental human trophoblast cells (Sarkar et al., 2001). There are several possible sources of amniotic fluid cortisol, apart from the maternal, including the fetal adrenal, and 11b-HSD1 in the placenta and the fetal membranes which converts cortisone to cortisol. These mature at different stages of gestation, but the effect of maternal anxiety on the amniotic fluid/maternal cortisol correlation was little affected by covarying gestational age. We do not know how long the acute anxiety experienced while awaiting amniocentesis lasted, or how long maternal cortisol takes to cross the placenta. However covarying time of amniocentesis did not affect the effect of anxiety on the correlation. The results presented here may have a variety of clinical implications. For example, prednisolone is used to treat several maternal diseases during pregnancy including asthma, hyperemesis gravidarium, and rheumatoid arthritis. It is metabolised by 11-b HSD2 in a similar manner to cortisol, and is generally considered to be safe for that reason. This study suggests the possibility that the level of prednisolone exposure to the fetus may be greater for more anxious mothers. Additionally, these findings underscore the value that clinical interventions for pregnant anxious women may have for both understanding basic mechanisms of fetal programming and for preventive medicine. Although our results provide only indirect evidence for an alteration in 11b-HSD2, they do provide the first robust human evidence that placental function may be affected by maternal anxiety. Accordingly, the findings are an important step in translating key animal findings and in understanding the mechanisms of fetal programming.
Role of the funding sources Funding for this study was provided by March of Dimes; March of Dimes had no further role in study design; in the collection, analysis and interpretation of data; in the writing of
Antenatal anxiety and maternal and amniotic fluid cortisol correlation the report; and in the decision to submit the paper for publication.
Conflict of interest None of the authors have any actual or potential conflict of interest related to the submitted manuscript.
Acknowledgement We should like to thank March of Dimes for funding and Diana Adams for help with the recruitment of the subjects, scoring of the questionnaires and data collection, and Kieran O’Donnell for helpful discussion of the findings.
References Bergman, K., Sarkar, P., O’Connor, T.G., Modi, N., Glover, V., 2007. Maternal stress during pregnancy predicts cognitive ability and fearfulness in infancy. J. Am. Acad. Child Adolesc. Psychiatry 46, 1454—1463. Clifton, V.L., 2005. Sexually dimorphic effects of maternal asthma during pregnancy on placental glucocorticoid metabolism and fetal growth. Cell Tissue Res. 322, 63—71. Davis, E.P., Glynn, L.M., Schetter, C.D., Hobel, C., Chicz-Demet, A., Sandman, C.A., 2007. Prenatal exposure to maternal depression and cortisol influences infant temperament. J. Am. Acad. Child Adolesc. Psychiatry 46, 737—746. Dy, J., Guan, H., Sampath-Kumar, R., Richardson, B.S., Yang, K., 2008. Placental 11beta-hydroxysteroid dehydrogenase type 2 is reduced in pregnancies complicated with idiopathic intrauterine growth Restriction: evidence that this is associated with an attenuated ratio of cortisone to cortisol in the umbilical artery. Placenta 29, 193—200. Gluckman, P.D., Hanson, M.A., Spencer, H.G., 2005. Predictive adaptive responses and human evolution. Trends Ecol. Evol. 20, 527—533. Herbert, J., Goodyer, I.M., Grossman, A.B., Hastings, M.H., de Kloet, E.R., Lightman, S.L., Lupien, S.J., Roozendaal, B., Seckl, J.R., 2006. Do corticosteroids damage the brain? J. Neuroendocrinol. 18, 393—411. Holmes, M.C., Abrahamsen, C.T., French, K.L., Paterson, J.M., Mullins, J.J., Seckl, J.R., 2006. The mother or the fetus? 11beta-hydroxysteroid dehydrogenase type 2 null mice provide evidence for direct fetal programming of behavior by endogenous glucocorticoids. J. Neurosci. 26, 3840—3844. Jansson, T., Powell, T.L., 2007. Role of the placenta in fetal programming: underlying mechanisms and potential interventional approaches. Clin. Sci. (Lond.) 113, 1—13. Kajantie, E., Dunkel, L., Turpeinen, U., Stenman, U.H., Wood, P.J., Nuutila, M., Andersson, S., 2003. Placental 11 beta-hydroxysteroid dehydrogenase-2 and fetal cortisol/cortisone shuttle in small preterm infants. J. Clin. Endocrinol. Metab. 88, 493—500. Kammerer, M., Adams, D., Castelberg Bv, B., Glover, V., 2002. Pregnant women become insensitive to cold stress. BMC Pregnancy Childbirth 2, 8. Mairesse, J., Lesage, J., Breton, C., Breant, B., Hahn, T., Darnaudery, M., Dickson, S.L., Seckl, J., Blondeau, B., Vieau, D., Maccari, S., Viltart, O., 2007. Maternal stress alters endocrine function of the
435
feto-placental unit in rats. Am. J. Physiol. Endocrinol. Metab. 292, E1526—E1533. O’Connor, T.G., Heron, J., Golding, J., Beveridge, M., Glover, V., 2002. Maternal antenatal anxiety and children’s behavioural/ emotional problems at 4 years Report from the Avon Longitudinal Study of Parents and Children. Br. J. Psychiatry 180, 502—508. O’Connor, T.G., Heron, J., Golding, J., Glover, V., 2003. Maternal antenatal anxiety and behavioural/emotional problems in children: a test of a programming hypothesis. J. Child. Psychol. Psychiatry 44, 1025—1036. Obel, C., Hedegaard, M., Henriksen, T.B., Secher, N.J., Olsen, J., Levine, S., 2005. Stress and salivary cortisol during pregnancy. Psychoneuroendocrinology 30, 647—656. Salaria, S., Chana, G., Caldara, F., Feltrin, E., Altieri, M., Faggioni, F., Domenici, E., Merlo-Pich, E., Everall, I.P., 2006. Microarray analysis of cultured human brain aggregates following cortisol exposure: implications for cellular functions relevant to mood disorders. Neurobiol. Dis. 23, 630—636. Sarkar, P., Bergman, K., Fisk, N.M., Glover, V., 2006. Maternal anxiety at amniocentesis and plasma cortisol. Prenat. Diagn. 26, 505— 509. Sarkar, P., Bergman, K., Fisk, N.M., O’Connor, T.G., Glover, V., 2007a. Ontogeny of foetal exposure to maternal cortisol using midtrimester amniotic fluid as a biomarker. Clin. Endocrinol. (Oxf.) 66, 636—640. Sarkar, P., Bergman, K., Fisk, N.M., O’Connor, T.G., Glover, V., 2007b. Amniotic fluid testosterone: relationship with cortisol and gestational age. Clin. Endocrinol. (Oxf.) 67, 743—747. Sarkar, S., Tsai, S.W., Nguyen, T.T., Plevyak, M., Padbury, J.F., Rubin, L.P., 2001. Inhibition of placental 11beta-hydroxysteroid dehydrogenase type 2 by catecholamines via alpha-adrenergic signalling. Am. J. Physiol. Regul. Integr. Comp. Physiol. 281, R1966— R1974. Schneider, M.L., Moore, C.F., Kraemer, G.W., Roberts, A.D., DeJesus, O.T., 2002. The impact of prenatal stress, fetal alcohol exposure, or both on development: perspectives from a primate model. Psychoneuroendocrinology 27, 285—298. Schulte, H.M., Weisner, D., Allolio, B., 1990. The corticotrophin releasing hormone test in late pregnancy: lack of adrenocorticotrophin and cortisol response. Clin. Endocrinol. (Oxf.) 33, 99— 106. Seckl, J.R., Holmes, M.C., 2007. Mechanisms of disease: glucocorticoids, their placental metabolism and fetal ‘programming’ of adult pathophysiology. Nat. Clin. Pract. Endocrinol. Metab. 3, 479—488. Spielberger, C.D., Gorusch, R.L., Rushene, R.E. (Eds.), 1970. STAI Manual for The State-Trait Anxiety Inventory. Consulting Psychologists Press Inc., Pao Alto, CA. Talge, N.M., Neal, C., Glover, V., 2007. Antenatal maternal stress and long-term effects on child neurodevelopment: how and why? J. Child Psychol. Psychiatry 48, 245—261. Van den Bergh, B.R., Mulder, E.J., Mennes, M., Glover, V., 2005. Antenatal maternal anxiety and stress and the neurobehavioural development of the fetus and child: links and possible mechanisms. A review. Neurosci. Biobehav. Rev. 29, 237—258. Weinstock, M., 2001. Alterations induced by gestational stress in brain morphology and behaviour of the offspring. Prog. Neurobiol. 65, 427—451. Welberg, L.A., Thrivikraman, K.V., Plotsky, P.M., 2005. Chronic maternal stress inhibits the capacity to up-regulate placental 11beta-hydroxysteroid dehydrogenase type 2 activity. J. Endocrinol. 186, R7—R12.
a v a i l a b l e a t w w w. s c i e n c e d i r e c t . c o m
j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / p s y n e u e n
Association between maternal and amniotic fluid cortisol is moderated by maternal anxiety Vivette Glover a,*, Kristin Bergman a, Pampa Sarkar a,b,c, Thomas G. O’Connor d a
Institute of Reproductive and Developmental Biology, Imperial College London, Hammersmith Campus, Du Cane Road, London W12 0NN, UK b Department of Obstetrics and Gynaecology, Wexham Park Hospital, Slough SL2 4HL,UK c Centre for Fetal Care, Queen Charlotte’s and Chelsea Hospital, Du Cane Road, London W12 0HS, UK d Department of Psychiatry, University of Rochester Medical Center, 300 Crittenden Boulevard, Rochester, NY 14642, USA Received 15 January 2008; received in revised form 3 October 2008; accepted 9 October 2008
KEYWORDS Cortisol; Anxiety; Stress; Amniotic fluid; Pregnancy; 11b-HSD2
Summary Maternal stress or anxiety during pregnancy can lead to neurodevelopmental and other problems in the child, and cortisol is one possible mediator. Animal models show that maternal prenatal stress can affect placental function, including regulation of placental 11bHSD2, the main barrier to the placental passage of cortisol. It is not known whether a parallel process exists in humans. The aim of the current study was to determine whether maternal anxiety increases the association between maternal plasma cortisol and amniotic fluid cortisol. The sample consisted of 262 women having amniocentesis, with normal pregnancies, who completed Spielberger State and Trait anxiety scales, from whom a plasma sample and an aliquot of amniotic fluid was obtained. The correlation between maternal and amniotic fluid cortisol was strongly dependent on both State and Trait maternal anxiety; in the most anxious State quartile r(62) = .59, p < .001 and in the least r(60) = .05, ns, a significant difference ( p < .0015). The moderating effect of maternal anxiety on the association between maternal plasma and amniotic fluid cortisol remained when gestational age, maternal age, fetal sex, medication and time of collection were controlled for. There was no difference in amniotic fluid cortisol levels between the most and least anxious groups of mothers. However, the finding that there is a stronger correlation between maternal and fetal cortisol among more anxious pregnant women does suggests that the maternal emotional state can affect the function of the placenta. # 2008 Elsevier Ltd. All rights reserved.
1. Introduction
* Corresponding author. Tel.: +44 207 594 2136; fax: +44 207 594 2138. E-mail address: [email protected] (V. Glover).
There is good evidence from animal studies, and increasing evidence in humans, that prenatal stress can have long termeffects on the child, including on neurodevelopment (Van den Bergh et al., 2005; Talge et al., 2007). The increased risk for
0306-4530/$ — see front matter # 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.psyneuen.2008.10.005
Antenatal anxiety and maternal and amniotic fluid cortisol correlation child cognitive and behavioral problems is found with moderate elevations of anxiety and in non-selected samples e.g., (O’Connor et al., 2002; O’Connor et al., 2003), and is independent of postnatal anxiety and depression (O’Connor et al., 2002; Bergman et al., 2007). In rodent models it has been shown that the effects of prenatal stress on the offspring are at least partly mediated via the hypothalamic— pituitary—adrenal (HPA) axis, particularly exposure to glucocorticoids (Weinstock, 2001; Herbert et al., 2006; Seckl and Holmes, 2007). Non-human primate models also point to the mediating role of the HPA axis (Schneider et al., 2002). The potentially widespread role for exposure to increased cortisol in human fetal brain development has been strengthened by a study showing, by microarray analysis, that increased cortisol exposure affects the expression of over a thousand genes in fetal brain cells (Salaria et al., 2006). It is, however, less clear that cortisol, or the function of the maternal HPA axis, mediates the long-term effects of antenatal maternal stress on the fetus in humans. Only modest associations between maternal antenatal anxiety and maternal cortisol levels have been reported (Sarkar et al., 2006; Obel et al., 2005) and there is some evidence that maternal prenatal stress/anxiety predicts infant outcomes independent of maternal prenatal cortisol levels (Davis et al., 2007; Bergman, Sarkar, O’Connor and Glover unpublished observations). Also, the maternal HPA axis becomes hypo-responsive to stress as gestation increases (Schulte et al., 1990; Kammerer et al., 2002) and at least in some studies, this is the stage at which antenatal stress has the greatest effect on fetal development (O’Connor et al., 2002). Perturbations in the maternal environment must be transmitted across the placenta in order to affect the fetus, and it is possible that alterations in the function of the placenta may play an important part in regulating the exposure of the fetus to maternal hormone levels, independently of a direct increase in maternal cortisol (Jansson and Powell, 2007). Two recent rodent studies have shown that maternal prenatal stress can affect the function of the placenta, including the expression and activity of 11beta-hydroxysteroid dehydrogenase 2 (11b-HSD2) (Welberg et al., 2005; Mairesse et al., 2007). This is the enzyme that metabolises cortisol to the inactive cortisone in the placenta, and thus regulates maternal to fetal transfer of glucocorticoids. If the enzyme is downregulated by prenatal stress, as shown in rodents by Mairesse et al. (2007), one would expect more cortisol to pass from mother to fetus, and possibly contribute to fetal programming effects. In humans lower placental levels of 11bHSD2. (Kajantie et al., 2003; Dy et al., 2008), and mutations in the gene encoding it (Seckl and Holmes, 2007) have been found to be associated with intrauterine growth restriction, possibly due to increased fetal exposure to cortisol. In the current study we aimed to translate the rodent study of Mairesse et al. (2007) as part of a series of studies on prenatal anxiety/stress and child development (Sarkar et al., 2006, 2007a,b; Bergman et al., 2007). We have previously reported a significant association between maternal plasma and amniotic fluid cortisol, which increases with gestational age (Sarkar et al., 2007a). Here we have tested the hypothesis that the association between amniotic fluid and maternal plasma cortisol is increased with maternal anxiety and is greater among more anxious mothers. We also determined
431
whether amniotic fluid cortisol was raised in the more anxious mothers.
2. Subjects and methods This study was conducted at Queen Charlotte’s and Chelsea Hospital, London and involved obtaining maternal plasma and amniotic fluid samples from pregnant women undergoing amniocentesis for karyotyping to identify chromosome anomalies, the majority for Down’s syndrome. A total of 365 subjects with a gestational age range of 15—37 weeks were recruited between December 2001 and December 2004. All women gave written informed consent as approved by the institutional ethics committee. All subjects underwent preliminary ultrasound assessment. Gestational age was established to the nearest day by plotting ultrasound-derived fetal biometry. Fetal sex was determined on karyotype. Only samples from singleton pregnancies with a structurally normal fetus and appropriate fetal growth on ultrasound were included in the study. Exclusion criteria were fetal structural abnormalities, samples visually contaminated with blood, and those subsequently associated with fetal aneuploidy or fetal death. Of the 365 patients recruited, 48 were excluded as above. Of the remaining 317 cases, amniotic fluid cortisol data were available on n = 311 cases and maternal plasma data were available on n = 271 cases. Self-rating Spielberger questionnaires (Spielberger et al., 1970) were given to the women to complete in the 15 min prior to amniocentesis, to assess their anxiety at the time ‘‘how you feel right now’’ (State) and their habitual anxiety ‘‘how you generally feel’’ (Trait). The final study population comprised 262 cases with complete data on maternal and fetal cortisol and anxiety ratings. Information on key covariates, time of collection, gestational age, fetal sex, and maternal age were also recorded. Eighteen of the women were taking some type of medication (SSRI, 1; antihypertensive, 2; antiasthma, 10; antiepileptic 1; steroids, 1; other, 3). Maternal blood was drawn immediately before the procedure, centrifuged and supernatant plasma stored at 80 8C until batch assay. During amniocentesis an additional aliquot of up to 4 ml of amniotic fluid surplus to clinical requirement was drawn for the study and stored at 80 8C until assay. Time of collection was recorded to the nearest 15 min. Total cortisol in amniotic fluid was assayed by radio-immunoassay (Coat-A-Count, DPC, Los Angeles, CA), cortisol having been extracted by dichloromethane and reconstituted prior to assay (Sarkar et al., 2007a,b). Our intra- and inter-assay coefficients of variation for the cortisol assay in amniotic fluid were 4.4% and 6.5% respectively. Some of the cortisol data has been reported previously (Sarkar et al., 2006, 2007a,b).
2.1. Data analysis Normality of distribution of data was checked in a P-P plot. Non-normally distributed variables underwent natural log (ln) transformation, and normal distribution was attained for maternal and amniotic fluid cortisol but not for gestational age. Statistical analyses were performed using ln data, which are substantially identical to non-transformed data.
432
V. Glover et al.
Maternal age, fetal sex, gestational age, and time of collection were included as covariates in the analyses.
Table 1 Clinical characteristics of women undergoing amniocentesis. Data shown as mean (SD) [range] n = 262.
3. Results
Maternal age (years) Gestation (weeks) Time of collection
The characteristics of the sample are shown in Table 1. It may be seen that the cortisol levels in maternal plasma were about 28 fold higher than those in the amniotic fluid. There was a positive correlation between amniotic fluid cortisol with maternal plasma levels (r = .32, p < .001), in the group as a whole. Fig. 1 shows the correlations between maternal plasma and amniotic fluid cortisol in the mothers according to maternal State anxiety quartiles. In the most anxious group the correlation was r(62) = .59, p < .001; there was a substantial and significant decrease in the magnitude of the correlation in the less anxious quartiles, in stepped fashion, with the correlation falling to r(60) = .05, ns in the least anxious group (see figure legend for significant contrasts). Subsequent analyses that covaried gestational age, time of amniotic fluid collection, fetal sex, and maternal age and medication yielded substantively similar results. For example, the correlation between maternal plasma cortisol and amniotic fluid cortisol was .51 for the most anxious women and .01 for the least anxious women (the difference between correlations was significant at p < .01). A comparable change in the magnitude of the correlation between amniotic fluid and plasma cortisol was also found when we considered Trait anxiety (Fig. 2). Thus, the correlation between maternal plasma cortisol and amniotic fluid
35.3 (5.4) [18—47] 17.7 (3.6) [15—37] 11.30 (1.4) [9.05—17.30]
Ethnicity, number (%) Caucasian 189 (72.1) Indian/subcontinent 23 (8.8) Black 21 (8) Middle eastern 15 (5.7) Asian-far eastern 10 (1.5) Other 4 (1.5) Spielberger State Anxiety 49.7 (13.6) [20—77] Spielberger Trait Anxiety 36.3 (8.8) [21—63] Maternal cortisol (nmol/l) 579 (240) [157—1770] Amniotic fluid cortisol (nmol/l) 20.6 (12.5) [.66—100]
cortisol was r(67) = .49, p < .001 among women in the top quartile of Trait anxiety, whereas the correlation was r(64) = .13 among women in the least anxious quartile. These patterns also held after adjusting for gestational age, time of amniotic fluid collection, fetal sex, medication and maternal age. The correlation between State and Trait anxiety in the whole sample was moderate, r(262) = .35, p < .0001), implying that the stepped decrease in the link between maternal plasma and amniotic fluid cortisol according to maternal anxiety across the two measures, is noteworthy and not redundant.
Figure 1 Bivariate correlations between ln maternal plasma (nmol/l) and ln amniotic fluid (AF) cortisol (nmol/l) according to maternal State anxiety (quartiles). (a) Quartile 1. Speilberger State anxiety 20-38: r(60) = .05; (b) Quartile 2. Speilberger State anxiety 39—49: r(71) = .29; (c) Quartile 3 Speilberger State anxiety 50—61: r(69) = .40; (d) Quartile 4. Speilberger State anxiety 62—77: r(62) = .59. Correlations were significantly different (using Fisher r to z transformation) between Quartiles 1 and 3 ( p < .05) and Quartiles 1 and 4 ( p < .001) and Quartiles 2 and 4 ( p < .05).
Antenatal anxiety and maternal and amniotic fluid cortisol correlation
433
Figure 2 Bivariate correlations between ln maternal plasma (nmol/l) and ln amniotic fluid (AF) cortisol (nmol/l) according to maternal Trait anxiety (quartiles). (a) Quartile 1. Speilberger Trait anxiety 21—29: r(64) = .13; (b) Quartile 2. Speilberger Trait anxiety 30—34: r(64) = .37, p < .01; (c) Quartile 3. Speilberger Trait anxiety 35—41: r(67) = .40, p < .001; (d) Quartile 4. Speilberger Trait anxiety 42—63: r(67) = .49, p < .001. Correlations between Quartiles 1 and 4 were significantly different (using Fisher r to z transformation) at p < .05.
We next examined whether the level of amniotic fluid cortisol varied significantly as a function of maternal State anxiety. We found no evidence of a significant correlation (r(262) = .09, ns); Table 2 shows the raw (untransformed) means (SD) for the amniotic fluid cortisol for mothers in the least to most state anxiety quartiles, differences that fell short of statistical significance (F(3,258) = .46); the same pattern was observed with the transformed amniotic fluid cortisol data. In contrast, there was a significant difference in maternal plasma cortisol among women in the State anxiety quartiles (from least to most anxious quartiles): F(3,258) = 4.04, p < .01), although the bivariate correlation was modest r(262) = .20, p < .01. There were no significant differences in maternal plasma or amniotic fluid cortisol according to maternal Trait anxiety quartile. Table 2 Cortisol levels (nmol/l), mean (SD), in maternal plasma and amniotic fluid by State and Trait Anxiety quartiles. Maternal plasma
Amniotic fluid
State anxiety Least anxious (n = 60) 2nd quartile (n = 71) 3rd quartile (n = 69) Most anxious (n = 62)
505 556 612 640
(175) (212) (233) (306)
18.9 21.0 21.3 20.9
(12.6) (12.9) (9.3) (14.9)
Trait anxiety Least anxious (n = 64) 2nd quartile (n = 64) 3rd quartile (n = 67) Most anxious (n = 67)
541 570 608 596
(202) (213) (265) (269)
20.4 20.7 21.0 20.2
(12.4) (10.3) (10.6) (16.0)
3.1. Supplementary analysis Results so far indicated that maternal anxiety may be associated with altered placental function, indexed by the moderating effect of maternal anxiety on the association between maternal plasma and amniotic fluid cortisol, but not increased exposure of the fetus to cortisol. A set of supplemental analyses were carried out to examine further this lack of association between maternal anxiety and amniotic fluid cortisol and to examine differences according to fetal sex. Based on our previous findings that the association between maternal plasma cortisol and amniotic fluid cortisol was greater at later gestation age (Sarkar et al., 2007a,b), exploratory analyses examined if a significant link between prenatal anxiety and amniotic fluid cortisol was evident at later gestational age; it was not, and neither did time of collection appear to influence the link between prenatal anxiety and amniotic fluid cortisol. We next considered if the moderating effect of maternal anxiety on the link between maternal plasma and amniotic fluid cortisol was different in male and female fetuses. The correlation between maternal plasma and amniotic fluid cortisol was comparable in females (r(130) = .37, p < .001) and males (r(132) = .30, p < .001); there was no evidence that the moderating effect of maternal anxiety on the link between maternal plasma and amniotic fluid cortisol differed according to fetal sex.
4. Discussion This study is the first to examine the possible effects of maternal mood on placental function in humans. It supports
434 the hypothesis that if the mother is anxious while pregnant then there is a greater association between her cortisol level and that of her fetus, and suggests that the emotional state of the mother may have a direct effect on placental function. There was a similar stepped relationship between maternal/ amniotic fluid cortisol correlation and both State and Trait anxiety (Figs. 1 and 2), implying a robust effect of maternal anxiety on the correlation. Comparable results for Trait (how the individual generally feels) and State (how the person feels now) anxiety is noteworthy. We have previously reported that State anxiety is considerably elevated in these women awaiting amniocentesis, and this can cause a modest increase in their cortisol (Sarkar et al., 2006). However Trait anxiety is not raised in these women, and this together with the dose—response pattern shown in the figures, imply that this effect is not limited to clinical extremes. There is some evidence that placental glucocorticoid metabolism can be sexually dimorphic (Clifton, 2005). However, there was no significant difference in the moderating effect of maternal anxiety on maternal/amniotic fluid cortisol correlation dependent on the sex of the fetus. The results presented here are compatible with a down regulation of placental 11b-HSD2, as found in the prenatal stress rodent model of Mairesse et al. (2007). However, obviously our results provide only indirect evidence for such a down regulation. It should also be noted that Welberg et al. (2005), also using a rodent model, found that acute stress up regulated the placental enzyme, and that chronic stress reduced this up regulation, although there are several differences between the experimental designs of these two studies, in type of stress and timing, which may be relevant. The importance of placental 11b-HSD2 in determining outcome for the offspring, independent of maternal HPA axis function, has been shown by the experiments of Holmes et al. (2006). Using knock out mice for 11b-HSD2 they demonstrated the key role of fetoplacental 11b-HSD2 in glucocorticoid programming in the offspring, resulting in both lower birthweight and increased anxiety. The placenta clearly plays a crucial role in moderating fetal exposure to maternal factors, and presumably in preparing the fetus for the environment in which it is going to find itself (Gluckman et al., 2005). One notable aspect of our results was that there was no difference in the cortisol level in amniotic fluid from the most and least anxious women. That is, the change in maternal and amniotic fluid cortisol correlation, associated with maternal anxiety, was apparently not directly linked with increased fetal cortisol exposure. How that is reconciled with the strong moderation effects is not certain. It might be that the ‘‘signal strength’’, the correlation of r = .09 between maternal anxiety and amniotic fluid cortisol, is muted by other intermediating processes. It is notable that Mairesse et al. (2007) also found no difference in the corticosterone level in the rat fetus despite direct evidence for down regulation of both 11b-HSD2 expression and activity. However ACTH levels and adrenal size were significantly reduced in the fetuses from the prenatally stressed dams. These authors suggest that these changes in the fetal HPA axis are adjustments made by the fetus in response to prior increased exposure to maternal glucocorticoids. It is of interest that in the offspring of 11b-HSD2 knockout mice (Holmes et al., 2006) there was also a hypotrophy of the adrenal glands. Thus the contribution by the fetal adrenal
V. Glover et al. gland to the amniotic fluid cortisol may be reduced, despite an increased contribution from the mother, leaving the overall level unchanged. We were unable to examine either ACTH or fetal adrenal volume in our study. It will clearly be important to study further the effect of both acute and chronic maternal anxiety on placental 11b-HSD2, both by direct assay of the placental enzyme activity and expression, and by measuring cord blood or amniotic fluid cortisol/cortisone ratio. It is also important to note that our studies were carried out at a median gestational age of 17 weeks. It is possible that there is a stronger association between maternal anxiety and amniotic fluid cortisol later in gestation, when some studies suggest maternal anxiety has a stronger effect on fetal programming (e.g. O’Connor et al., 2002). The question arises of what may be the mediator between maternal anxiety and a putative down regulation of 11bHSD2, if not increased maternal cortisol. The peripheral sympathetic system is a plausible candidate; it is possible that the sympathetic system remains more responsive to anxiety or stress during pregnancy than the HPA axis. In a rat model it has been demonstrated that whereas the HPA axis becomes unreactive to an acute stressor as pregnancy progresses, the reactivity of the peripheral sympathetic system is unaffected (Douglas personal communication). Noradrenaline has been shown to down regulate 11b-HSD2 in isolated placental human trophoblast cells (Sarkar et al., 2001). There are several possible sources of amniotic fluid cortisol, apart from the maternal, including the fetal adrenal, and 11b-HSD1 in the placenta and the fetal membranes which converts cortisone to cortisol. These mature at different stages of gestation, but the effect of maternal anxiety on the amniotic fluid/maternal cortisol correlation was little affected by covarying gestational age. We do not know how long the acute anxiety experienced while awaiting amniocentesis lasted, or how long maternal cortisol takes to cross the placenta. However covarying time of amniocentesis did not affect the effect of anxiety on the correlation. The results presented here may have a variety of clinical implications. For example, prednisolone is used to treat several maternal diseases during pregnancy including asthma, hyperemesis gravidarium, and rheumatoid arthritis. It is metabolised by 11-b HSD2 in a similar manner to cortisol, and is generally considered to be safe for that reason. This study suggests the possibility that the level of prednisolone exposure to the fetus may be greater for more anxious mothers. Additionally, these findings underscore the value that clinical interventions for pregnant anxious women may have for both understanding basic mechanisms of fetal programming and for preventive medicine. Although our results provide only indirect evidence for an alteration in 11b-HSD2, they do provide the first robust human evidence that placental function may be affected by maternal anxiety. Accordingly, the findings are an important step in translating key animal findings and in understanding the mechanisms of fetal programming.
Role of the funding sources Funding for this study was provided by March of Dimes; March of Dimes had no further role in study design; in the collection, analysis and interpretation of data; in the writing of
Antenatal anxiety and maternal and amniotic fluid cortisol correlation the report; and in the decision to submit the paper for publication.
Conflict of interest None of the authors have any actual or potential conflict of interest related to the submitted manuscript.
Acknowledgement We should like to thank March of Dimes for funding and Diana Adams for help with the recruitment of the subjects, scoring of the questionnaires and data collection, and Kieran O’Donnell for helpful discussion of the findings.
References Bergman, K., Sarkar, P., O’Connor, T.G., Modi, N., Glover, V., 2007. Maternal stress during pregnancy predicts cognitive ability and fearfulness in infancy. J. Am. Acad. Child Adolesc. Psychiatry 46, 1454—1463. Clifton, V.L., 2005. Sexually dimorphic effects of maternal asthma during pregnancy on placental glucocorticoid metabolism and fetal growth. Cell Tissue Res. 322, 63—71. Davis, E.P., Glynn, L.M., Schetter, C.D., Hobel, C., Chicz-Demet, A., Sandman, C.A., 2007. Prenatal exposure to maternal depression and cortisol influences infant temperament. J. Am. Acad. Child Adolesc. Psychiatry 46, 737—746. Dy, J., Guan, H., Sampath-Kumar, R., Richardson, B.S., Yang, K., 2008. Placental 11beta-hydroxysteroid dehydrogenase type 2 is reduced in pregnancies complicated with idiopathic intrauterine growth Restriction: evidence that this is associated with an attenuated ratio of cortisone to cortisol in the umbilical artery. Placenta 29, 193—200. Gluckman, P.D., Hanson, M.A., Spencer, H.G., 2005. Predictive adaptive responses and human evolution. Trends Ecol. Evol. 20, 527—533. Herbert, J., Goodyer, I.M., Grossman, A.B., Hastings, M.H., de Kloet, E.R., Lightman, S.L., Lupien, S.J., Roozendaal, B., Seckl, J.R., 2006. Do corticosteroids damage the brain? J. Neuroendocrinol. 18, 393—411. Holmes, M.C., Abrahamsen, C.T., French, K.L., Paterson, J.M., Mullins, J.J., Seckl, J.R., 2006. The mother or the fetus? 11beta-hydroxysteroid dehydrogenase type 2 null mice provide evidence for direct fetal programming of behavior by endogenous glucocorticoids. J. Neurosci. 26, 3840—3844. Jansson, T., Powell, T.L., 2007. Role of the placenta in fetal programming: underlying mechanisms and potential interventional approaches. Clin. Sci. (Lond.) 113, 1—13. Kajantie, E., Dunkel, L., Turpeinen, U., Stenman, U.H., Wood, P.J., Nuutila, M., Andersson, S., 2003. Placental 11 beta-hydroxysteroid dehydrogenase-2 and fetal cortisol/cortisone shuttle in small preterm infants. J. Clin. Endocrinol. Metab. 88, 493—500. Kammerer, M., Adams, D., Castelberg Bv, B., Glover, V., 2002. Pregnant women become insensitive to cold stress. BMC Pregnancy Childbirth 2, 8. Mairesse, J., Lesage, J., Breton, C., Breant, B., Hahn, T., Darnaudery, M., Dickson, S.L., Seckl, J., Blondeau, B., Vieau, D., Maccari, S., Viltart, O., 2007. Maternal stress alters endocrine function of the
435
feto-placental unit in rats. Am. J. Physiol. Endocrinol. Metab. 292, E1526—E1533. O’Connor, T.G., Heron, J., Golding, J., Beveridge, M., Glover, V., 2002. Maternal antenatal anxiety and children’s behavioural/ emotional problems at 4 years Report from the Avon Longitudinal Study of Parents and Children. Br. J. Psychiatry 180, 502—508. O’Connor, T.G., Heron, J., Golding, J., Glover, V., 2003. Maternal antenatal anxiety and behavioural/emotional problems in children: a test of a programming hypothesis. J. Child. Psychol. Psychiatry 44, 1025—1036. Obel, C., Hedegaard, M., Henriksen, T.B., Secher, N.J., Olsen, J., Levine, S., 2005. Stress and salivary cortisol during pregnancy. Psychoneuroendocrinology 30, 647—656. Salaria, S., Chana, G., Caldara, F., Feltrin, E., Altieri, M., Faggioni, F., Domenici, E., Merlo-Pich, E., Everall, I.P., 2006. Microarray analysis of cultured human brain aggregates following cortisol exposure: implications for cellular functions relevant to mood disorders. Neurobiol. Dis. 23, 630—636. Sarkar, P., Bergman, K., Fisk, N.M., Glover, V., 2006. Maternal anxiety at amniocentesis and plasma cortisol. Prenat. Diagn. 26, 505— 509. Sarkar, P., Bergman, K., Fisk, N.M., O’Connor, T.G., Glover, V., 2007a. Ontogeny of foetal exposure to maternal cortisol using midtrimester amniotic fluid as a biomarker. Clin. Endocrinol. (Oxf.) 66, 636—640. Sarkar, P., Bergman, K., Fisk, N.M., O’Connor, T.G., Glover, V., 2007b. Amniotic fluid testosterone: relationship with cortisol and gestational age. Clin. Endocrinol. (Oxf.) 67, 743—747. Sarkar, S., Tsai, S.W., Nguyen, T.T., Plevyak, M., Padbury, J.F., Rubin, L.P., 2001. Inhibition of placental 11beta-hydroxysteroid dehydrogenase type 2 by catecholamines via alpha-adrenergic signalling. Am. J. Physiol. Regul. Integr. Comp. Physiol. 281, R1966— R1974. Schneider, M.L., Moore, C.F., Kraemer, G.W., Roberts, A.D., DeJesus, O.T., 2002. The impact of prenatal stress, fetal alcohol exposure, or both on development: perspectives from a primate model. Psychoneuroendocrinology 27, 285—298. Schulte, H.M., Weisner, D., Allolio, B., 1990. The corticotrophin releasing hormone test in late pregnancy: lack of adrenocorticotrophin and cortisol response. Clin. Endocrinol. (Oxf.) 33, 99— 106. Seckl, J.R., Holmes, M.C., 2007. Mechanisms of disease: glucocorticoids, their placental metabolism and fetal ‘programming’ of adult pathophysiology. Nat. Clin. Pract. Endocrinol. Metab. 3, 479—488. Spielberger, C.D., Gorusch, R.L., Rushene, R.E. (Eds.), 1970. STAI Manual for The State-Trait Anxiety Inventory. Consulting Psychologists Press Inc., Pao Alto, CA. Talge, N.M., Neal, C., Glover, V., 2007. Antenatal maternal stress and long-term effects on child neurodevelopment: how and why? J. Child Psychol. Psychiatry 48, 245—261. Van den Bergh, B.R., Mulder, E.J., Mennes, M., Glover, V., 2005. Antenatal maternal anxiety and stress and the neurobehavioural development of the fetus and child: links and possible mechanisms. A review. Neurosci. Biobehav. Rev. 29, 237—258. Weinstock, M., 2001. Alterations induced by gestational stress in brain morphology and behaviour of the offspring. Prog. Neurobiol. 65, 427—451. Welberg, L.A., Thrivikraman, K.V., Plotsky, P.M., 2005. Chronic maternal stress inhibits the capacity to up-regulate placental 11beta-hydroxysteroid dehydrogenase type 2 activity. J. Endocrinol. 186, R7—R12.