UBC Faculty Research and Publications

Hair cortisol as a hypothalamic-pituitary-adrenal axis biomarker in pregnant women with asthma: a retrospective… Smy, Laura; Shaw, Kaitlyn; Amstutz, Ursula; Smith, Anne; Berger, Howard; Carleton, Bruce; Koren, Gideon Jul 20, 2016

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


52383-12884_2016_Article_962.pdf [ 686.33kB ]
JSON: 52383-1.0361951.json
JSON-LD: 52383-1.0361951-ld.json
RDF/XML (Pretty): 52383-1.0361951-rdf.xml
RDF/JSON: 52383-1.0361951-rdf.json
Turtle: 52383-1.0361951-turtle.txt
N-Triples: 52383-1.0361951-rdf-ntriples.txt
Original Record: 52383-1.0361951-source.json
Full Text

Full Text

RESEARCH ARTICLE Open AccessHair cortisol as a hypothalamic-pituitary-adrenal axis biomarker in pregnant womenwith asthma: a retrospective observationalstudyLaura Smy1,2, Kaitlyn Shaw3,4, Ursula Amstutz3,4,5, Anne Smith3,4ˆ, Howard Berger6, Bruce Carleton3,4,7*and Gideon Koren1,2AbstractBackground: Cortisol is a hormone involved in many physiological functions including fetal maturation andepigenetic programming during pregnancy. This study aimed to use hair cortisol as a biomarker of chronic inhaledcorticosteroid (ICS) exposure and assess the potential effects of asthma on the hypothalamic-pituitary-adrenal (HPA)axis in pregnant women. We hypothesized that pregnant women with asthma treated with ICS would exhibit lowerhair cortisol concentrations, indicative of adrenal suppression, compared to women with asthma not using ICS andwomen who do not have asthma.Methods: We performed an observational retrospective cohort study. Hair samples were analyzed from pregnantwomen with asthma, with (n = 56) and without (n = 31) ICS treatment, and pregnant women without asthma(n = 31). Hair samples were segmented based on the growth rate of 1 cm/month and analyzed by enzymeimmunoassay to provide cortisol concentrations corresponding to preconception, trimesters 1–3, and postpartum.Hair cortisol concentrations were compared within and among the groups using non-parametric statistical tests.Results: Hair cortisol concentrations increased across trimesters for all three groups, but this increase wasdampened in women with asthma (P = 0.03 for Controls vs. ICS Treated and Controls vs. No ICS). ICS Treatedwomen taking more than five doses per week had hair cortisol concentrations 47 % lower in third trimester thanControls. Linear regression of the third trimester hair cortisol results identified asthma as a significant factor whencomparing consistent ICS use or asthma as the predictor (F(1, 25) = 9.7, P = 0.005, R2adj = 0.257).Conclusions: Hair cortisol successfully showed the expected change in cortisol over the course of pregnancy andmay be a useful biomarker of HPA axis function in pregnant women with asthma. The potential impact ofdecreased maternal cortisol in women with asthma on perinatal outcomes remains to be determined.Keywords: Hair cortisol, Pregnancy, Asthma, Biomarker, HPA axis, Adrenal suppression* Correspondence: bcarleton@popi.ubc.caˆDeceased3Child & Family Research Institute, Vancouver, BC, Canada4Division of Translational Therapeutics, Department of Pediatrics, Universityof British Columbia, Vancouver, BC, CanadaFull list of author information is available at the end of the article© 2016 The Author(s). Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, andreproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link tothe Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.Smy et al. BMC Pregnancy and Childbirth  (2016) 16:176 DOI 10.1186/s12884-016-0962-4BackgroundCortisol has many important functions in the body in-cluding metabolism, regulation of blood pressure, androles in the inflammatory and stress responses. Whenrequired, the hypothalamus and pituitary gland are acti-vated to release a series of hormones that signal the ad-renal cortex to release cortisol. Asthma is a commonchronic inflammatory disease for which the recom-mended therapy for long-term control is inhaled cortico-steroids (ICS) [1]. ICS bind to glucocorticoid receptorsin lung epithelial cells and reduce airway inflammation.Binding also mimics endogenous cortisol and conse-quently initiates a negative feedback resulting in de-creased cortisol release from the adrenal glands. Whilegenerally believed to confer less systemic exposure thansystemic corticosteroids, in severe cases, the use of ICSmay cause adrenal insufficiency or crisis [2, 3].Classically, the matrices used to measure a patient’scortisol level are blood, saliva, or urine. However, thesematrices do not account for the diurnal nature of corti-sol secretion and only reflect the point in time whensampling is conducted [4]. More recently, hair analysishas emerged as a viable alternative for cortisol detection.Hair cortisol levels correlate with a 24-h urine sample(r = 0.33, p = 0.04) and multiple saliva samples col-lected over 7 days (r = 0.41, p = 0.03) [4, 5]. Since hairgrows on average one centimeter per month (cm/mo)([6], p. 2), the cortisol detected in a one-centimeterhair segment represents the average cortisol level overthe corresponding one-month period.Hair cortisol has previously been used to assess the ef-fects of various medical conditions on the hypothalamic-pituitary-adrenal (HPA) axis, including psychologicaland physical stressors. It is already known that cortisolconcentrations increase in relation to a stressful event,and these changes were reflected in the hair of childrenfearful of beginning school and individuals who recentlyexperienced a traumatic event [7–9]. Similarly, hair cor-tisol concentrations in patients diagnosed with Cushingsyndrome corresponded to the characteristic increasedendogenous cortisol concentrations, with 86 % sensitivityand 98 % specificity for the detection of cyclic Cushing’ssyndrome [10, 11]. Recently, we examined hair cortisol ofchildren with asthma due to concerns of potential adverseeffects from long-term ICS therapy [12]. Our resultsshowed a 55 % decrease in hair cortisol concentration dur-ing ICS therapy compared to prior to ICS therapy suggest-ing significant ICS-induced HPA axis suppression andproviding further support for using hair as a matrix formeasuring cortisol.In 2001 in the United States, the prevalence of asthmain pregnancy was 8 % [13]. Only one study has examinedhormone concentrations, including cortisol, amongpregnant women with asthma with and without ICStreatment as compared to pregnant women withoutasthma and found no difference among the three groups[14]. Given that cortisol is a vital factor in fetal lung,gastrointestinal tract, kidney, and thyroid maturation[15], our objective was to use hair cortisol as a bio-marker to investigate cortisol changes over the course ofpregnancy in the context of potential adverse effects ofICS on the HPA axis in pregnant women with asthma.We hypothesized that pregnant women with asthmatreated with ICS would exhibit lower hair cortisol con-centrations, suggestive of adrenal suppression, comparedto women with asthma not using ICS and women whodo not have asthma.MethodsStudy design, participants, and ethicsWe performed a retrospective observational cohort study.Three groups of pregnant women were recruited fromJune 2012 to December 2014: women with ICS-treatedasthma (ICS Treated), and two comparison groups con-sisting of women without asthma (Controls) and adisease-matched group of women with asthma not treatedwith ICS (No ICS). The sample size was not formally de-termined due to the unavailability of data on hair cortisolconcentrations in pregnant women with and withoutasthma when the protocol was created. Participants wererecruited in person in obstetric clinics at the BritishColumbia Women’s Hospital and Health Centre inVancouver, British Columbia, and St. Michael’s Hospital inToronto, Ontario. Additionally, women were recruitedthrough the Hospital for Sick Children’s MotheriskProgram teratology information service via telephone andmail. All pregnant women were eligible provided theycould read and understand English, did not use any cor-ticosteroid products on their scalp, or have any knownmedical conditions characterized by high cortisol levelssuch as Cushing’s syndrome. Women were recruited atany time during pregnancy up to 6 months postpartum.Research ethics board approval was obtained from eachinstitution and informed written consent was obtainedfrom all participants.Data collectionRelevant clinical information was obtained through med-ical record review and/or patient interview and includedasthma and ICS treatment history and concomitantmedications. Factors reported to affect hair cortisollevels, such as frequency of hair washing, days since lastwashing and chemical treatment (color or relaxer), werealso collected [16]. Additionally, the Perceived StressScale (PSS), a validated tool to assess stress levels experi-enced in the previous month, was administered towomen enrolled in Ontario [17].Smy et al. BMC Pregnancy and Childbirth  (2016) 16:176 Page 2 of 10Hair sample collection and analysisA lock of hair approximately 3 mm thick (equivalent toone-half the diameter of a pencil) was cut from the ver-tex posterior region of the head as close to the scalp aspossible. Based on an average growth rate of 1 cm/mo,each hair sample was further cut into segments 2–3.6 cm in length (the majority were 3 cm long) to cor-respond to preconception (PC), trimesters 1–3 (T1, T2,T3), and immediate postpartum (PP) time points asavailable depending on the hair length and collectiondate ([6], p. 2). Day 0 of each hair sample was consideredfourteen days prior to date of hair collection to accountfor the fact that 10–14 days of hair resides below thescalp ([6], p. 35). Segmented hair samples were proc-essed for hair cortisol extraction as previously reported[18], with a few minor modifications. In brief, each sam-ple of 10–25 mg of hair was washed twice with isopro-panol and allowed to dry, and then finely minced withscissors and extracted overnight in methanol. After re-moving all of the supernatant, samples were dried underN2 at 37 °C, reconstituted with 125 μL of phosphatebuffered saline, and vortexed for one minute. Initially,samples were reconstituted with 250 μL, but this re-sulted in some samples having results below the quanti-tation limit (0.33 pmol/mL or 0.12 ng/mL).The samples were analyzed using the Salimetrics HighSensitivity Salivary Cortisol Enzyme Immunoassay (EIA)kit (Salimetrics, Philadelphia, PA) following the manu-facturer’s instructions. All of the % cross-reactivity re-ported by the manufacturer is less than 0.6 % for othersteroid hormones, such as prednisolone, prednisone, orcortisone, except for dexamethasone, which was 19.2 %.Additionally, the cross-reactivity was determined foreach of the ICS available in Canada using an 8000 ng/mL solution run six times on two different EIA plates.The results were not detectable for fluticasone propionate,budesonide and ciclesonide, 0.01 % for beclomethasonedipropionate, and 0.03 % for mometasone furoate. Haircortisol concentrations measured using this kit have beencorrelated to two different liquid chromatography-massspectrometry methods with a Spearman rho of ≥ 0.92(P < 0.0001) [19].In addition to the quality control material includedwith the kit, a pooled in-house hair sample was run as athird quality control and the results were evaluatedusing Westgard Rules for acceptance or rejection. Theresults of the EIA were considered acceptable if two ofthree control values were within expected range. Basedon the coefficients of variation for the participant sam-ples, the average intra-day coefficient of variation was6.3 %. The average inter-day coefficients of variationwere 1.5 and 4.5 % for the high and low quality control,respectively. Because all samples are run in duplicatewith EIA analysis, any duplicates that had a coefficientof variation greater than 15 or 20 % for results greaterthan or lower than approximately 3 pmol/mL (1 ng/mL),respectively, were reanalyzed or reprocessed based onsample availability. Cortisol concentrations are reportedas a ratio to the hair sample weight (pmol/g).Statistical analysisStatistical analysis was performed using Graphpad Prismsoftware, version 5.0c (Graphpad Software, Inc., La Jolla,CA) and SPSS, version 22.0 (IBM Corporation, New York,NY). Comparisons between the groups for hair cortisolconcentrations, demographic information, and clinical in-formation were performed using one-way analysis of vari-ance with Tukey’s multiple comparison test, the Kruskal-Wallis test with Dunn’s multiple comparison test, andChi-square test or Freeman-Halton extension of theFisher’s exact probability test as appropriate for normallyor not normally distributed, and categorical data. Furtherpost hoc analyses were performed using the unpaired t-testwith Welch’s correction, and Mann–Whitney U test withBonferroni correction for multiple comparisons to calcu-late the adjusted p-values (Padj), as needed.The Friedman test was used to compare hair cortisolconcentrations for the five time points within eachgroup. Because most women did not have results for allfive time points, post hoc analysis for the Friedman testwas performed using the Wilcoxon matched-pairs signedrank test. A natural log transformation was required forthe hair cortisol data prior to performing the Wilcoxontests to best satisfy statistical assumptions. The Holm-Bonferroni correction was applied to p-values to correctfor multiple comparisons. Additionally, the linear regres-sion for median hair cortisol concentrations from PC toT3 for each group were compared using analysis of co-variance with the Bonferroni correction for multiplecomparisons to calculate the adjusted p-values. Univari-ate and multiple linear regressions were performed withthe variables “consistent ICS use” (defined as use of ICSfor ≥ 5 doses per week), “intranasal corticosteroid use”(yes/no), and “asthma” (yes/no) for T3 to determine ifthere was any influence of these variables on haircortisol concentrations in that trimester.Spearman correlations were calculated for the haircortisol results with previously published confounders,including the age of hair sample, body mass index (BMI)(pre-pregnancy and at time of hair collection), numberof hair washes per week, days since last washing, andPSS score. Because hair chemical treatment is a binaryoutcome, a point biserial correlation was performedusing natural log transformed hair cortisol concentra-tions with concentrations ≥ 276 pmol/g removed to bestsatisfy the assumption of normally distributed data,which was unsuccessful for the PC concentrations in thetreated asthma group. Correlations with the hairSmy et al. BMC Pregnancy and Childbirth  (2016) 16:176 Page 3 of 10segment length were added post hoc when the segmentlengths were finalized and some deviated from 3 cm.In all cases, two-tailed p-values were calculated andconsidered significant if ≤ 0.05.ResultsParticipant results and demographic comparisonHair samples were analyzed for 118 pregnant women,consisting of 31 Controls, 31 No ICS, and 56 ICSTreated. Fourteen additional hair samples from the ICSTreated group could not be analyzed due to insufficientquantity, inaccurate segmentation, ICS being used totreat a condition other than asthma, or hair being col-lected too early in T1 (i.e., if the hair segment was lessthan 2 cm) or >6 months postpartum. Some of the firsthair samples analyzed had cortisol concentrations belowthe quantitation limit (0.33 pmol/mL) but could not berepeated due to insufficient quantity of the original hairsample; therefore, these results were recalculated usingthe quantitation limit. This recalculation predominantlyaffected the No ICS group (n = 5 patients/12 segments)compared to the Controls (n = 3 patients/3 segments) orICS Treated (n = 2 patients/3 segments).Comparisons of demographics, hair variables, andmedication use among the groups are listed in Table 1.Overall, there were no significant differences among thethree groups except for their use of intra-nasal cortico-steroids and beta-agonists, which, as anticipated, weremore frequent among women with asthma (Table 1). As-sociations between the hair cortisol results and factorspreviously reported to affect hair cortisol concentrationswere further explored but no significant confounding ef-fects were found (see Additional file 1: Table S1). There-fore, the test statistics were not adjusted for any of thesefactors. Of the women recruited postpartum or in T3,only one woman in the No ICS group receivedTable 1 Comparison of demographics, hair variables, and medication use among the three groups of pregnant womenControls(n = 31)No ICS(n = 31)ICS treated(n = 56)PDemographicsAge, years, mean (SD) 33.8 (4.3) 31.6 (6.0) 33.3 (5.5) 0.240kBMI, kg/m2, median (IQR, n)Pregnant 28.8 (25.4–32.5, 31) 28.2 (25.0–34.2, 25) 27.8 (24.7–32.2, 41) 0.716lPre-Pregnancy 24.6 (20.9–27.2, 31) 25.8 (22.8–31.3, 25) 24.8 (21.8–29.4, 42) 0.554lPSS Score, mean (SD, n) 12 (5–26) 15 (7–9) 15 (6–23) 0.210kBirth DataaGestational age, weeks, median (IQR, n) 39.6 (39.0–40.9, 30) 39.1 (37.6–40.6, 28) 39.1 (38.1–40.3, 51) 0.155lBirth weight, kg, median (IQR, n) 3.41 (3.11–3.65, 31) 3.37 (2.76–3.64, 30) 3.31 (2.89–3.63, 52) 0.701lHair Sample VariablesSample age, daysb, mean (SD) 334 (82) 325 (93) 315 (103) 0.661k# Washes per week, median (IQR, n) 3.5 (2.5–4.5, 31) 3.0 (2.5–5.5, 24) 4.0 (3–6.6, 42) 0.141l# Days since last washed, median (IQR, n) 1 (0–2, 29) 1 (0–1, 23) 1 (0–1, 37) 0.375lChemical treatmentc, n (%) 19 (61.3) 17 (of 26, 65.4) 25 (of 47, 53.2) 0.563mMedication Use During PregnancyInhaled corticosteroid used, Yes/No No No Yes –Oral corticosteroid usee, n (%) 0 (0) 3 (9.7) 4 (7.1) 0.280mIntranasal corticosteroid usef, n (%) 3 (9.7)h 2 (6.5)h 16 (28.6) 0.014mTopical corticosteroid usef, n (%) 5 (16.1) 2 (6.4) 7 (12.5) 0.343mOther steroid hormone use (e.g., progesterone)f, n (%) 3 (9.7) 4 (12.9) 6 (10.7) 0.867mBeta-agonistf,g, n (%) 2 (6.4)i 24 (77.4)j 54 (96.4) <0.0001mNumber of other classes of medications usedf, median (IQR) 2 (1–3) 2 (0–4) 2 (1–3) 0.650lBMI body mass index, PSS perceived stress scale, IQR interquartile range, SD standard deviationaInclusive of data for twin births, the frequency of which was not significantly different among the groups. bThe age is calculated to the oldest part of the hairsample at the beginning of the preconception segment. cFor No ICS and ICS Treated, information regarding chemical treatment (color or relaxer) was notavailable for all women. The total number of women is indicated in the parentheses. dInhaled corticosteroid use includes any use within the time captured by thetested hair segment, regardless of frequency or duration. eOral corticosteroid use is reported if within one month prior to the tested hair segment. fUse within thelast 12 months, which may not be during pregnancy. gIncludes beta-agonist drugs that are short and long-acting, including use of combination inhaler products.hSignificantly different from ICS Treated, P < 0.05. iSignificantly different from both No ICS and ICS Treated, P < 0.0001. jSignificantly different from ICS TreatedP < 0.01. kOne-way analysis of variance with Tukey’s multiple comparison test. lKruskal-Wallis test with Dunn’s multiple comparison test. mChi-squared or Freeman-Halton extension of the Fisher’s exact probability test, as appropriate. Significant P-values are in boldSmy et al. BMC Pregnancy and Childbirth  (2016) 16:176 Page 4 of 10corticosteroids due to threatened preterm labour duringthe time point captured by her hair sample. It is un-known whether any of the women received hydrocorti-sone stress dose treatment during labour.Hair cortisol concentrations increase during pregnancyHair cortisol concentrations for each time point areshown in Fig. 1. A similar increase in cortisol over thecourse of pregnancy from PC to T3, followed by a de-cline PP, was evident for all three groups, although thistrend was less pronounced for the two groups of womenwith asthma. There were seven patients (five ICSTreated and two No ICS) with a decline in hair cortisolconcentrations from PC to T2. Additionally, there was asubgroup of three women with 11 samples with hair cor-tisol concentrations ≥ 276 pmol/g consisting of one Con-trol for segments PC to PP, and two ICS Treated forsegments PC to T3 and T3 to PP (Fig. 1). The cortisolconcentrations for the Controls differed significantlyacross all time points (χ2(4) = 9.6, P = 0.028, n = 4) and in-creased from 7.9 pmol/g (IQR 3.8–17.0 pmol/g, n = 29)in PC to 21.1 pmol/g (IQR 14.7–31.0 pmol/g, n = 11) inT3. The ICS Treated group also showed overall signifi-cant changes in cortisol (χ2(4) = 21.4, P < 0.001, n = 7) aswell as an increase between PC (7.9 pmol/g, IQR 5.4–14.2 pmol/g, n = 50) and T3 (14.2 pmol/g, IQR 10.2–21.7pmol/g, n = 19). Only the No ICS group did not have asignificant overall change in cortisol across all time points(χ2(4) = 2.1, P = 0.768, n = 3) but did show a similar trend ofincreasing cortisol from PC (8.2 pmol/g, IQR 4.8–12.4pmol/g, n = 29) to T3 (13.0 pmol/g, IQR 8.8–15.7 pmol/g,n = 9), although the increase from PC to T3 was also notsignificant (χ2(4) = 2.0, P = 0.583, n = 8). Generally, post hocanalyses revealed significant differences over the course ofpregnancy between hair cortisol concentrations during PCand T1 compared to T2 and onward for the Controls andICS Treated (Fig. 1).Hair cortisol increase during pregnancy is dampened inwomen with asthmaWhen the median hair concentrations from PC to T3 wereplotted for each group to determine the slopes of the re-gression lines for the change in cortisol during pregnancy,there was a significant difference among the three slopes(F(2,6) = 14.8, P = 0.005) (Fig. 2a). Comparing the slopes toone another, the change in median hair cortisol concentra-tions from PC to T3 for the Controls was significantly dif-ferent from the ICS Treated (F(1,4) = 22.6, Padj = 0.026) anda b cFig. 1 Scatter plots of median hair cortisol concentrations. Median hair cortisol concentrations (horizontal bar) in pmol/g of hair are shown foreach group of pregnant women (a - Controls, b – No ICS, c – ICS Treated) by pregnancy time point consisting of preconception (PC), firsttrimester (T1), second trimester (T2), third trimester (T3), and postpartum (PP). Hair cortisol is plotted on a log10 y-axis. Sample sizes for each timepoint are shown below the x-axis. The change in cortisol over the five time points was significant for the Controls and ICS Treated. Post hoc analysisshowed significant differences between time points as indicated in the figure. When analyses were repeated with all hair cortisol concentrations≥ 276pmol/g removed, the results were not greatly changed. The alternate p-values are shown in parentheses, ( ). Holm-Bonferroni correction for multiplecomparisons was applied to all p-values. *P≤ 0.05, **P≤ 0.01, ***P≤ 0.001Smy et al. BMC Pregnancy and Childbirth  (2016) 16:176 Page 5 of 10No ICS (F(1,4) = 20.2, Padj = 0.033), whereas there was nosignificant difference between the No ICS and ICSTreated (F(1,4) = 0.1, Padj = 2.29). When the subgroup ofwomen with samples with concentrations ≥ 276 pmol/gwere excluded as potential confounders, the comparisonbetween the Controls and No ICS was no longer signifi-cant (F(1,4) = 13.5, Padj = 0.064) (Fig. 2b).Hair cortisol concentrations are lower in third trimesterfor women with asthmaAll groups had a similar median hair cortisol concentra-tion at PC and T1 with differences becoming apparentin T2 and T3 (Fig. 3a, b). When cortisol levels for eachindividual time point were compared among the groups,although there was a visible difference between the Con-trols and two asthma groups for T3, the results were notsignificant (χ2(2) = 5.1, P = 0.078) (Fig. 3a) and performingthe statistical analyses with and without the concentra-tions ≥ 276 pmol/g yielded similar results. However, ifthe cortisol results for all women with asthma werecombined, the mean T3 cortisol concentration was sig-nificantly lower compared to the Controls (t(34) = 2.189,P = 0.036). Furthermore, when women who reported ICSuse less than five doses per week and concentrations≥ 276 pmol/g were excluded as potential confounders,the T3 median hair cortisol concentration for the ICSTreated group was significantly lower (47 %) than theControls (19.9 pmol/g vs. 10.6 pmol/g, U = 9, Padj =0.029) (Fig. 3b). Univariate and multiple linear regressionanalysis to determine the influence of consistent ICSuse, intranasal corticosteroids, or asthma on the T3 haircortisol concentration showed that asthma was the onlysignificant factor (F(1, 25) = 9.7, P = 0.005, R2adj = 0.257).When the samples with concentrations ≥ 276 pmol/gwere included (with ICS use still restricted to five ormore doses per week), the median hair cortisol concen-trations at T3 for ICS Treated group was still 48 % lowerthan the Controls, but the difference was no longer sig-nificant (21.1 pmol/g vs. 11.0 pmol/g, U = 29, Padj =0.386) (Fig. 3c).DiscussionThis study is the first to investigate the use of hair cortisolas a biomarker to assess asthma and the potential effectsof ICS on the HPA axis in pregnant women. The 2- to 3-fold increase in serum or salivary cortisol over the courseof normal pregnancy compared to a non-pregnant state iswell-documented [20–22]. Our findings, which showed anincrease in hair cortisol concentrations over the course ofpregnancy for all groups, are in line with previous haircortisol research [23–25], and, specifically, the 2- to 3-foldincrease in hair cortisol concentrations for the Controlscorresponds well with the previous serum and salivaresearch. Moreover, we also found a suppressed adrenalresponse over the course of pregnancy in women withasthma compared to women without asthma, most not-ably in T3.Our results are in contrast to findings reported previ-ously that did not find a significant difference in serumcortisol concentrations among pregnant women withasthma, with and without ICS treatment, and pregnantwomen without asthma [14]. Limitations of the previousbFig. 2 Comparison of the change in hair cortisol concentrationsduring pregnancy. The linear regressions for median hair cortisolconcentrations for each group of pregnant women from PC to T3are shown. The similarity of the slopes for the No ICS and ICSTreated groups can be seen. The difference of those slopes from theslope for the Controls is significant if the samples withconcentrations ≥ 276 pmol/g are included (a), but when they areexcluded (b), the comparison between the Controls and No ICS isno longer significant. PC = preconception, T1 = first trimester,T2 = second trimester, T3 = third trimesterSmy et al. BMC Pregnancy and Childbirth  (2016) 16:176 Page 6 of 10study was that serum cortisol was only assessed on oneoccasion per trimester for each participant in a 3-h timewindow, and the sample collection timing varied over a4 to 9-week period of each trimester. Moreover, thesample collection for T3 occurred during 25–34 weeksof gestation, which is considered late T2 or early T3.This sampling pattern may not have fully captured thedynamic changes in cortisol that occur during preg-nancy, including the rise in cortisol during T3 as seen inour study. Thus, our results suggest that using hair asthe sample matrix, which is representative of the averagecortisol levels for the entire trimester, may be more sen-sitive for detecting changes in cortisol between each tri-mester and is a significant advantage of our study.Based on previous reports of the association betweendecreased cortisol production and ICS therapy [3, 12], itwas anticipated that pregnant women using ICS mighthave decreased cortisol concentrations compared toboth comparison groups. A significant decrease was ap-parent for women who used more than five ICS dosesper week throughout pregnancy, but only in T3 whencompared to Controls and if the three women with haircortisol concentrations ≥ 276 pmol/g were excluded.Unexpectedly, the No ICS group showed a diminishedadrenal response over the course of pregnancy from PCto T3 similar to the ICS Treated group. One possible ex-planation for the lower hair cortisol concentrations inwomen with asthma may be sustained overwork andresultant fatigue of the HPA axis [26]. Research showsthat the initial response to stress is increased cortisolproduction with decreased expression of pro-inflammatorycytokines; however, with chronic exposure to stress hor-mones there is a decrease in immune system sensitivityand response to cortisol, ultimately resulting in increasedpro-inflammatory cytokines [26]. Pregnancy generates aninflammatory state and approximately one-third ofwomen experience increased asthma symptoms whenpregnant [27]. It is possible that the added physiologicalstress due to pregnancy in combination with asthma, orthe woman’s asthma severity or chronic state, exacerbatesHPA axis fatigue through chronic exposure to stress hor-mones and ultimately leads to a decreased cortisol re-sponse regardless of ICS use. This is supported by thelinear regression analysis of our data for T3 that indicates,between consistent ICS use, intranasal corticosteroid use,and asthma, asthma accounted for approximately 26 % ofthe decrease observed. Also, in support of our findings,patients with chronic asthma were previously found tohave a decreased response to adrenocorticotropic hor-mone stimulation, but research is limited in this area [28].Alternatively, the less-pronounced increase in hair cortisolduring pregnancy in women with asthma may be due todecreased cortisol sensitivity from a reduction in gluco-corticoid receptors [26], as found in children with asthmawho were shown to have a 5.5-fold decrease in expressionof the glucocorticoid receptor [29]. Further research com-paring hair cortisol concentrations in healthy adults tothose with asthma, with and without ICS treatment, mayconfirm whether the observed difference in cortisol con-centrations is due to HPA axis fatigue from the physio-logical stress of pregnancy rather than asthma chronicityor severity.The prevalence of adrenal insufficiency or suppressionin pregnancy is currently not well documented. This isin part due to the dynamic and significant increases ofcortisol during pregnancy potentially masking any deficita b cFig. 3 Bar graphs comparing median hair cortisol concentrations. The median hair cortisol concentrations (y-axis) at each time point (x-axis) forControls, No ICS, and ICS Treated are shown. Sample sizes for each time point are shown below the x-axis. a. Comparison inclusive of all womenin the ICS Treated group who reported ICS use captured in the hair sample, regardless of frequency or duration, showing no significant differenceamong the three groups. b Comparison excluding women with hair cortisol concentrations ≥ 276 pmol/g and women in the ICS Treated groupwho reported ICS use less than five doses per week showing a significant difference between Controls and ICS Treated women in T3. However,when the samples with concentrations ≥ 276 pmol/g were included, the comparison was no longer significant (c). *P ≤ 0.05, error bars indicatethe interquartile range, PC = preconception, T1 = first trimester, T2 = second trimester, T3 = third trimester, PP = postpartumSmy et al. BMC Pregnancy and Childbirth  (2016) 16:176 Page 7 of 10[21], and a lack of reference intervals specific to preg-nancy. Previous research has largely focused on the fetaland perinatal outcomes of fetal exposure to increased ma-ternal cortisol concentrations [30, 31], leaving a large gapin knowledge on the potential adverse effects of decreasedcortisol levels. Our study suggests that women withasthma during pregnancy experience adrenal suppression,which may represent adrenal hypo-functionality that isless severe than that of adrenal insufficiency but could po-tentially still impact pregnancy outcomes and normal fetalmaturation. Research shows that women with asthma,with or without ICS treatment, have similar perinatal out-comes to women with autoimmune adrenal insufficiency.Both groups are more likely to require a Cesarean sectionor have a preterm birth [32–34], potentially indicating alink with decreased cortisol levels. Our data do not sug-gest women with asthma experience preterm birth(Table 1), but our sample size may be too small to detect asignificant difference. Yet, the results of a recent publica-tion investigating the determinants of maternal hair corti-sol at delivery support an association between pretermbirth and decreased cortisol. Although they did not takeasthma or other diseases into consideration, Braig et al.[35] found that women who had premature babies(≤ 37 weeks of gestation) had significantly lower hair corti-sol levels in the three months prior to delivery (β = −0.16,P = 0.029). However, this significance did not remain whenthe regression model was adjusted for all other significantvariables considered by the researchers (mutually adjustedβ = −0.10, P = 0.157). Nevertheless, our research and thatof Braig et al. supports the premise that hair cortisol is auseful tool for future research to ascertain whether thereis an association between decreased maternal cortisol andpregnancy outcomes. We were unable to find any fur-ther research reporting on pregnancy outcomes forwomen with lower-than-normal cortisol productionsuggesting that the condition has possibly gone un-detected until now, thus indicating an advantage of haircortisol as a biomarker of adrenal function duringpregnancy.The surges in cortisol from fetal and maternal sourcesduring pregnancy both likely contribute to fetal maturation.Multiple mechanisms, such as the release of placental cor-ticotropin releasing hormone and adrenocorticotropic hor-mone, result in increased maternal cortisol as pregnancyprogresses, and the surge in cortisol in late pregnancy isinvolved in epigenetic processes that program fetalcardiovascular, neurologic, endocrine, and metabolic sys-tems [21, 31]. If the required surge in cortisol is dimin-ished, as was evident in our study in women with asthma,organ systems that rely on cortisol for maturation andprogramming may be adversely affected. Only one studyhas examined the long-term effects of asthma and asthmatreated with ICS on childhood disease. The DanishNational Birth Cohort followed children to 6 years of ageand found that children exposed to ICS in utero weremore likely to experience ‘endocrine, metabolic disorders’(hazard ratio (HR) = 1.84, CI95% 1.13–2.99) and digestivesystem diseases (HR = 1.54, CI95% 1.18–2.02) [36]. Second-ary analyses assessing the effects of maternal asthma, com-bining those with and without ICS treatment, revealed anincreased risk of diseases of the respiratory system (HRadj =1.43, CI95% 1.34–1.52), nervous system (HRadj = 1.43, CI95%1.18–1.73), and digestive system (HRadj = 1.17, CI95% 1.04–1.32) [37], all of which rely on cortisol for proper fetal mat-uration or programming [15, 30, 31]. Although our presentstudy was not designed to draw conclusions on the effect ofdecreased cortisol levels and disease risk in children bornto women with asthma, the growing evidence may warrantfurther research in this area.Limitations of our study include its relatively smallsample size and possible discrepancies in how the hairwas collected (e.g., scalp location, distance from scalp).Given the observed expected change in hair cortisolconcentrations over the course of pregnancy, any in-accuracy due to improperly collected samples is not ob-vious, nor expected to be significant. Additionally, allmedication use was self-reported, either to the studypersonnel or a healthcare provider. Generally, womentend to reduce their ICS use during pregnancy [38, 39],and some women in our study reported this, but issueswith recall may affect the reporting of ICS and othermedications. Finally, a positive bias could have occurredfor the recalculated cortisol concentrations that werebelow the method quantitation limit. This predomin-antly affected the No ICS group for PC, T1, and T2 andmay have reduced potential differences between thisgroup and the Controls. Although the degree of biascannot definitively be determined, future studies involv-ing a larger number of women would be beneficial and re-solve any uncertainty.ConclusionsOur findings suggest that hair cortisol may be a usefulbiomarker of HPA axis function during pregnancy andsensitive enough to detect the effects of asthma, both withand without ICS treatment, on systemic cortisol levels.Using hair cortisol analysis, we are the first to show thatpregnant women with asthma are potentially unable tomount the expected cortisol response seen in later preg-nancy regardless of ICS use. Perinatal outcomes that areknown to be associated with maternal asthma may thusbe a result of decreased maternal cortisol that adverselyimpacts fetal maturation and epigenetic programming.Building upon our current work, future research on theeffects of maternal cortisol levels on pregnancy outcomescould benefit from using hair cortisol analysis as an assess-ment tool.Smy et al. BMC Pregnancy and Childbirth  (2016) 16:176 Page 8 of 10Additional fileAdditional file 1: Table S1. Table of correlations between hair cortisolconcentrations and potential confounding factors. (PDF 88 kb)AbbreviationsBMI, body mass index; cm/mo = centimetres per month; EIA, enzymeimmunoassay; HPA, hypothalamic-pituitary-adrenal axis; HR, hazard ratio; ICS,inhaled corticosteroids; IQR, interquartile range; n, sample size; Padj, adjustedp-value; PC, preconception; PP, postpartum; PSS, Perceived Stress Scale; SD,standard deviation; T1, first trimester; T2, second trimester; T3, third trimesterAcknowledgementsWe would like to thank the staff at St. Michael’s Hospital Obstetric Clinic andthe Motherisk Program for kindly aiding with the participant recruitmentprocess. Additionally, we would like to thank Rachel Zabel for her assistancewith performing the hair cortisol analysis. LS’s education is supported by theUniversity of Toronto Scace Graduate Fellowship, Ontario GraduateScholarship, and a Canadian Pharmacogenomics Network for Drug SafetyTraining Award.FundingFunding for this study is provided by The Canadian Institutes of HealthResearch Drug Safety and Effectiveness Network. The funding body did nothave any role in the design of the study; collection, analysis, andinterpretation of the data; or writing of the manuscript.Availability of data and materialsData and materials will not be shared as per Research Ethics Board approvalto protect participant confidentiality.Authors’ contributionsGK, BC, and UA participated in project conception and grant writing. GKprovided project oversight and HB provided obstetrical oversight. AS(deceased) coordinated the project at the University of British Columbia. LScoordinated the project at The Hospital for Sick Children and St. Michael’sHospital. LS and KS recruited participants, collected data and hair samples,and wrote the manuscript. LS performed the hair sample analysis and dataanalysis. LS, KS, UA, HB, BC, and GK all participated in data interpretation.Finally, UA, HB, BC, and GK edited the manuscript. All authors, except AS,have read and approve of the final version of the manuscript.Competing interestsThe authors declare that they have no competing interests.Consent for publicationNot applicable.Ethics approval and consent to participateEthics approval was obtained from The Hospital for Sick Children ResearchEthics Board (#1000032684), St. Michael’s Hospital Research Ethics Board(#13-333), and University of British Columbia / Children’s and Women’sHealth Centre of British Columbia Research Ethics Board (#H11-02444).Informed written consent was obtained from all participants.Author details1Division of Clinical Pharmacology and Toxicology, The Hospital for SickChildren, Toronto, ON, Canada. 2Pharmaceutical Sciences, Leslie Dan Facultyof Pharmacy, University of Toronto, Toronto, ON, Canada. 3Child & FamilyResearch Institute, Vancouver, BC, Canada. 4Division of TranslationalTherapeutics, Department of Pediatrics, University of British Columbia,Vancouver, BC, Canada. 5University Institute of Clinical Chemistry, InselspitalBern University Hospital, University of Bern, Bern, Switzerland. 6Department ofObstetrics & Gynecology, St. Michael’s Hospital, Toronto, ON, Canada.7Pharmaceutical Outcomes Programme, Child & Family Research Institute,950 W 28th Avenue, Vancouver, BC V5Z 4H4, Canada.Received: 14 December 2015 Accepted: 12 July 2016References1. Global Initiative for Asthma: Global strategy for asthma management andprevention, 2015. http://www.ginasthma.org. Accessed 1 Dec 2015.2. Lipworth BJ. Systemic adverse effects of inhaled corticosteroid therapy: asystematic review and meta-analysis. Arch Intern Med. 1999;159:941–55.3. Todd GRG, Cerini CL, Ross-Russell R, Zahra S, Warner JT, McCance D. Surveyof adrenal crisis associated with inhaled corticosteroids in the UnitedKingdom. Arch Dis Child. 2002;87:457–61.4. Sauve B, Koren G, Walsh G, Tokmakejian S, Van Uum SH. Measurement of cortisol inhuman hair as a biomarker of systemic exposure. Clin Invest Med. 2007;30:E183–91.5. van Holland BJ, Frings-Dresen MH, Sluiter JK. Measuring short-term andlong-term physiological stress effects by cortisol reactivity in saliva and hair.Int Arch Occup Environ Health. 2012;85:849–52.6. Kintz P, editor. Analytical and practical aspects of drug testing in hair. BocaRaton: Taylor & Francis Group, LLC; 2007.7. Groeneveld MG, Vermeer HJ, Linting M, Noppe G, van Rossum EF, vanIJzendoorn MH. Children’s hair cortisol as a biomarker of stress at schoolentry. Stress. 2013;16:711–5.8. Luo H, Hu X, Liu X, Ma X, Guo W, Qiu C, et al. Hair cortisol level as abiomarker for altered hypothalamic-pituitary-adrenal activity in femaleadolescents with posttraumatic stress disorder after the 2008 Wenchuanearthquake. Biol Psychiatry. 2012;72:65–9.9. Steudte S, Kolassa IT, Stalder T, Pfeiffer A, Kirschbaum C, Elbert T. Increasedcortisol concentrations in hair of severely traumatized Ugandan individualswith PTSD. Psychoneuroendocrinology. 2011;36:1193–200.10. Thomson S, Koren G, Fraser LA, Rieder M, Friedman TC, Van Uum SH. Hairanalysis provides a historical record of cortisol levels in Cushing’s syndrome.Exp Clin Endocrinol Diabetes. 2010;118:133–8.11. Manenschijn L, Koper JW, van den Akker EL, de Heide LJ, Geerdink EA, deJong FH, et al. A novel tool in the diagnosis and follow-up of (cyclic)Cushing’s syndrome: measurement of long-term cortisol in scalp hair. J ClinEndocrinol Metab. 2012;97:E1836–43.12. Smy L, Shaw K, Smith A, Russell E, Van Uum S, Rieder M, et al. Hair cortisolas a novel biomarker of HPA suppression by inhaled corticosteroids inchildren. Pediatr Res. 2015;78:44–7.13. Kwon HL, Belanger K, Bracken MB. Asthma prevalence among pregnant andchildbearing-aged women in the United States: estimates from nationalhealth surveys. Ann Epidemiol. 2003;13:317–24.14. Hodyl NA, Stark MJ, Osei-Kumah A, Bowman M, Gibson P, Clifton VL. Fetalglucocorticoid-regulated pathways are not affected by inhaled corticosteroiduse for asthma during pregnancy. Am J Respir Crit Care Med. 2011;183:716–22.15. Liggins GC. The role of cortisol in preparing the fetus for birth. Reprod FertilDev. 1994;6:141–50.16. Wosu AC, Valdimarsdottir U, Shields AE, Williams DR, Williams MA. Correlatesof cortisol in human hair: implications for epidemiologic studies on healtheffects of chronic stress. Ann Epidemiol. 2013;23:797–811. e2.17. Cohen S, Williamson GM. Perceived stress in a probability sample of theUnited States. In: Spacapan S, Oskamp S, editors. The Social Psychology ofHealth. Newbury Park: Sage; 1988. p. 31–67.18. Pereg D, Gow R, Mosseri M, Lishner M, Rieder M, Van Uum S, et al. Hair cortisoland the risk for acute myocardial infarction in adult men. Stress. 2011;14:73–81.19. Russell E, Kirschbaum C, Laudenslager ML, Stalder T, de Rijke Y, van RossumEF, et al. Toward standardization of hair cortisol measurement; results of thefirst international inter-laboratory round robin. Ther Drug Monit. 2014;37:71–5.20. Mikkelsen AL, Felding C, Hasselbalch H. Urinary free cortisol duringpregnancy. Acta Obstet Gynecol Scand. 1984;63:253–6.21. Kamoun M, Mnif MF, Charfi N, Kacem FH, Naceur BB, Mnif F, et al. Adrenaldiseases during pregnancy: pathophysiology, diagnosis and managementstrategies. Am J Med Sci. 2014;347:64–73.22. Suri D, Moran J, Hibbard JU, Kasza K, Weiss RE. Assessment of adrenalreserve in pregnancy: defining the normal response to the adrenocorticotropinstimulation test. J Clin Endocrinol Metab. 2006;91:3866–72.23. D’Anna-Hernandez KL, Ross RG, Natvig CL, Laudenslager ML. Hair cortisol levelsas a retrospective marker of hypothalamic-pituitary axis activity throughoutpregnancy: comparison to salivary cortisol. Physiol Behav. 2011;104:348–53.24. Kirschbaum C, Tietze A, Skoluda N, Dettenborn L. Hair as a retrospectivecalendar of cortisol production-Increased cortisol incorporation into hair inthe third trimester of pregnancy. Psychoneuroendocrinology. 2009;34:32–7.25. Krumbholz A, Anielski P, Reisch N, Schelling G, Thieme D. Diagnostic valueof concentration profiles of glucocorticosteroids and endocannabinoids inhair. Ther Drug Monit. 2013;35:600–7.Smy et al. BMC Pregnancy and Childbirth  (2016) 16:176 Page 9 of 1026. Tian R, Hou G, Li D, Yuan TF. A possible change process of inflammatorycytokines in the prolonged chronic stress and its ultimate implications forhealth. ScientificWorldJournal. 2014;2014:780616.27. Clifton VL, Murphy VE. Maternal asthma as a model for examining fetalsex-specific effects on maternal physiology and placental mechanisms thatregulate human fetal growth. Placenta. 2004;25(Suppl A):S45–52.28. Robson AO, Kilborn JR. Studies of adrenocortical function in continuousasthma. Thorax. 1965;20:93–8.29. Miller GE, Chen E. Life stress and diminished expression of genes encodingglucocorticoid receptor and beta2-adrenergic receptor in children withasthma. Proc Natl Acad Sci U S A. 2006;103:5496–501.30. Moisiadis VG, Matthews SG. Glucocorticoids and fetal programming part 1:outcomes. Nat Rev Endocrinol. 2014;10:391–402.31. Moisiadis VG, Matthews SG. Glucocorticoids and fetal programming part 2:mechanisms. Nat Rev Endocrinol. 2014;10:403–11.32. Björnsdottir S, Cnattingius S, Brandt L, Nordenström A, Ekbom A, Kämpe O,et al. Addison’s disease in women is a risk factor for an adverse pregnancyoutcome. J Clin Endocrinol Metab. 2010;95:5249–57.33. Murphy VE, Namazy JA, Powell H, Schatz M, Chambers C, Attia J, et al. Ameta-analysis of adverse perinatal outcomes in women with asthma. BJOG.2011;118:1314–23.34. Wen SW, Demissie K, Liu S. Adverse outcomes in pregnancies of asthmaticwomen: results from a Canadian population. Ann Epidemiol. 2001;11:7–12.35. Braig S, Grabher F, Ntomchukwu C, Reister F, Stalder T, Kirschbaum C, et al.Determinants of maternal hair cortisol concentrations at delivery reflectingthe last trimester of pregnancy. Psychoneuroendocrinology. 2015;52:289–96.36. Tegethoff M, Greene N, Olsen J, Schaffner E, Meinlschmidt G. Inhaledglucocorticoids during pregnancy and offspring pediatric diseases: anational cohort study. Am J Respir Crit Care Med. 2012;185:557–63.37. Tegethoff M, Olsen J, Schaffner E, Meinlschmidt G. Asthma duringpregnancy and clinical outcomes in offspring: a national cohort study.Pediatrics. 2013;132:483–91.38. Enriquez R, Griffin MR, Carroll KN, Wu P, Cooper WO, Gebretsadik T, et al.Effect of maternal asthma and asthma control on pregnancy and perinataloutcomes. J Allergy Clin Immunol. 2007;120:625–30.39. Schatz M, Leibman C. Inhaled corticosteroid use and outcomes inpregnancy. Ann Allergy Asthma Immunol. 2005;95:234–8.•  We accept pre-submission inquiries •  Our selector tool helps you to find the most relevant journal•  We provide round the clock customer support •  Convenient online submission•  Thorough peer review•  Inclusion in PubMed and all major indexing services •  Maximum visibility for your researchSubmit your manuscript atwww.biomedcentral.com/submitSubmit your next manuscript to BioMed Central and we will help you at every step:Smy et al. BMC Pregnancy and Childbirth  (2016) 16:176 Page 10 of 10


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items