Skip to main content

Table 1 Summary of existing literature exploring the association between cortisol and perinatal depression

From: Perinatal maternal depression and cortisol function in pregnancy and the postpartum period: a systematic literature review

Authors

Subjects:

Design

Measurement of cortisol (IV)

Depression and stress reactivity measure (DV)

Relevant findings

1. Handley et al. (1977) [70]

N = 18 healthy pregnant women

Cohort Study

1 sample at 4 time points: 2, 3, 4, 5 days postpartum. Time: 9–9:30 am

Blood plasma, method of cortisol determination unspecified

MAACL, BDI, Hildreth Feeling Scale, the Blues Index devised by Handley et al. (1980).

Plasma cortisol was significantly and positively correlated with the Hildreth Feeling scale (R 2 = 0.12, p < 0.05).

2. Handley et al. (1980). [44]

N = 71 healthy pregnant women

Longitudinal study

1 sample at 9 time points: 36, 38 weeks gestation, 1 to 5 days postpartum, at delivery and 6 weeks postpartum. Time: 9:00–9:30 am

Blood plasma, method of cortisol determination not specified.

MAACL, VAS,BDI

Global Ranking Scale

Blues Index (Handley et al. 1980)

Cortisol was higher in “cases” and “severe cases” of depression than non-cases from 38 weeks gestation to 5 days post-partum. Global Ranking Scale: R 2 = 0.16, MAACL: R 2 = 0.14, VAS: R 2 = 0.19, BDI: R 2 = 0.12, Blues Index = N/S.

3.Balbi et al. (1980) [74]

N = 25 healthy pregnant women

Cohort Study

4 samples at 1 time point: on 4 days postpartum, a sample was retrieved every 6 hours for 24 hours.

Blood plasma, RIA

HAMD

The depressed group (n = 6) had significantly higher cortisol levels than controls.

4. Kuevi et al. (1983). [75]

N = 44 healthy pregnant women

Cohort Study

1 sample from 4 time points: 2,3,4,5 days postpartum and for N = 35, 2–3 hrs after the last breast feed. N = 18 also had an antenatal sample at 36 weeks gestation. Time: 10:00 am–12:00 pm

Blood plasma, RIA

A questionnaire including; self-rating mood scale, VAS, and questions on the frequency and duration of crying.

No significant relationship between mood and cortisol was found.

5.Brinsmead et al. (1985) [76]

N = 19 healthy pregnant women.

Cohort Study

1 sample from 3 time points: 36–38thweeks gestation, during labor, 4 days postpartum. Time not reported

Blood plasma, RIA

POMS, Caroll Depression Inventory, a set of 5 self-rated visual analogue scales

No significant association was found between maternal blues and cortisol.

6. Feksi et al. (1984). [77]

N = 5 postpartum primiparous women who experienced severe blues and 5 matched mothers who did not experience any depressive symptoms.

Case controlled pilot study

20 samples: Samples collected at 6 am, 12 pm, 6 pm and 10 pm daily, from day 1 to day 5 postpartum

Saliva, RIA

Semi-structured interview (Pitt, 1973).

VAS, DACL

No significant relationship between cortisol and mood was found.

7. Gard et al. (1986) [49]

N = 52 healthy pregnant women

Longitudinal study

1 samples at 2 time points: 36–38 weeks gestation and between 1–5 days postpartum. Time not reported.

Blood plasma, method of cortisol determination not specified.

MAACL, BDI, unidentified retrospective antenatal interview assessing mood.

No significant relationship between cortisol and mood was found.

8. Harris et al. (1989) [48]

N = 147 postpartum women

Cohort study

3 samples across 2 days at 6–8 weeks postpartum: between 1:30–3 pm, 10 pm and 8 am the next day.

Saliva and blood plasma, RIA.

EPDS, cut-off score: 12

Raskin 3 Area Depression Rating Scales, MADRAS

No significant relationship between cortisol and mood was found.

9. Ehlert et al. (1990). [71]

N = 70 postpartum women, 29 developed postpartum blues.

Longitudinal study

15 samples:

From the first day after delivery to 5 days postpartum, all women provided 3 saliva samples daily (8 am, 3 pm, 8 pm)

Saliva, RIA

BDI

DACL

BFS (adjective checklist)

Women who experienced postpartum blues showed significantly higher cortisol levels in the morning on days where symptoms were present, in comparison to women who did not experience the blues (t(59) = -2.35, p < 0.022, R 2 = 0.10)

10. Smith et al. (1990) [45]

N = 97 primiparous Australian women (28 weeks gestation)

Divided into women whose mood either improved (n = 46)or deteriorated (n = 36) from 38 weeks gestation to 2 days postpartum.

Longitudinal study

1 sample from 4 time points: 28 and 38 weeks’ gestation, during labour and day 2 postpartum. Time: between 8–10 am.

Blood plasma, RIA

POMS

MADRS

There were no significant differences in cortisol levels between groups (F = 1.75, p > 0.05).

11. O’Hara et al. (1991) [46]

N = 182 healthy pregnant women.

Longitudinal study

9 time points: At 34, 36, 38 weeks; 3× blood samples daily and 24 hour urine samples collected. 1× blood sample daily on 1, 2,3,4,6, and 8 days postpartum and 24 hour urine samples on 2& 4 days postpartum. Time: “before breakfast” On day 4 postpartum, an additional blood sample was drawn at 4 pm.

Blood plasma and urine, RIA

BDI, VAS, Maternal Blues Index (Handley et al., 1980).

No significant association was found between cortisol and mood.

12. Okano et al. (1992). [52]

N = 47 healthy pregnant women

Attrition rate: 19 %

Longitudinal study

1 sample from 3 time points: 30–41 weeks, 3rd/4th day postpartum and 1 month postpartum. Time: 10 am during pregnancy and 1 month postpartum, 7 am on 3rd/4th day postpartum.

Blood plasma, RIA

Semi-structured interview adapted from SADS

SRDS, Stein Scale for Maternity Blues

Cortisol levels were significantly higher 3–4 days postpartum in the “blues” group in comparison to those without depressive symptoms, R 2 = 0.18.

13. Pedersen et al. (1993). [50]

N = 12 healthy pregnant women

Case controlled study

I sample from 6 time-points: 38 weeks, and 1, 3, 6, 9 and 12 days postpartum. Time not reported.

Urine and blood plasma samples, method of cortisol determination not specified.

VMAS, CSI

HRSD

Morning serum cortisol levels were significantly higher 6 days postpartum in the group with depressive symptoms (via HRSD) than controls (R 2 = 0.08).

No significant differences in urinary cortisol between groups at any time point.

14. Taylor et al. (1994) [73]

N = 163 healthy postpartum women.

Cohort Study

1 sample, 1 time point: 3 days postpartum. Time: 10:30 am–12:00 pm

Blood plasma, RIA

The Kennerley Blues Scale

EPDS, cut-off:10

Cortisol levels were significantly higher in the blues group than non-blues group (as identified by the Kennerley Scale), R 2 = 0.07.

15. Harris et al. (1994). [78]

N = 130 primiparous, healthy pregnant women

Longitudinal study

Saliva: 8 am, 10 pm (2 weeks pre-term until 35–40 days postpartum daily) and additional 2 pm samples on 1, 2,3,4,5 days postpartum.

Blood samples: 1 sample at 2 weeks before delivery, 1, 5, and 35 days postpartum. Time not reported.

Saliva, RIA

EPDS (did not specify cut-off score), Stein Scale for Maternity Blues, BDI, MADRAS.

There were no significant associations between blues and cortisol (neither mean concentrations at the times of plasma sampling nor the decrements in concentrations from before delivery to day 5 postpartum)

16. Mahomed et al. (1995) [90]

N = 189 healthy pregnant, primiparous women

Prospective study

1 sample, 1 time point (cortisol): when in established labour. Time not reported.

Blood plasma, RIA

Pitts Depression Inventory.

No significant associations between mood and cortisol.

17.Harris et al. (1996) [47]

N = 130 healthy pregnant, primiparous women

Longitudinal study

Saliva: 8 am, 10 pm (2 weeks pre-term until 35–40 days postpartum daily) and additional 2 pm samples on 1, 2,3,4,5 days postpartum.

Blood samples: 1 sample at 2 weeks pre-term, 1, 5, and 35 days postpartum. Time not reported

Saliva and Blood, RIA

EPDS (did not specify cut-off score), MADRAS, Stein Scale for Maternity Blue, Raskin 3 Area Depression Rating Scales, a semi-structured interview for depression using DSM-III-R criteria for major depression.

Depressed women had significantly (p < 0.05) lower evening (10 pm) cortisol on pre-natal day 14 (using all measures), pre-natal day 1 (using the Raskin, MADRAS and semi-structured interview), pre-natal days 2–7 pooled and 3 days postpartum (using the Raskin & MADRAS), R 2 = 0.10–0.15.

18. Abou-Saleh et al. (1998). [79]

N = 61 women (23 pregnant women and 38 non-gravid controls). 3 groups: postpartum women, pregnant women and controls.

Cross-sectional study

1 sample, 1 time point: Cortisol measured 7 days postpartum, between 9–10 am

Serum cortisol, RIA

EPDS, cut-off score: 11

PSE

There was no significant relationship between cortisol and mood.

19.Lundy et al. (1999) [56]

N = 63 pregnant women (36 with depression)

Cortisol samples retrieved from a subsample of 43 (25 depressed, 18 non-depressed).

Case controlled study

1 sample at 2 time points: between 27–35 weeks gestation and shortly after term. Time: “morning hours”.

Urine (not 24 hr samples), method of cortisol determination not specified

CES-D, DIS

Depressed mothers had significantly higher prenatal cortisol levels than non-depressed mothers, F (1, 42) = 4.16, p < 0.05, R 2 = 0.03.

20. Susman et al. (1999). [80]

N = 59 pregnant healthy adolescents (13–19 year olds)

Longitudinal design

1 sample at 3 time points: early pregnancy (8–16 and 9–12 weeks), late pregnancy (32–34 weeks) and 3–4 weeks postpartum.

Time: 8:30 am.

Blood plasma, RIA

DISC-2.1(administered across all stages)

No significant relationship was found between cortisol and mood at any time-point.

21. Parry et al. (2003) [53]

N = 40, 20 depressed and 20 non-depressed postpartum women

Case controlled study

Every 30 minutes from 6 pm to 11 pm, sometime within the first 12 months postpartum.

Blood plasma, Unspecified

HRDS, BD1, EPDS (did not report cut-off score), SCID, VAS.

Hypocortisolemia was indicated in postpartum depressed women, in comparison to controls. Insufficient data to obtain effect size.

22. Field et al. (2004) [57]

N = 140 pregnant women (70 depressed, 70 non-depressed)

1 sample at 2 time points: Average. 20.1 weeks gestation, within 24 h following delivery. Time: “morning”

Urine samples, Unspecified

CES-D

Mothers with depressive symptoms had elevated cortisol levels in comparison to controls at 20.1 weeks (on average). R 2 = 0.05.

23. Diego et al. (2004). [94]

N = 80 pregnant women, 23–27 weeks gestation

1 sample at 2 time points: 23–27 weeks gestation, within 2 weeks postpartum. Time: 11 am–1 pm

Urine sample (not 24 hour), RIA

CES-D

Women expressing depressive symptoms during both pregnancy and postpartum and only during pregnancy had significantly higher cortisol levels than non-depressed women during mid gestation (R 2 = 0.19, R 2 = 0.31 respectively)

24. Field et al. (2006) [88]

N = 300 depressed pregnant women at approx. 20 weeks gestation

Cross-sectional study

1 sample, 1 time point: 20 weeks gestation, Time: “first morning urine sample”

Urine samples, RIA

CES-D, SCID

Cortisol significantly and positively associated with CES-D scores at 20 weeks gestation (F = 6.72, p = 0.01, R 2 = 0.02).

25. Nierop et al. (2006). [65]

N = 57 healthy multiparous pregnant women

Cross-sectional study

6 samples on a single day: Cortisol samples were measured immediately before and after the TSST and 10, 20, 45 and 60 minutes after testing. Time not reported.

Saliva samples, EIA

Trier Social Stress Test (TSST)

EPDS, cut-off score: 9

The group likely to develop depression had greater psychological reactivity to psychosocial stress and greater increases in cortisol levels. Cortisol over time × group effect: F (2.41, 25.74) =2.99, p = 0.04, R 2 = 0.05.

26. Groer et al. (2007) [54]

N = 25 depressed and 175 non-depressed mothers (at 4–6 weeks postpartum)

Case controlled study

1 sample at 1 time point: Between 4–6 weeks postpartum. Time: before 8 am for saliva and between 8–11 am for blood samples.

Saliva and blood plasma, EIA

POMS-D

Depressed mothers had significantly lower salivary cortisol levels than the control group (p < .05). Serum cortisol concentrations were not significantly different between groups. Insufficient data to obtain effect size.

27. Davis et al. (2007) [81]

N = 247 healthy pregnant women

1 sample at 4 time points: 19.1, 24.9, 30.8 weeks gestation, 8 weeks postpartum. Time: Mean 2:20 pm, SD: 1.5 hrs

Saliva, RIA

CES-D

No significant relationship between mood and cortisol was found.

28. Evans et al. (2008). [66]

N = 180 pregnant women at 36 weeks gestation.

Based on psychiatric diagnosis, 4 groups were formed: n = 121 controls, 16 depressed, 34 had anxiety, and 9 comorbid.

Case Controlled study

1 sample at 3 time points (between 33–39 weeks); upon presentation of task (baseline), before the psychophysiology session started (anticipation) and after the session (reaction). Time: 10:30–11:30 am.

Blood serum, RIA

SCID, CES-D, PES

Psychophysiology task: Stroop task, mental arithmetic task or controlled breathing task.

Women with co-morbid depression and anxiety had higher salivary cortisol levels than controls (p = 0.01). However those with either depression or anxiety alone did not differ significantly from controls.

29. Field et al. (2008) [59]

N = 430 healthy pregnant women

1 sample at 3 time points: approx. 22 and 32 weeks gestation, 2 days postpartum. Time: “mid-morning”

Urine sample, RIA

SCID, CES-D

At 22 weeks gestation, depressed women (as identified by the SCID) had higher cortisol levels than non-depressed women. Insufficient data provided to calculate effect size.

30. Fan et al. (2009). [87]

N = 308 pregnant or recently delivered women.

n = 77 each in 4 groups, representing each trimester and 1 week postpartum.

Cross-sectional study

1 sample at 3 time points: each group (trimester), between 9–10 am

Blood/serum samples, RIA

HAMD, SCL-90

No significant relationship between cortisol and mood was found.

31. Figueiredo et al. (2009). [82]

N = 91 healthy pregnant, primiparous women

Longitudinal

1 sample at 2 time points: Between 21–28 weeks and 3 months postpartum. Time not reported

24-hour urine samples, EIA

EPDS, cut-off score: 10

Cortisol was not a significant predictor of maternal depression

32. Yim et al. (2009). [93]

N = 100 healthy pregnant women.

Longitudinal study

1 sample at 5 time points: Blood samples were obtained at 15.3, 19.2, 25.0, 31.0 and 36.7 week’s gestation. Time not reported.

Blood plasma, RIA

CES-D, EPDS, cut-off score: 10

At no time during pregnancy were cortisol levels associated with concurrent depressive symptoms or postnatal depression (p > .53 for all comparisons)

33. Diego et al. (2009) [63]

N = 80 pregnant women (40 depressed, 40 non-depressed)

Longitudinal study

1 sample at 1 time point: between 18–20 weeks gestation. Time: “mid-morning”

Urine, RIA

SCID, CES-D

Depressed women had significantly higher prenatal cortisol concentrations than non-depressed women (determined by the SCID & CES-D), F (1, 74) =7.92, p = 0.006, R 2 = 0.14.

34. Cheng et al. (2010) [67]

N = 46 healthy pregnant women at or over 36 gestational weeks.

Longitudinal

2 samples at 2 time points: 36 weeks gestation and 4–6 weeks postpartum. Cortisol was collected at waking & 30 minutes after awakening.

Saliva, method of cortisol determination unidentified

CES-D

No significant relationship between prenatal or postnatal CAR and CES-D scores.

35. Taylor et al. (2009) [2]

N = 21 depressed and 30 non-depressed women at 7.5 weeks gestation

Cohort study

Samples obtained 30 min, 3 and 12 hours post-waking for 2 consecutive days at 7.5 weeks postpartum.

Saliva, EIA

EPDS, cut-off score: 13

Depressed women had a significantly reduced morning rise (at 30 minutes post-waking) in cortisol concentrations than controls, R 2 = 0.34.

36. Pluess et al. (2010). [68]

N = 66 healthy pregnant women

Longitudinal study

4 time points: 35 and 36th gestation weeks, 2 consecutive days during 10–12 weeks gestation. Samples obtained immediately, 30, 45 and 60 minutes after waking.

Saliva, EIA

EPDS, cut-off score: 13

No significant relationship between cortisol and CAR was found.

37. Parcells, D.A. (2010) [60]

N = 59 healthy pregnant women

Longitudinal Study

1 sample at 2 times points (cortisol): 26–28 and 32–34 weeks gestation. Time: between 10:00–11:30 am.

Saliva, STAT Fax

2100 microplate reader

SCID,BDI-II

No significant association between SCID diagnoses and cortisol. However, cortisol significantly differed between women with BDI-II scores greater than 12 and less than 12. Insufficient data to obtain effect size.

38.O’Keane et al. (2011). [83]

N = 70 healthy pregnant women

Longitudinal study

1 sample at 2 time points: 36 weeks and 3 days postpartum. Time: 11.00 am and 3 pm.

Blood plasma, EIA

EPDS, cut-off score:11

28 item Blues Questionnaire (Kennerley & Gath, 1989)

No significant association between cortisol and depression (antenatal or postnatal) was found.

39. Giesbrecht et al. (2012) [61]

N = 83 healthy pregnant women

Longitudinal Study

3 consecutive days between 6–37 weeks gestation with the following sampling schedule; upon waking, 30–45 min after waking and semi-randomly with the anchor times of 11:00 am, 4:00 pm, and 8:00 pm

Saliva, EIA

POMS-15, EDPS, did not specify cut-off score.

CAR and negative mood were significantly associated (after accounting for the diurnal variations across the 3 days), R 2 = 0.29.

40. Tsubouchi et al. (2011) [84]

N = 69 healthy pregnant women

1 sample at 5 time points: 1sttrimester (10–12 weeks), 2ndtrimester(20–22 weeks), early 3rdtrimester(30–32 weeks), late 3rdtrimester(37–39 weeks) and 1 month postpartum. Time: between 9:00 am and 1 pm.

Saliva, EIA.

Zung self-rating depression scale (cut off score: 42)

General Health Questionnaire -28

Participants identified as “chronically stressed” had lower cortisol levels during the 2nd and 3rd trimesters than controls. However no significant difference was found in the 1st trimester or postpartum. Insufficient data to obtain effect size.

41. Salacz et al. (2012) [89]

N = 79 pregnant women in their 36–38th gestational week

Cross-sectional study

1 sample at 1 time point: 36–38th gestational week, before 8 am.

Blood plasma, RIA

BDI-IA

No significant relationship between cortisol levels and mood found

42. Voegtline et al. (2013) [65]

N = 112 pregnant women between 24 and 38 weeks gestation.

Longitudinal study

1 sample at 5 time points: 24–26 weeks, 27–29 weeks, 30–32 weeks, 33–35 weeks and 36–38 weeks. Time: between 1–3 pm.

Saliva, EIA

CES-D

Women who reported more depressive symptoms between 30–32 weeks had higher cortisol levels than controls, R 2 = 0.05, p < 0.05).

43. Peer et al. (2013) [69]

N = 78 healthy pregnant Canadian immigrant women

4 times per day for 2 consecutive days at 19 weeks gestation: immediately post-waking, 30 and 60 minutes post-waking (CAR). Time: between 9:00 pm–10:00 pm.

Saliva, EIA

EPDS, cut-off score: 12.

Evening cortisol levels were significantly higher in women with high levels of depressive symptoms (n = 8) than those with low levels of depressive symptoms (n = 45). There were no significant differences for CAR., R 2 = 0.17.

44. Shelton et al. (2014)

N = 105 healthy pregnant women

Cohort Study

1 time point: between 16 and 26 weeks gestation. Time: “before noon” (mean time = 11:25 am).

Blood plasma, EIA

POMS-D

There was no significant relationship between POMS-D scores and cortisol, R 2 = 0.02.

45. O’Connor et al. (2014) [62]

N = 101 healthy pregnant women

Longitudinal Study

CAR was measured using five samples collected at; upon waking, 45 min, 2.5 hrs, 8 hrs and 12 hrs post-waking.

Two CAR measurements: on average, at 21.26 and 34.15 weeks gestation.

Saliva, EIA

EPD (did not specify cut-off scores), SCID

SCID diagnosis of depression were significantly and negatively associated with cortisol upon initial waking. Insufficient data to obtain effect size.

46. Luiza et al. (2015) [91]

N = 50 healthy pregnant women recruited at approx. 11 weeks gestation

Case-Controlled Study

1 sample at 1 time point: urine and blood samples collected between 6–16 weeks gestation. Time: not reported for blood samples & urine samples were obtained “first thing in the morning”.

Urine and Blood plasma, EIA.

EDS, cut-off score: 11

There was no significant relationship between cortisol and EDS scores.

47. Shimizu et al. (2015) [104]

N = 65 healthy Japanese postpartum women.

Cohort study

1 sample at 2 time points: 1 month and 4 months postpartum. Time not reported.

Urine samples, Unspecified.

EPDS (Japanese version), cut-off score: 8-9

There was no significant relationship between cortisol and EPDS scores.

  1. This table lists and provides details of existing literature examining the association between cortisol and perinatal depression
  2. Abbreviations: BDI-IA Becks Depression Inventory revised, CSI Childcare Stress Inventory, CES-D Centre for Epidemiological Studies Depression Scale, DACL Depressive Adjective Check List, DIS Diagnostic Interview Schedule, DISC- 2.1 Diagnostic Interview Schedule for Children, EIA enzyme immunoassay, EPDS Edinburgh Postnatal Depression Scale, EDS Edinburgh Depression Scale, HAMD Hamilton Rating Scale for Depression, HRSD Hamilton Rating Scale for Depression, LES Life Experiences Survey, MAACL Multiple Affect Adjective Checklist, MADRAS Montgomery-Asberg Depression Rating Scale, POMS Profile of Mood States, PES Pregnancy Experiences Scale, PSE Present State Examination, RIA radioimmunoassay, SCID Structured Clinical Inventory for DSM Disorders, SCL-90 Symptom Checklist-90, SADS Schedule for Affective Disorders and Schizophrenia, SRDS Zung Self-rating Depression Scale, VAS Visual Analogue Scale for Mood and Anxiety, VMAS Visual Analogue Mood Scales, CAR Cortisol Awakening Response, TSST Trier Social Stress Test