Skip to main content
NCBI home page
Indian Journal of Psychiatry logoLink to Indian Journal of Psychiatry
. 2022 Oct 12;64(5):440–448. doi: 10.4103/indianjpsychiatry.indianjpsychiatry_561_21

A systematic review and meta-analysis of serum and plasma cortisol levels in depressed patients versus control

Manoj K Sahu 1, Rajesh K Dubey 1, Alka Chandrakar 1, Mahesh Kumar 1, Mahendra Kumar 1,✉
PMCID: PMC9707655  PMID: 36458076

Abstract

Depression is associated with hyperactivity of the hypothalamo pituitary adrenal axis. Cortisol is a steroid hormone, released from the adrenal gland and is considered to be a biological marker of stress and anxiety. Serum or plasma cortisol levels have been previously studied in depressive patients but reported contradictory results. The present meta analysis aims to assess the serum or plasma concentration of cortisol in depressive patients compared with controls. We have conducted a systematic review with sequential meta analysis according to the Preferred Reporting Items for Systematic Reviews and Meta Analyses (PRISMA) guidelines. Web of Science, PubMed, Scopus, and PsycINFO databases, and published reference lists were searched up to January 2021. We have conducted a systematic review on PubMed for the following search (MeSH) terms (“Hydrocortisone”[Mesh]) AND “Depressive Disorder”[Mesh]). The RevMan 5.3 and Open Meta Analyst software was used with the standard mean difference (SMD) and 95% confidence intervals (CIs). The Jamovi and Open Meta Analyst Software were used to evaluate the publication bias, sensitivity analysis, and meta regression as possible sources of heterogeneity. Seventeen studies having a combined population (n) of 1400 (743 depressive patients and 657 controls) had satisfied the inclusion criteria for serum or plasma cortisol. The pooled SMD of the serum or plasma cortisol levels in depressive patients compared with controls was 1.18, (95% CI: 0.84, 1.52; P < 0.00001) with I2 = 85% (Ph < 0.00001). This meta analysis indicates a statistically significant mean difference in serum or plasma cortisol between depressed patients and controls. Meta analysis concluded that serum or plasma cortisol can differentiate depressed patients from nondepressed controls.

Keywords: Depression, plasma cortisol, serum cortisol

INTRODUCTION

Depression is a common and morbid psychiatric illness. It is a leading cause of disability. Worldwide more than 264 million people are affected by depression.[1] The World Health Organization has reported major depressive disorder as the second leading cause of global YLDs (Years Lived with Disability). Severe depression may significantly decrease quality of life and lead to death by suicide,[2] for which the risk rate is 6% to 15% among patients with this affective disorder. Various studies indicate that patients with premorbid history of depression are more likely to have depression in the context of cardiovascular disease.[3,4,5,6]

Cortisol is a steroid hormone of glucocorticoid class and its plays an important role in the development and expression of behavioral changes in human. The hypothalamic-pituitary-adrenal (HPA) axis has been the most studied of all endocrine axes in psychiatry. Depression is well-recognized to be associated with hypercortisolemia as a result of hyperfunctioning of hypothalamo-pituitary-adrenal axis.[7,8] Increased cortisol levels are a well-known response to stress. HPA axis abnormalities play important role in the pathophysiology of mood disorders.[9,10] Stress occurs due to negative life events and it can contribute to the development of depression; this link is bidirectional.[11] Cortisol has been considered a biological marker of stress and anxiety, and cortisol levels have been studied in relation to psychological changes in patients.

Cortisol can be measured in three ways: blood, saliva, or urine. A systematic review and meta-analysis on salivary cortisol levels were conducted by Knorr et al.[12] in depressed patients versus control and reported that salivary cortisol is unable to discriminate between persons with and without depression. Serum or plasma cortisol levels have been previously studied in depressive patients. Studies have reported contradictory results on serum and plasma cortisol levels in depressive patients. In the review of literature, we could not find any systematic review and meta-analysis on serum and plasma cortisol levels in depressed patients versus control persons. Also, many studies have preferred salivary cortisol levels due to its noninvasiveness and direct proportionality with the serum/plasma cortisol levels. Few studies have also mentioned that saliva cortisol cannot always replace serum cortisol levels due to error in accuracy when serum cortisol level is only marginally high.[13] Hence, we have included serum/blood cortisol levels to be more accurate in our findings and other factors influencing the results. Hence, we have focused on studies done on plasma and serum cortisol levels. The main aim of the current systematic review and meta-analysis is to find whether serum or plasma cortisol level differs in depressive patients compared with controls without depression.

MATERIAL AND METHODS

Protocol and search strategies

This study was done according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [Figure 1].[14] The databases of Web of Science, PubMed/Medline, Scopus, Psycinfo, and other databases were comprehensively searched manually with key terms (Serum or plasma cortisol) AND (Depression) AND (Serum or plasma cortisol level in Depressive Patients), before dexamethasone serum or plasma cortisol and depression, corticosteroid*, cortisone* AND depressive disorder, mood, affective, bipolar up to January 2021 to January 1990, only with the English language. We have conducted a systematic review on PubMed (database) for the following search (MeSH) term (“Hydrocortisone”[MeSH]) AND “Depressive Disorder”[MeSH]). Reference lists and citing articles of the included article were reviewed to identify other relevant studies. The research protocol was registered in the International Prospective Register of Systematic Reviews (https://www.crd.york.ac.uk/prospero/display_record.php?ID = CRD42021233603, PROSPERO: CRD42021233603).

Figure 1.

Figure 1

Open in a new tab

The PRISMA flow diagram of the study selection

Study selection

Initially two reviewers contributed to the study selection. Eligible studies from January 1990 to January 2021 were comprehensively searched by conducting a literature search on Web of Science, PubMed/Medline, Scopus, and Psycinfo. The first reviewer had independently screened the abstract of studies to check if they followed the inclusion criteria and the second reviewer had re-evaluated.

The inclusion criteria were:

  • (i)

    Detection of cortisol levels in the serum or plasma (fasting state and before diurnal changes) of the depressive disorder patients and control groups (case-control or comparative, cross-sectional, prospective, and experimental studies).

  • (ii)

    The depressive patient’s diagnosis was in accordance with the clinical and/or diagnostic and statistical manual of mental disorders (DSM) and International Classification of Diseases (ICD) (World Health Organization [WHO]) criteria.

  • (iii)

    The controls were reported without depression and other severe diseases.

Exclusion criteria

Studies were excluded when:

  • (i)

    The HPA axis was pharmacologically challenged before first cortisol measurement.

Data extraction

One reviewer had extracted the relevant data of every study. The second reviewer had rechecked them. The extracted data from all studies included in the meta-analysis are shown in Table 1.

Table 1.

Characteristics of 17 studies involved in the meta-analysis for serum or plasma cortisol

The first author Year Country or Place Study type/design No. of patients/controls Mean age of patients/controls Measurement method Time Score*
Thirthalli, et al.[8] 2013 Asia Open labeled study 58/18 54.3/34.0 DSM IV, HDRS, ELISA (ng/mL) 8:00 AM 7
Shabaan, et al.[15] 2015 Egypt Case control study 30/30 43.5/42.8 DSM V, HDRS, ELISA _ 8
Ahmed, et al.[16] 2017 Egypt Case control study 20/20 31.00/32.50 DSM IV TR, Beck’s Depression Inventory, IMMULITE 1000 analyzers by Siemens (Siemens Medical Solutions Diagnostics: Los Angeles, California, USA). (µg/dL) 8:00 to 9:00 AM 6
Theodor, et al.[17] 2016 Europe Comparative study 15/15 44.27/39.6 DSM – IV, NA, NA (µg/dL) 8:00 AM 6
Martinac, et al.[18] 2017 Europe Cross sectional study 49/40 49.43/47.20 MINI questionnaire, HDRS, Lumiimunochemical method (Roche Diagnostics GmbH, Germany). (nmol/L) 8:00 AM 8
Papakostas, et al.[19] 2013 US America Pilot Study 36/43 42.5/30 DSM IV, HDRS, ELISA (µg-mL−1) 9:00 to 17:00 h 7
Cizza, et al[20] 2012 US America Experimental 92/44 36.0/35.3 DSM IV, HDRS, (µg/dL) 8:00 AM 7
Islam, et al[21] 2018 Asia Case control study 247/248 33.03/33.55 DSM V, NA, ELISA (µg/dL) 10:00 AM and 12:00 PM 7
Bakheet, et al.[22] 2013 Asia Case control study 21/20 43.5/42.8 DSM IV, MADRS, ELISA (µg/dL) 8:00 AM 7
Li, et al.[23] 2013 Asia NA 15/15 34.7/32.9 Chinese Classification of Mental Disorders, HDRS, photoluminescence method (nmol/L) Daytime 7
Cubała, et al.[24] 2014 Poland Europe Cross sectional case control study 20/20 33.5/30.5 DSM IV, HDRS, Chemiluminescence microparticle Immunoassay (CMIA) using a commercially available kit (Architects Cortisol, Abbott Diagnostics, USA). (nmol/L) 8:00 to 9:00 AM 7
Piwowarska, et al.[25] 2009 Poland Europe NA 08/21 53.5/34.56 DSM IV, HDRS, high performance liquid chromatography (HPLC) (ng/mL) Morning 6
Rudzki, et al.[26] 2017 Poland Europe Exploratory research 34/29 42.26/42.79 DSM IV R, HDRS, ELISA (IBL International, Germany: Catalog number RE52061). (µg/mL) 8:00 to 9:00 AM 7
Tang, et al[27] 2019 Australia Cross s ectional 60/60 31.90/31.83 DSM V, ELISA kit (Abcam), with absorbance detection at 450 nm. (ng/mL) Morning 7
Jiang, et al[28] 2000 Taiwan Experimental 12/08 35.4/36.7 DSM III R, HDRS, radioimmunoassay (RIA) kit (solid phase Coat A Count, RBI/Sigma, Natick, MA, USA) (µg/dL) Morning 7
Osran, et al[29] 1993 Presented at the 143rd annual meeting of the APA, New York Brief report 9/9 34.8/34.1 DSM III R, HDRS, Radioimmunoassay (µg/dL) 8 AM 6
Bertollo, et al.[30] 2020 Brazil America Cross sectional 17/17 53/48 Beck’s depression inventory, immunoenzymatic assay was performed using automatic chemiluminescence equipment, model Immulite 2000, (mcg/dL) Morning 7
Open in a new tab

ELISA, enzyme-linked immunosorbent assay; HDRS, Hamilton Depression Rating Scale; MADRS, Montgomery–Asberg Depression Rating Scale; NA, not available; * Newcastle–Ottawa Quality Assessment Scale (NOS) score

Quality evaluation

Reviewers had estimated the quality of each study included. We have applied Newcastle–Ottawa Quality Assessment Scale (NOS) with score ≥7 being high quality.[31] Quality assessment of the studies was based on the selection of study groups, comparability of groups, and ascertainment of exposure/outcome.

Data analysis

Data were analyzed with the Review Manager 5.3 software (RevMan 5.3, The Cochrane Collaboration, Oxford, United Kingdom). We had computed the standard mean difference (SMD) on 95% confidence intervals (CIs) with random effects. We had analyzed the estimation of variance for serum or plasma cortisol levels between the depressed patients versus control group. P values less than 0.05 and 0.01 levels were considered statistically significant. The I2 statistic was calculated for the assessment of total percentage variation, which can be attributed to heterogeneity rather than chance. With regard to the Q statistic, heterogeneity was determined if I2 >50. The Q-statistics was applied to examine whether there was more heterogeneity in the results than could be expected from chance alone between the studies. Within this data set, we examined the subgroup analysis whether there were systematic differences at the time of measurement, types of measurement (principle), geographical area of participants, and nature of study.

The jamovi. (Version 1.0) and JASP (Version 0.14.1)[32,33,34] were used to evaluate the publication bias among the studies. The potential publication bias was explored by using funnel plots, Egger’s test,[35] and trim-and-fill methods.[36] Open Meta-Analyst[37] Software was used for sensitivity analysis and meta-regression as possible sources of heterogeneity, such that P < 0.05 showed a significant existence of publication bias. Continuous moderating variables, such as publication year, and age, were analyzed with meta-regression analysis if the number of included studies was at least 10. The unit of measurement of cortisol was μg/dL, ng/mL, nmol/L, μg/mL, and μg · mL−1, in serum or plasma. The Z-test was used to show the significance of the pooled SMD. To figure out the results stability, a leave-one-out meta-analysis was performed for multiple meta-analyses by excluding one study from each analysis. It is common for studies to produce exaggerated effect sizes, which may distort the overall results. To find the influence of each study on the overall effect-size estimate leave-one-out meta-analysis is useful.

RESULTS

The selection of the studies was done based on Figure 1. Out of 1926 studies identified in the databases, 17 studies were selected and entered in the meta-analysis.

Characteristics of included studies

Initially, 1926 hits were identified. A total of 17 studies fulfilled the study entry criteria and finally, 17 studies with 1400 population were pooled (743 for depressive and 657 for control) [Figure 2]. Characteristics of 17 studies involved in the meta-analysis are shown in Table 1. The studies were published from 1993 to 2021 and included (743 depressive patients and 657 controls) for serum or plasma cortisol. Sample sizes ranged from 18 to 495. In most studies, depressive patients were diagnosed as per DSM[8,15,16,17,19,20,21,22,23,24,25,26,27,28,29] but in one study patients were diagnosed by Chinese Classification of Mental Disorders[23] and the other two study patients were diagnosed with MINI questionnaire[18] and Beck’s depression inventory.[30] In the meta-analysis for serum or plasma cortisol two studies were reported from Egypt,[15,16] five from Europe,[17,18,24,25,26] three from America or south America,[19,20,30] five from Asia,[8,21,22,23,28] one from Australia,[27] and one presented at the annual meeting of the American Psychiatric Association.[29] In seven studies, the detection method of serum and plasma cortisol levels was enzyme-linked immunosorbent assay (ELISA),[8,15,19,21,22,26,27] whereas the remaining methods were the radioimmunoassay in two studies,[28,29] chemiluminescence microparticle. Immunoassay in two studies,[24,30] IMMULITE 1000 analyzers by Siemens in one study,[16] photoluminescence method in one study,[23] Lumiimunochemical method in one study,[18] and high-performance liquid chromatography (HPLC) in one study.[25] But the detection method was not mentioned in two of the studies.[17,20] The unit of measurement of cortisol was “mcg/dL or μg/dL” in eight studies,[16,17,20,21,22,28,29,30] ng/mL in three studies,[8,25,27] nmol/L in three studies,[18,23,24] μg/mL in one study,[26] μg · mL −1[19] in one study for serum or plasma. Other data of the studies are shown in Table 1. According to the NOS criteria, 13 studies out of the 17 showed high quality, the score of quality assessment ranged from 6 to 8 [Table 1].

Figure 2.

Figure 2

Open in a new tab

Forest plot of standard mean differences in serum or plasma cortisol between patients with depression and controls in 17 studies with an overall pooled standard mean difference based on a random-effects meta-analysis

Meta-Analysis

Serum cortisol (Pooled standard mean difference, SMD)

In a random effects meta-analysis, the pooled standard mean difference (SMD) of the serum cortisol levels in depressive patients compared with controls group was 1.18, (95% CI: 0.84, 1.52; P < 0.00001) with I2 = 85% (Ph < 0.00001) [Figure 2].

Subgroup analysis

The subgroup analysis was done based on five characteristics: geographical area of participants, detection method of serum or plasma cortisol, nature of study, time of cortisol assessment, and unit measurement of cortisol [Table 2]. There was a significant difference in serum or plasma cortisol levels between depressive patients and controls in Asia (SMD = 1.68; 95% CI: 0.87, 2.49; P = 0.0001), Europe (SMD = 1.06; 95% CI: 0.45, 1.67; P = 0.0007), America or South America (SMD = 0.65; 95% CI: 0.15, 1.16; P = 0.01), and in Egypt (SMD = 1.14; 95% CI: 0.79, 1.49; P = 0.01). There was a significant difference in the serum or plasma cortisol levels between depressive patients and controls when performed with ELISA method (SMD = 1.53; 95% CI: 0.89, 2.17; P = 0.00001) and also with other measurement methods group (SMD = 1.69; 95% CI: 0.92, 2.45; P = 0.0001). There was a significant difference in the serum or plasma cortisol between depressive patients and controls in cross sectional, case control (or comparative) study group (SMD = 1.14; 95% CI: 0.88, 1.41; P = 0.00001), and in also with other method study group (SMD = 0.62; 95% CI: 0.28, 0.96; P = 0.0003). There was a significant difference in the serum or plasma cortisol between depressive patients and controls in only morning time assessment (SMD = 1.02; 95% CI: 0.71, 1.33; P = 0.00001). There was a significant difference in the serum or plasma cortisol between depressive patients and controls when performed with μg/dL or mcg/dL unit measurement group (SMD = 0.94; 95% CI: 0.53, 1.35; P = 0.00001), ng/mL unit measurement group (SMD = 1.78; 95% CI: 0.66, 2.90; P = 0.002), and nmol/L unit measurement group (SMD = 2.32; 95% CI: 0.48, 4.16; P = 0.01).

Table 2.

Random-effects analysis based on the area of participants, detection method of cortisol, nature of study, time of cortisol assessment, and unit measurement of cortisol in depressive patients vs. controls

Subgroups Number of Studies SMD (95% CI) P I2 P h
Method
 ELISA 7 1.12 (0.78, 1.45) <0.00001 75% 0.0004
 Others 8 1.69 (0.92, 2.45) <0.00001 86% <0.0001
Area
 Asia 5 1.68 (0.87, 2.49) <0.0001 88% <0.00001
 Europe 5 1.06 (0.45, 1.67) 0.0007 78% 0.001
 North America or South America 4 0.65 (0.15, 1.16) 0.01 68% 0.03
 Egypt 2 1.29 (0.27, 2.30) 0.01 81% 0.02
Nature of Study
Observational Nature 9 1.14 (0.88,1.41) <0.00001 59% 0.01
 Other 6 0.62 (0.28, 0.96) 0.0003 52% 0.06
Time of assessment
 Morning 14 1.02 (0.71, 1.33) <0.00001 79% <0.00001
 Other time 2 3.57 (−2.35, 9.50) 0.24 97% <0.00001
Measurement Unit
 ng/mL 3 1.78 (0.66, 2.90) 0.002 89% <0.0001
 µg/dL or Mcg/dL 8 0.94 (0.53, 1.35) <0.00001 77% <0.0001
 nmol/L 3 2.32 (0.48, 4.16) 0.01 94% <0.00001
Open in a new tab

Meta–Regression

Meta-regression analysis showed an insignificant statistical correlation of the pooled SMD of the serum or plasma cortisol levels with the year of publication, age of patients, and age of control. The result showed that the year of publication (Correlation coefficient: −0.028, P value 0.458) and age of patients (Correlation coefficient: −0.007, P value 0.843) or age of controls (Correlation coefficient: −0.034, P value 0.447) were not one of the reasons for heterogeneity.

Publication bias

The jamovi. (Version 1.0) was used to evaluate the publication bias among the studies by funnel plot by Rosenthal’s fail-safe and Kendall’s Tau tests. The funnel plot of the overall analysis is shown in Figure 3. Visual inspection of the funnel plot asymmetry and Rank correlation test suggests little evidence of publication bias [Figure 3]. Kendall’s Tau was significant at 0.05 level (Tau = 0.456, P = 0.011), which further indicates a presence of publication bias. Rosenthal’s fail-safe test suggested that there should have to be at least 1735 unpublished, nonsignificant comparisons to raise the overall P value to greater than 0.01, which is considered robust.

Figure 3.

Figure 3

Open in a new tab

Funnel plot of random effect analysis: Serum or plasma cortisol levels in the depressed patients vs. the controls

In addition, the statistical significance was assessed by the Egger’s regression asymmetry test. The Egger’s test identified evidence of significant publication bias (Regression test for Funnel plot asymmetry, z = 2.4449; P = 0.014). An adjusted analysis using “The fill and trim method” suggests a standard mean difference in the random-effects model is 1.37 (0.952 to 1.7989), which is statistically significant.

The level of serum or plasma cortisol in depressive patients was significantly higher than in controls in random effect. The largest weight of the studies of Islam et al.[21] was substantially decreased in a random-effects model of serum cortisol. In forest plot of random-effects analysis: the green box on each line shows the standard mean difference (SMD) for each study and the black diamond at the bottom of the graph shows the overall SMD of the seventeen studies.

Sensitivity analysis

The sensitivity analyses by removing each one study and cumulative analysis were done on overall analysis. The omission of study 4 (Cizza et al.)[20] or 7 (Li et al.)[23] seems to have a relatively larger influence (when compared with other studies) on the estimation of the overall effect size. Omitting study 4 causes the overall risk ratio to increase by roughly 0.07, whereas omitting study 7 causes the overall risk ratio to decrease by roughly 0.013 [see Figure 4].

Figure 4.

Figure 4

Open in a new tab

Forest plot of leave-one-out meta-analysis with random effect for standard mean difference in serum or plasma cortisol levels in the depressive patients vs. controls

DISCUSSION

To our knowledge, the current study is the first systematic review and meta-analysis done to examine the pooled SMD of serum and plasma cortisol levels between the depressive patients and controls. We have performed a broad search by including about 1900 records. The current study have accessed data from 17 studies and have evaluated 1400 participants for serum or plasma cortisol levels. The main finding of our current meta-analysis is that the serum or plasma cortisol level is significantly higher in patients with depressive disorder as compared with the controls. The findings of our current meta-analysis result are in line with the findings of the other study done recently.[12] Our findings are consistent with the previous studies, which have reported high levels of serum cortisol in depressive patients.[38]

We did further subgroup analysis with respect to five characteristics: 1) geographical area of the participants, 2) detection method of cortisol levels, 3) nature of study, 4) time of cortisol assessment, and 5) measuring unit of cortisol levels. The findings revealed that the Asian, European, American, South American, and Egyptian population had higher serum or plasma cortisol levels in depressive patients as compared with the controls. The differences were significant when the samples were processed by ELISA and various other methods. The difference was significant only when the cortisol samples were collected in the morning. The difference was significant when the cortisol measurement units like mcg/dL or μg/dL”, ng/mL, and nmol/L were used. The differences were significant in both the observational and interventional types of studies. The reason where we found insignificant results was probably due to the lesser number of studies done pertaining to the respective subgroup categories. The findings of our study are in line with findings of the other recent meta-analyses, which too have reported a lower number of studies hence influencing the confidence interval.[39]

The findings in our study have suggested important differences in serum or plasma cortisol levels. It is increased with 1.18 random-effects model of serum or plasma cortisol. Cortisol is a physiological parameter of stress and anxiety, and its association with psychiatric disorders among the depressive patients has been studied. The results of the study are in consistent with Wolkowitz et al.[38] who have also reported a high level of serum cortisol in patients with depression. On the other hand, the results are against the findings of Vythilingam et al.[40] who has reported low plasma cortisol levels in patients with depression. Hence, there are studies with controversial findings on the cortisol levels.[38,40] The current study results will add to the previous study findings. High cortisol concentrations have also been linked to other psychiatric illnesses like mania and delirium,[41] which are also associated with sleep disturbances, decreased attention, loss of libido, psychomotor disorders, anxiety, and suicidal ideations. Hypercortisolism is frequent in patients with major depression because depression causes disregulation of the HPA axis and reduction of hippocampal volume due to recurrent episodes of depression. Studies have also reported that in depression, adrenal androgens in contrast to cortisol, are partially regulated by mechanisms independent of adrenocorticotropic hormone.[29]

We found significant heterogeneity among study results as well as a substantial overlap between values for patients and controls, which may be explained by the differences in the geographical area of participants, collection technique, assessment time, methods and depression assessment tools. It is important to improve the study designs and include more homogeneous population and methods to reduce the heterogeneity in studies. Further studies are needed to investigate the cortisol level in depressive patients on the basis of updated classification systems DSM-5 and ICD-11. Many studies have preferred salivary cortisol levels due to its noninvasiveness and direct proportionality with the serum/plasma cortisol levels. Still, a few studies have mentioned that saliva cortisol cannot always replace serum cortisol levels due to errors in accuracy when serum cortisol level is only marginally high.[13] The current study had included serum/blood cortisol levels to be more accurate in our findings and other factors influencing the results. More longitudinal studies can be done, which can correlate the levels of cortisol with the improvement in depressive symptoms, establishing cortisol level as a prognostic biomarker.

The most important limitation is the significant heterogeneity in the selected studies. The systematic literature search was limited to articles published in English. Search strategy was designed to detect serum or plasma cortisol in depressive patients but in this search strategy broader terms were not included. The number of included studies is low, for this reason, other subgroup analyses with regard to gender, socioeconomic status, and some other measurement unit were not possible. The less number of studies also influenced the results of the meta-analysis. Only 14 studies with morning cortisol levels were included to reduce heterogeneity. The risk of bias for the quality assessment criterion was not assessed. This may also affect the standard mean differences.

CONCLUSIONS

The current meta-analyses indicate a statistically significant mean difference in morning serum or plasma cortisol between depressed patients and controls. A higher serum or plasma cortisol level was found in depressive patients compared with controls. Based on the findings of this meta-analysis, it can be concluded that serum or plasma cortisol level can differentiate depressed patients from nondepressed controls, it is a good indicator to differentiate between depressive patients and controls. The findings are consistent with previous studies reporting the relationship of depression with the functioning of the HPA axis.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.

REFERENCES

  • 1.GBD 2017 Disease and Injury Incidence and Prevalence Collaborators. Global, regional, and national incidence, prevalence, and years lived with disability for 354 diseases and injuries for 195 countries and territories, 1990-2017: A systematic analysis for the Global Burden of Disease Study 2017. Lancet. 2018;392:1789–858. doi: 10.1016/S0140-6736(18)32279-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Alves Vde M, Francisco LC, Belo FM, de-Melo-Neto VL, Barros VG, Nardi AE. Evaluation of the quality of life and risk of suicide. Clinics (Sao Paulo) 2016;71:135–9. doi: 10.6061/clinics/2016(03)03. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Shankman SA, Nadelson J, McGowan SK, Sovari AA, Vidovich MI. The predictive power of depression screening procedures for veterans with coronary artery disease. Vasc Health Risk Manag. 2012;8:233–8. doi: 10.2147/VHRM.S29424. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Caro MA, Sowden GL, Mastromauro CA, Mahnks S, Beach SR, Januzzi JL, et al. Risk factors for positive depression screens in hospitalized cardiac patients. J Cardiol. 2012;60:72–7. doi: 10.1016/j.jjcc.2012.01.016. [DOI] [PubMed] [Google Scholar]
  • 5.Mallik S, Spertus JA, Reid KJ, Krumholz HM, Rumsfeld JS, Weintraub WS, et al. Depressive symptoms after acute myocardial infarction:Evidence for highest rates in younger women. Arch Intern Med. 2006;166:876–83. doi: 10.1001/archinte.166.8.876. [DOI] [PubMed] [Google Scholar]
  • 6.Strik JJMH, Lousberg R, Cheriex EC, Honig A. One year cumulative incidence of depression following myocardial infarction and impact on cardiac outcome. J Psychosom Res. 2004;56:59–66. doi: 10.1016/S0022-3999(03)00380-5. [DOI] [PubMed] [Google Scholar]
  • 7.Shanmugasegaram S, Russell KL, Kovacs AH, Stewart DE, Grace SL. Gender and sex differences in prevalence of major depression in coronary artery disease patients:A meta-analysis. Maturitas. 2012;73:305–11. doi: 10.1016/j.maturitas.2012.09.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Thirthalli J, Naveen GH, Rao MG, Varambally S, Christopher R, Gangadhar BN. Cortisol and antidepressant effects of yoga. Indian J Psychiatry. 2013;55:405–8. doi: 10.4103/0019-5545.116315. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Holsboer F. Neuroendocrinology of mood disorders. In: Bloom FE, Kupfer DJ, editors. Psychopharmacology –4th Generation of Progress. New York: Raven Press; 1995. pp. 957–69. [Google Scholar]
  • 10.Plotsky PM, Owens MJ, Nemeroff CB. Psychoneuroendocrinology of depression:Hypothalamic-pituitary-adrenal axis. Psychiatr Clin N Am. 1998;21:293–307. doi: 10.1016/s0193-953x(05)70006-x. [DOI] [PubMed] [Google Scholar]
  • 11.Phillips AC, Carroll D, Der G. Negative life events and symptoms of depression and anxiety:Stress causation and/or stress generation. Anxiety Stress Coping. 2015;28:357–71. doi: 10.1080/10615806.2015.1005078. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Knorr U, Vinberg M, Kessing LV, Wetterslev J. Salivary cortisol in depressed patients versus control persons:A systematic review and meta-analysis. Psychoneuroendocrinology. 2010;35:1275–86. doi: 10.1016/j.psyneuen.2010.04.001. [DOI] [PubMed] [Google Scholar]
  • 13.Vieira-Correa M, Giorgi RB, Oliveira KC, Hayashi LF, Costa-Barbosa FA, Kater CE. Saliva versus serum cortisol to identify subclinical hypercortisolism in adrenal incidentalomas:Simplicity versus accuracy. J Endocrinol Invest. 2019;42:1435–42. doi: 10.1007/s40618-019-01104-8. [DOI] [PubMed] [Google Scholar]
  • 14.Moher D, Liberati A, Tetzlaff J, Altman DG PRISMA Group. Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement. PLoS Med. 2009;6:e1000097. doi: 10.1371/journal.pmed.1000097. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Shabaan MAS, Shaaban I, Alatram AR, Gad EF. The relationship of serum cortisol levels with severity of depression and age in major depression. Al-Azhar Assiut Med J. 2015;13:79–82. [Google Scholar]
  • 16.Ahmed S, Moussa F, Moustafa A, Ayoub DR. Cortisol level in depressed patients and its relation with suicidal risk and anhedonia. Egypt J Neurol Psychiatr Neurosurg. 2016;53:193–9. [Google Scholar]
  • 17.Theodor M, Iosif GG, Elena BG, Melinda F, Marieta GG, Andreea S, et al. Increased cortisol levels in depression:A comparative study evaluating the correlation of hypercortisolemia with prosocial coping mechanisms. Acta Med Marisiensis. 2016;62:68–72. [Google Scholar]
  • 18.Martinac M, Babić D, Bevanda M, Vasilj I, Glibo DB, Karlović D, et al. Activity of the hypothalamic-pituitary-adrenal axis and inflammatory mediators in major depressive disorder with or without metabolic syndrome. Psychiatr Danub. 2017;29:39–50. [PubMed] [Google Scholar]
  • 19.Papakostas GI, Shelton RC, Kinrys G, Henry ME, Bakow BR, Lipkin SH, et al. Assessment of a multi-assay, serum-based biological diagnostic test for major depressive disorder:A pilot and replication study. Mol Psychiatry. 2013;18:332–9. doi: 10.1038/mp.2011.166. [DOI] [PubMed] [Google Scholar]
  • 20.Cizza G, Mistry S, Nguyen VT, Eskandari F, Martinez P, Torvik S, et al. Do premenopausal women with major depression have low bone mineral density? A 36-month prospective study. PLoS One. 2012;7:e40894. doi: 10.1371/journal.pone.0040894. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Islam MR, Islam MR, Ahmed I, Moktadir AA, Nahar Z, Islam MS, et al. Elevated serum levels of malondialdehyde and cortisol are associated with major depressive disorder:A case-control study. SAGE Open Med. 2018;6:2050312118773953. doi: 10.1177/2050312118773953. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Bakheet MS. Serum cortisol levels in depression patients. Int J Biomed Sci. 2013;1:005–008. [Google Scholar]
  • 23.Li H, Gao Z, Wu Q, Huang P, Lin C, Chen G. Relationship of hypothalamus-pituitary-adrenal (HPA) axis function and suicidal behavior in patients with depression. Shanghai Arch Psychiatry. 2013;25:32–9. doi: 10.3969/j.issn.1002-0829.2013.01.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Cubałan WJ, Landowski J. C-reactive protein and cortisol in drug-naïve patients with short-illness-duration first episode major depressive disorder:Possible role of cortisol immunomodulatory action at early stage of the disease. J Affect Disord. 2014;152-154:534–7. doi: 10.1016/j.jad.2013.10.004. [DOI] [PubMed] [Google Scholar]
  • 25.Piwowarska J, Wrzosek M, Zaleska MR, Pokraœniewicz BR, Skalski M, Matsumoto H, et al. Serum cortisol concentration in patients with major depression after treatment with clomipramine. Pharmacol Rep. 2009;61:604–11. doi: 10.1016/s1734-1140(09)70112-4. [DOI] [PubMed] [Google Scholar]
  • 26.Rudzki L, Pawlak D, Pawlak K, Waszkiewicz N, Małus A, Konarzewska B, et al. Immune suppression of IgG response against dairy proteins in major depression. BMC Psychiatry. 2017;17:268. doi: 10.1186/s12888-017-1431-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Tang AL, Thomas SJ, Larkin T. Cortisol, oxytocin, and quality of life in major depressive disorder. Qual Life Res. 2019;28:2919–28. doi: 10.1007/s11136-019-02236-3. [DOI] [PubMed] [Google Scholar]
  • 28.Jiang HK, Wang JY, Lin JC. The central mechanism of hypothalamic pituitary adrenocortical system hyperfunction in depressed patients. Psychiatry Clin Neurosci. 2000;54:227–34. doi: 10.1046/j.1440-1819.2000.00663.x. [DOI] [PubMed] [Google Scholar]
  • 29.Osran H, Reist C, Chen CC, Lifrak ET, Chicz-DeMet A, Parker LN. Adrenal androgens and cortisol in major depression. Am J Psychiatry. 1993;150:806–9. doi: 10.1176/ajp.150.5.806. [DOI] [PubMed] [Google Scholar]
  • 30.Bertollo AG, Grolli RE, Plissari ME, Gasparin VA, Quevedo J, Réus GZ, et al. Stress and serum cortisol levels in major depressive disorder:A cross-sectional study. AIMS Neurosci. 2020;7:459–69. doi: 10.3934/Neuroscience.2020028. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Wells GA, Shea B, O’Connell D, Peterson J, Welch V, Losos M, et al. The Newcastle-Ottawa scale (NOS) for assessing the quality if nonrandomized studies in meta-analyses. [Last accessed on 2022 Jul 07]. Available from: http://www.ohri.ca/programs/clinical_epidemiology/oxford.asp .
  • 32.JASP Team, JASP (Version 0.14.1), 2020;[Computer software] [Google Scholar]
  • 33.Hedges LV, Olkin I. San Diego, CA: Academic Press; 1985. Statistical Methods for Meta-Analysis. [Google Scholar]
  • 34.Viechtbauer W. Conducting meta-analyses in R with the meta for package. J Stat Softw. 2010;36:1–48. [Google Scholar]
  • 35.Egger M, Davey Smith G, Schneider M, Minder C. Bias in meta-analysis detected by a simple, graphical test. BMJ. 1997;315:629–34. doi: 10.1136/bmj.315.7109.629. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Duval S, Tweedie R. Trim and fill: A simple funnel-plot-based method of testing and adjusting for publication bias in meta-analysis. Biometrics. 2000;56:455–63. doi: 10.1111/j.0006-341x.2000.00455.x. [DOI] [PubMed] [Google Scholar]
  • 37.Wallace, Byron C, Dahabreh IJ, Trikalinos TA, Lau J, Trow P, et al. Closing the gap between methodologists and end-users:R as a computational back-end. J Stat Softw. 2012;49:5. [Google Scholar]
  • 38.Wolkowitz OM, Reus VI. Treatment of depression with antiglucocorticoid drugs. Psychosom Med. 1999;61:698–711. doi: 10.1097/00006842-199909000-00011. [DOI] [PubMed] [Google Scholar]
  • 39.Volkert J, Gablonski TC, Rabung S. Prevalence of personality disorders in the general adult population in Western countries:Systematic review and meta-analysis. Br J Psychiatry. 2018;213:709–15. doi: 10.1192/bjp.2018.202. [DOI] [PubMed] [Google Scholar]
  • 40.Vythilingam M, Gill JM, Luckenbaugh DA, Gold PW, Collin C, Bonne O. Low early morning plasma cortisol in posttraumatic stress disorder is associated with co-morbid depression but not with enhanced glucocorticoid feedback inhibition. Psychoneuroendocrinology. 2010;35:442–50. doi: 10.1016/j.psyneuen.2009.08.006. [DOI] [PubMed] [Google Scholar]
  • 41.Stokes PE, Sikes CR. Hypothalamic-pituitary-adrenal axis in affective disorders. In: Meltzer HY, editor. Psychopharmacology: The Third Generation of Progress. New York: Raven Press; 1987. [Google Scholar]

Articles from Indian Journal of Psychiatry are provided here courtesy of Wolters Kluwer -- Medknow Publications

RESOURCES