Skip to main content
NCBI home page
Biological Psychiatry Global Open Science logoLink to Biological Psychiatry Global Open Science
. 2024 Feb 18;4(3):100294. doi: 10.1016/j.bpsgos.2024.100294

Assessing the Role of Cortisol in Anxiety, Major Depression, and Neuroticism: A Mendelian Randomization Study Using SERPINA6/SERPINA1 Variants

Io Ieong Chan a,c,∗, Anise MS Wu b,c
PMCID: PMC10959652  PMID: 38525495

Abstract

Background

Previous evidence informed by the toxic stress model suggests that higher cortisol causes anxiety and major depression, but clinical success is lacking. To clarify the role of cortisol, we used Mendelian randomization to estimate its associations with anxiety, major depression, and neuroticism, leveraging the largest available genome-wide association studies including from the Psychiatric Genomics Consortium, the UK Biobank, and FinnGen.

Methods

After meta-analyzing 2 genome-wide association studies on morning plasma cortisol (n = 32,981), we selected single nucleotide polymorphisms (SNPs) at p < 5 × 10−8 and r2 < 0.3 in the SERPINA6/SERPINA1 gene region encoding proteins that influence cortisol bioavailability. We applied these SNPs to summary genetic associations with the outcomes considered (n = 17,310–449,484), and systolic blood pressure as a positive outcome, using inverse-variance weighted meta-analysis accounting for correlation. Sensitivity analyses addressing SNP correlation and confounding by childhood maltreatment and follow-up analyses using only SNPs that colocalized with SERPINA6 expression were conducted.

Results

Cortisol was associated with anxiety (pooled odds ratio [OR] 1.16 per cortisol z score; 95% CI, 1.04 to 1.31), but not major depression (pooled OR 1.02, 95% CI, 0.95 to 1.10) or neuroticism (β −0.025; 95% CI, −0.071 to 0.022). Sensitivity analyses yielded similar estimates. Cortisol was positively associated with systolic blood pressure, as expected. Using rs9989237 and rs2736898, selected using colocalization, cortisol was associated with anxiety in the UK Biobank (OR 1.32; 95% CI, 1.01 to 1.74) but not with major depression in FinnGen (OR 1.14; 95% CI, 0.95 to 1.37).

Conclusions

Cortisol was associated with anxiety and may be a potential target for prevention. Other targets may be more relevant to major depression and neuroticism.

Keywords: Anxiety, Cortisol, Depression, Mendelian randomization, Neuroticism

Plain Language Summary

Previous observational studies and randomized controlled trials remain inconclusive about whether cortisol causes mental disorders. In this Mendelian randomization study, using genetic variants from the SERPINA6/SERPINA1 gene region as instrumental variables for cortisol to minimize bias from confounding, we found that morning plasma cortisol was positively associated with anxiety but not major depression or neuroticism. From a practical perspective, these findings provide supporting evidence that lowering cortisol may reduce the risk of anxiety but not depression.

Plain Language Summary

Previous observational studies and randomized controlled trials remain inconclusive about whether cortisol causes mental disorders. In this Mendelian randomization study, using genetic variants from the SERPINA6/SERPINA1 gene region as instrumental variables for cortisol to minimize bias from confounding, we found that morning plasma cortisol was positively associated with anxiety but not major depression or neuroticism. From a practical perspective, these findings provide supporting evidence that lowering cortisol may reduce the risk of anxiety but not depression.

Globally, anxiety and major depression are among the most prevalent mental disorders (1), and neuroticism is a major risk factor for both (2). Neuroticism is typically considered to be a core personality trait (3) and has been described as a tendency toward emotional instability in reaction to stressful situations (4). Anxiety and major depression are highly disabling and have risen substantially in incidence and prevalence in recent years as a likely consequence of increasing environmental stress (5). While reducing stress can effectively prevent mental disorders (6), stress reduction on a population level remains highly challenging, if not impossible (7). As such, insights into the neurobiology of mental disorders that are seemingly affected by high levels of stress may inform molecular targets of prevention. Cortisol is a steroid hormone synthesized from cholesterol in the adrenal gland and is widely known as the “stress hormone” because its level is closely and positively correlated with perceived stress (8). Cortisol, regulated by the hypothalamus-pituitary-adrenal (HPA) axis (9), exhibits a circadian rhythm and typically peaks in the morning.

Chronic activation of the HPA axis resulting in persistently high levels of circulating cortisol is a hallmark of toxic stress response (10) and has frequently been associated with anxiety and major depression, but observational evidence remains inconclusive (11). Childhood maltreatment, which increases the risk of mental disorders (12) and has been associated with both cortisol blunting (13) and hyperactivity (14), may confound the observed effects of cortisol toward or away from the null, respectively. Previous randomized controlled trials (RCTs), which are free of confounding by design, did not find CRF1 (corticotropin-releasing factor receptor 1) antagonists, an HPA-targeting medication, to have an effect on either anxiety or depression (15,16). However, it was unclear whether the tested compounds reduced cortisol sufficiently (17). Conversely, RCTs of other HPA-targeting medications, such as V1b (vasopressin 1B) receptor antagonists, and mifepristone, a glucocorticoid receptor antagonist, have shown favorable effects on major depression (18). Given the mixed findings from observational studies and RCTs (19), the use of a quasi-experimental approach, such as Mendelian randomization (MR) (20), may help clarify the role of cortisol.

Using genetic variants as instrumental variables (IVs) for cortisol (21) to mimic the deconfounding property of RCTs, previous MR studies did not clearly show an association between cortisol and major depression or neuroticism (22,23). However, in one of the studies, the analysis did not account for the correlation between IVs (22), making the estimates difficult to interpret. Furthermore, a recent systematic review showed that very few MR studies on anxiety have used standardized criteria for diagnosis (24). Therefore, the role of cortisol in anxiety remains to be elucidated.

In this MR study, to clarify the role of cortisol, and by implication the HPA axis in anxiety and major depression, and to overcome some of the challenges in previous observational and MR studies, we used an extended set of genetic variants from the SERPINA6/SERPINA1 gene region that encode proteins that influence cortisol bioavailability as IVs for cortisol. We pooled the largest available genome-wide association studies (GWASs) to assess the associations of morning plasma cortisol with anxiety and major depression. We also included neuroticism because of its strong link with anxiety and major depression. We used systolic blood pressure (SBP) as the positive control outcome because exogenous cortisol increases SBP (25).

Methods and Materials

Study Design

This is a 2-sample MR study based on the largest available GWASs on anxiety, major depression, and neuroticism (Figure 1). MR relies on the 3 IV assumptions (26) (Figure 2). First, the IV is associated with the exposure of interest. Second, the IV is independent of confounders of exposure and outcome. Third, the IV is associated with the outcome only through affecting the exposure. In this MR study, we used genetic variants from the SERPINA6/SERPINA1 gene region as IVs. The SERPINA6 and SERPINA1 genes encode corticosteroid-binding globulin and α₁-antitrypsin, respectively. Both proteins have well-established biological functions that influence cortisol bioavailability (27). To strengthen the validity of the results, we conducted sensitivity analyses addressing correlation of the IVs and potential confounding by childhood maltreatment. Additionally, we performed follow-up analyses using only IVs that showed evidence of colocalization with SERPINA6 expression. This study is reported in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology guidelines for MR (Table S1).

Figure 1.

Figure 1

Open in a new tab

Study design of the current Mendelian randomization study on morning plasma cortisol and anxiety, major depression, and neuroticism. ANGST, Anxiety NeuroGenetics Study; CORNET, Cortisol Network; GPC, Genetics of Personality Consortium; GWAS, genome-wide association study; ICBP, International Consortium for Blood Pressure; LIFE, Leipzig Research Centre for Civilization Diseases; PGC, Psychiatric Genomics Consortium; SESA, sensitivity to environmental stress; SNP, single nucleotide polymorphism.

Figure 2.

Figure 2

Open in a new tab

Directed acyclic graph illustrating the instrumental variable assumptions in Mendelian randomization and potential biasing pathways (in dashed lines). LD, linkage disequilibrium.

Data Source for Cortisol

To maximize power to detect associations with the outcomes considered, we obtained and combined summary genetic associations with morning plasma cortisol from the CORNET (Cortisol Network) consortium (28) and the Leipzig Research Centre for Civilization Diseases LIFE-Adult (95% of analyzed samples obtained between 7:00 and 10:30 am) and LIFE-Heart studies (95% of analyzed samples obtained between 6:00 and 11:45 am) (29), using fixed-effects inverse-variance weighted (IVW) meta-analysis (30), which resulted in a total of up to 32,981 participants. We only included studies that reported the use of morning samples to minimize measurement error given the diurnal and ultradian nature of cortisol secretion. The CORNET consortium, with a pooled mean age (SD) of 53.3 (11.2) years and 36.3% men, was a GWAS meta-analysis of 25,314 participants of European ancestry from 17 European population-based cohorts adjusted for age, sex, and genetic principal components (PCs) for population structure. The LIFE-Adult and LIFE-Heart studies, with a pooled mean age (SD) of 59.9 (12.0) years and 55.3% men, included 7667 participants of European ancestry from Leipzig, Germany. The GWAS adjusted for age, sex, and body mass index (BMI). The combined genetic associations with cortisol z scores (per pooled SD or ∼179 nmol/L) were checked for potential effect size inflation from confounding. We used the linkage disequilibrium (LD) score regression intercept, where a value substantially different from unity indicates the presence of confounding by population stratification (31).

Genetic Instrument for Cortisol

Because similar to previous GWASs, only genetic variants in the SERPINA6/SERPINA1 gene region (chromosome 14: 94,768,859–94,843,565) (28,32) were associated with morning plasma cortisol at genome-wide significance (p < 5 × 10−8), consistent with previous MR studies, we selected single nucleotide polymorphisms (SNPs) associated with cortisol in low (r2 < 0.3) LD (28,33). The SERPINA6/SERPINA1 SNPs have been shown to affect cortisol levels throughout the day (34), which makes them suitable IVs for long-term cortisol exposure.

Data Sources for Anxiety, Major Depression, and Neuroticism

We obtained summary genetic associations with anxiety, major depression, and neuroticism from a GWAS of participants of European ancestry mainly from the PGC (Psychiatric Genomics Consortium) (35), an international consortium of genetic studies of psychiatric disorders; the UK Biobank (36), an ongoing cohort study of ∼500,000 adults from Great Britain recruited at ages 40 to 69 years; and FinnGen (37), an expanding cohort study intended to recruit 500,000 participants from Finland. Here, we used FinnGen’s ninth public data release (data freeze 9), which consists up to 377,277 participants. Details of the included studies, including source population, sample size, overlap of participants, case and control definitions, and covariate adjustment, are detailed in Table S2.

Anxiety

We obtained summary genetic associations with anxiety disorders from 4 major data sources. The ANGST (Anxiety NeuroGenetics Study) Consortium included 3695 cases and 13,615 controls (38). The UK Biobank included 25,453 cases and 58,113 controls (39). Both the ANGST Consortium and UK Biobank defined cases based on self-report lifetime diagnosis and DSM-based diagnostic assessment. The iPSYCH Cohort, part of the The Lundbeck Foundation Initiative for Integrative Psychiatric Research, included 4584 cases and 19,225 controls (40). FinnGen included 40,191 cases and 277,526 controls (37). The iPSYCH Cohort and FinnGen defined cases based on linked medical records. In the ANGST Consortium, the UK Biobank, and FinnGen, the GWASs all adjusted for age, sex, and some genetic PCs, but in the iPSYCH Cohort, the GWAS only adjusted for genetic PCs.

Major Depression

We obtained summary genetic associations with major depression from 3 major data sources including the PGC (43,204 cases and 95,680 controls) (41), the UK Biobank (113,769 cases and 208,811 controls) (42), and FinnGen (43,280 cases and 329,192 controls) (37). All 3 studies defined cases based on medical records, and in the PGC and the UK Biobank, cases were also defined using DSM- or ICD-based diagnostic assessment and history of help seeking for mental health difficulties, respectively. All 3 GWASs adjusted for genetic PCs, and in the UK Biobank and FinnGen, other covariates such as age and sex were also included. In the UK Biobank, for statistical power, we selected the broadest case definition available for major depression due to its high genetic correlation with probable depression (rg = 0.871, p = 4.18 × 10−67) and ICD-coded major depressive disorder (rg = 0.863, p = 5.03 × 10−80). Furthermore, the broad depression phenotype was not correlated with the anxiety factor score (rg = 0.052 ± 0.11) (42), which makes bias via association with anxiety unlikely.

Neuroticism Overall and Subclusters

We obtained summary genetic associations with the neuroticism mean total score (in z scores) from a GWAS meta-analysis of the GPC (Genetics of Personality Consortium) and the UK Biobank that involved a total of 449,484 participants (43). Both GWASs adjusted for age and sex, and in the UK Biobank, genetic PCs, genotyping array, and Townsend deprivation index were also adjusted for. Given the heterogeneity between the neuroticism items, we also obtained summary genetic associations with the total scores (in z scores) of 3 neuroticism subclusters from a GWAS of the UK Biobank, including depressed affect (357,957 participants) (44), worry (348,219 participants) (44), and sensitivity to environmental stress (351,827 participants) (45), which adjusted for the same covariates as the GWAS of the UK Biobank neuroticism mean total score.

Data Source for SBP

We obtained summary genetic associations with SBP (n = 757,601) in mm Hg from a GWAS meta-analysis of the UK Biobank (36) and the International Consortium for Blood Pressure (46), which adjusted for age, age2, sex, and BMI.

Statistical Analysis

Instrument strength was calculated using the F statistic approximated by the squared SNP-cortisol association divided by the variance of the association (47). An F statistic <10 indicates weak instruments, which bias the MR estimates toward the observational direction, where there is overlap of participants for exposure and outcome studies, and toward the null, where there is no overlap (48). Power calculations were conducted based on the proportion of phenotypic variance explained by the IVs (49), using the more conservative estimate of 0.5% taken from previous findings for the SERPINA6/SERPINA1 locus (32) given the weak LD of the selected SNPs. SNPs for cortisol and outcomes were aligned on the same effect alleles. SNPs available for cortisol but not the outcomes and palindromic SNPs (A/T and G/C) with ambiguous effect allele frequency (0.42–0.58), which are difficult to align, were replaced by LD proxies (r2 = 1.0) where available (50).

To estimate the associations of cortisol with the outcomes considered, we used the IVW method with fixed effects to combine the SNP-specific Wald estimates, allowing for the correlation between SNPs in the weights. Where applicable, to increase precision, we meta-analyzed with fixed effects the estimates for the same outcome to give an overall estimate. Given the fact that we used SNPs from the same gene region, standard sensitivity analyses for MR, such as weighted median (51), which assumes that the majority of SNPs are valid IVs, and MR-Egger (52), which requires invalid IVs to not be associated with the outcome of interest via the same pleiotropic paths, are not suitable in this setting. To check whether the estimates were sensitive to the SNP correlation threshold (r2 < 0.3) (53), we repeated the analysis using r2 < 0.1 and 0.5 as thresholds. We also applied the IVW-PC method to all genome-wide significant SNPs in the SERPINA6/SERPINA1 gene region (54) to further address the correlation between SNPs. The IVW-PC method uses PC analysis to identify uncorrelated linear combinations of SNPs that explain the largest proportion of variance in cortisol. Compared to selecting SNPs based on an r2 cutoff, the MR estimates obtained from the IVW-PC method is less sensitive to differences in the SNP correlation matrix. To check whether the estimates could be confounded by LD with childhood adversity affecting steroid hormone exposure (55), we applied two-step cis-MR (TSCMR) to adjust for childhood adversity (56), using summary genetic associations with childhood adversity indicated by maltreatment from a GWAS meta-analysis of 185,414 individuals (57) that measured emotional, sexual, and physical abuse and emotional and physical neglect. TSCMR extends the 2-step MR approach used for mediation analysis (58). Specifically, TSCMR similarly uses the product and difference of coefficients methods, leveraging the SNP-confounder and confounder-outcome associations to adjust for potential bias via the confounder in the SNP-outcome associations. TSCMR differs from conventional MR methods used for estimating independent effects, such as multivariable MR, which typically uses uncorrelated IVs (59).

Where significant associations were found, we first applied leave-one-out analyses to identify possible outlier SNPs with large effects. Then, to further rule out potential bias from LD with unmeasured confounders that violated the IV assumptions, we repeated the primary analysis using SNPs with evidence of colocalization with SERPINA6 expression in the liver (60), given that SNPs in this specific region influence cortisol levels by affecting SERPINA6 expression (28). We applied approximate Bayes factor colocalization analysis (61) to summary genetic associations with morning plasma cortisol and gene expression of SERPINA6 in the GTEx (Genotype-Tissue Expression) project version 8 (n = 208) (62). We selected the SNP with the strongest posterior probability of colocalization within 100 kilobases of the SERPINA6/SERPINA1 gene region. We also included rs2736898, which was identified in a previous colocalization analysis using the Stockholm-Tartu Atherosclerosis Reverse Networks Engineering Task study (28).

We used a Bonferroni-corrected threshold of .017 (.05/3) for statistical significance given the 3 main outcomes considered, namely anxiety, major depression, and neuroticism. We used METAL software (http://genome.sph.umich.edu/wiki/METAL) to meta-analyze the GWAS estimates for cortisol, the R packages TwoSampleMR and MendelianRandomization to obtain MR estimates, TwoStepCisMR to obtain adjusted SNP-outcome associations, coloc to obtain colocalization estimates, LDlinkR to obtain LD proxies, and metafor to meta-analyze MR estimates. All statistical analyses were conducted using R (version 4.2.2, The R Foundation for Statistical Computing).

Ethics

This study only used publicly available data, so specific ethical approval was not required. Ethical approval for the studies included can be found in the original publications.

Results

From the combined CORNET and LIFE studies, there were 8 SNPs (Table S3) at p < 5 × 10−8 and r2 < 0.3 with a mean F statistic of 58.1 (32.2–114.4). All the SNPs or their proxies (rs11620763 replaced by rs7141205) were available. Power calculations showed that at a Bonferroni-corrected alpha level (5%/3), this study had 80% power to detect per morning plasma cortisol z score, a positive/negative association with anxiety, with an odds ratio (OR) of 1.19/0.81 and 1.12/0.88 for major depression in the studies combined and a β of ±0.07 to 0.08 for neuroticism total and by subcluster.

As shown in Figure 3, cortisol was positively associated with anxiety in the UK Biobank (OR 1.27; 95% CI, 1.03 to 1.57) and when meta-analyzed with the other 3 studies (OR 1.16; 95% CI, 1.04 to 1.31), although no significant associations were observed in the ANGST Consortium (OR 1.18; 95% CI, 0.70 to 1.98), iPSYCH Cohort (OR 0.89; 95% CI, 0.55 to 1.42), or FinnGen (OR 1.14; 95% CI, 0.98 to 1.33). Cortisol was positively associated with major depression (OR 1.20; 95% CI, 1.04 to 1.38) in FinnGen, but not in other studies or when meta-analyzed with the other 2 studies (OR 1.02; 95% CI, 0.95 to 1.10). Cortisol was not associated with neuroticism overall (β −0.025 z score; 95% CI, −0.071 to 0.022) or any of its subclusters. Repeating the analysis using different r2 values to select SNPs and the IVW-PC method (Table S4) gave similar estimates. Adjusting for childhood maltreatment (Tables S5, S6) also gave similar estimates. Cortisol was positively associated with SBP (Table S7) in the primary (β 0.660 mm Hg; 95% CI, 0.051 to 1.270) as well as the sensitivity analyses.

Figure 3.

Figure 3

Open in a new tab

Mendelian randomization estimates of the association of morning plasma cortisol with anxiety, major depression, and neuroticism using genetic variants from the SERPINA6/SERPINA1 gene region as instruments. p Values for heterogeneity for individual (between single nucleotide polymorphisms) and combined (between studies) Mendelian randomization estimates were obtained using Cochrane’s Q test. ANGST, Anxiety NeuroGenetics Study; OR, odds ratio; PGC, Psychiatric Genomics Consortium; Phet, p values for heterogeneity; SESA, sensitivity to environmental stress.

Leave-one-out analyses (Table S8) did not show any specific SNPs driving the associations of cortisol with anxiety in the UK Biobank or major depression in FinnGen. The regional association plots (Figure 4) showed well-defined peaks within the same region for morning plasma cortisol and SERPINA6 expression. Colocalization analysis in the GTEx project suggested moderate evidence (59.2%) of the 2 traits sharing 1 causal SNP, with rs9989237 being the lead SNP (Table S9). Using only rs9989237 and rs2736898, at reduced power, cortisol was nominally associated with anxiety (OR 1.32; 95% CI, 1.01 to 1.74) in the UK Biobank but not major depression (OR 1.14; 95% CI, 0.95 to 1.37) in FinnGen (Figure 5).

Figure 4.

Figure 4

Open in a new tab

Regional association plots of genetic associations with morning plasma cortisol and liver SERPINA6 expression highlighting rs9989237 (blue) and rs2736898 (red).

Figure 5.

Figure 5

Open in a new tab

Mendelian randomization estimates of the association of cortisol with anxiety in the UK Biobank (p for heterogeneity = .921) and major depression in FinnGen (p for heterogeneity = .804) using rs9989237 and rs2736898. OR, odds ratio.

Discussion

Consistent with the toxic stress model (10), which suggests that elevated cortisol contributes to the development of mental disorders, we found that morning plasma cortisol was positively associated with the risk of anxiety. Our findings also clarify that, contrary to one previous MR study (22) but consistent with another MR study (23), RCTs of CRF1 antagonists (15) and previous observational studies on personality (63), cortisol was not associated with major depression or neuroticism. Therefore, our findings suggest that elevated cortisol, and by implication overactivation of the HPA axis, primarily contributes to anxiety but not major depression and does not play a role in neuroticism.

Compared to previous MR studies that have examined the role of cortisol in mental disorders (22,23), our study has several strengths. First, we accounted for the correlation between SNPs, thereby minimizing the possibility of false positives. Second, we pooled the largest available studies to minimize the possibility of false negatives. Additionally, in addition to major depression, we included anxiety in our study and clarified the association of cortisol with anxiety. Notably, cortisol was associated with anxiety overall, but not in ANGST or iPSYCH, possibly due to the relatively smaller sample size. However, in iPSYCH, the lack of adjustment for age or sex may have resulted in confounding from population stratification (64). Our findings are also broadly consistent with previous prospective cohort studies of adolescents, which showed a positive association of cortisol with symptoms of anxiety (65) but not major depression (66). However, our findings are not entirely consistent with previous RCTs of CRF1 antagonists, which did not find an association of CRF1 inhibition with anxiety (16), or with RCTs of mifepristone and V1b receptor antagonists, which showed favorable effects on major depression. Depletion of CRF1 signaling in mice led to reduced symptoms related to anxiety and depression (67), and elevated cortisol was observed in patients with comorbid major depression and anxiety, but not in patients with major depression alone (68). Mifepristone may act via antiandrogen pathways to reduce depression (69). Animal studies suggest that it is likely that V1b receptor depletion does not reduce aggression via lowering androgen (70). It has been suggested that the effect of cortisol on mental disorders could be mediated by inflammation (71). However, given the anti-inflammatory properties of α₁-antitrypsin (72) and the negative associations of morning plasma cortisol with proinflammatory cytokines (73), the mediating role of inflammation may not fully explain the positive associations of cortisol with anxiety. Alternatively, cortisol suppresses androgens (74), which may in turn increase depressive symptoms (75). Additional research is warranted to better understand the mediating mechanisms, as well as the possibility of sex-specific effects (76).

Despite the MR design, which minimizes bias from environmental confounding, some limitations of the current study require attention. First, the MR estimates would be invalid if the IV assumptions were violated, and in the current study, we implemented several measures to ensure that this was not the case. We exclusively used SNPs from the SERPINA6/SERPINA1 gene region, which has well-established biological relevance to cortisol bioavailability (27). Furthermore, the selected SNPs were associated with morning plasma cortisol at genome-wide significance, and we used SBP as a positive control outcome for validation. In addition to the extensive sensitivity analyses addressing correlation between SNPs and potential confounding by childhood maltreatment, in the follow-up analysis, we leveraged colocalization to select SNPs for further validation, so it is unlikely that the positive findings with anxiety were driven by LD with unmeasured confounding. There was little sample overlap between exposure and outcome overall, except for ANGST at 39%. However, the F statistics were well above the rule-of-thumb value of 10 for all SNPs, with a mean of 58.1, indicating that bias was at most 0.67% for ANGST and was unlikely for outcomes where there was no sample overlap (48). In the LIFE studies, the GWAS adjusted for BMI, potentially introducing collider bias via the common causes of cortisol and BMI (77). However, using only SNPs from the SERPINA6/SERPIN1 gene region minimized this possibility. Second, we did not consider subtypes of anxiety and depression despite the known heterogeneity (78), primarily to avoid false negatives given the much smaller sample sizes of the subtypes (20). Furthermore, the overarching aim of this study was to examine the role of cortisol in the most prevalent mental disorders, namely anxiety and major depression, meaning that the use of broadly defined outcomes is necessary. Similarly, given the overarching aim of this study, we did not conduct a reverse MR, which addresses the question of whether differences in liability to mental disorders influence morning plasma cortisol. For the same reason, a colocalization analysis of morning plasma cortisol and mental disorders was not conducted because this addresses the question of whether a specific SNP may causally influence both cortisol and mental disorders irrespective of the direction of causality (61). Third, we assessed the lifetime but not the age-specific risk of anxiety and major depression, and as such our findings do not clarify whether a critical window of risk (20), for example adolescence (79), exists for exposure to higher cortisol. However, the use of lifetime diagnoses encompasses cases diagnosed at younger ages, which avoids selection bias (80). Future studies of younger participants are warranted to clarify the possibility of a critical window. Fourth, this study used a 2-sample MR design, meaning that only linear relationships could be assessed. Although we cannot preclude the possibility of a threshold effect for cortisol, very low or very high levels of cortisol are indicators of poor health (81), which can be a source of confounding in conventional observational studies but is less likely in MR studies. Nevertheless, the assessment of nonlinear effects is possible but will require that the exposure be measured in the outcome population (82), which is not often available. Fifth, using TSCMR, we ruled out the possibility of confounding by LD with childhood maltreatment, but whether the MR estimates could be modified by childhood maltreatment, which remains a possibility (83), warrants further elucidation and will have profound implications for the role and importance of childhood and adolescent well-being. Sixth, some cases of anxiety and major depression were based on self-report, so the possibility of bias due to misclassification exists (84) but is likely small given that most cases were ascertained based on DSM/ICD criteria. Seventh, cortisol secretion generally follows a diurnal and ultradian pattern but is highly variable between individuals (85). Random variation in morning plasma cortisol could bias the genetic estimates toward null. Replication using other cortisol measurements, such as hair cortisol, which is more stable over time (86), is warranted. Eighth, this MR study is based on GWASs of participants of European ancestry, mostly adjusted for but not stratified by sex. Replication by sex in other populations, which have different presentations of anxiety and depressive symptoms (87), is warranted.

From a public health and clinical perspective, stress is increasingly recognized as a serious public health issue that influences both physical and mental health (88). The current findings provide additional evidence that cortisol plays a causal role in mental health, particularly in anxiety. Our findings also support ongoing efforts aimed at targeting the cortisol pathway to alleviate anxiety and redirect focus toward other pathways for major depression (89).

Conclusions

Morning plasma cortisol, instrumented by SNPs from the SERPINA6/SERPINA1 gene region, was positively associated with anxiety but not major depression or neuroticism. Consistent with the toxic stress model, our findings suggest that cortisol can be targeted to reduce the risk of anxiety, while other pathways may be more relevant to major depression and neuroticism.

Acknowledgments and Disclosures

IIC and AMSW conceptualized the study and interpreted the results. IIC conducted the formal analyses and wrote the first draft, with critical feedback and revisions from AMSW.

We thank the study participants and researchers for making the genome-wide association studies summary data used in this study available.

Data sources for cortisol, anxiety, major depression, neuroticism, and systolic blood pressure can be found in Table S2. The analysis code and combined summary statistics for cortisol are available from the corresponding author upon request.

The authors report no biomedical financial interests or potential conflicts of interest.

Footnotes

Supplementary material cited in this article is available online at https://doi.org/10.1016/j.bpsgos.2024.100294.

Supplementary Material

Key Resources Table
mmc1.xlsx (20.8KB, xlsx)
Supplement
mmc2.xlsx (44.4KB, xlsx)

References

  • 1.GBD 2019 Mental Disorders Collaborators Global, regional, and national burden of 12 mental disorders in 204 countries and territories, 1990–2019: A systematic analysis for the Global Burden of Disease Study 2019. Lancet Psychiatry. 2022;9:137–150. doi: 10.1016/S2215-0366(21)00395-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Lahey B.B. Public health significance of neuroticism. Am Psychol. 2009;64:241–256. doi: 10.1037/a0015309. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Eysenck H.J. Biological basis of personality. Nature. 1963;199:1031–1034. doi: 10.1038/1991031a0. [DOI] [PubMed] [Google Scholar]
  • 4.Bolger N., Zuckerman A. A framework for studying personality in the stress process. J Pers Soc Psychol. 1995;69:890–902. doi: 10.1037//0022-3514.69.5.890. [DOI] [PubMed] [Google Scholar]
  • 5.COVID-19 Mental Disorders Collaborators Global prevalence and burden of depressive and anxiety disorders in 204 countries and territories in 2020 due to the COVID-19 pandemic. Lancet. 2021;398:1700–1712. doi: 10.1016/S0140-6736(21)02143-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Wilson N., McDaid S. The mental health effects of a Universal Basic Income: A synthesis of the evidence from previous pilots. Soc Sci Med. 2021;287 doi: 10.1016/j.socscimed.2021.114374. [DOI] [PubMed] [Google Scholar]
  • 7.Alvaredo F., Chancel L., Piketty T., Saez E., Zucman G. The elephant curve of global inequality and growth. AEA Papers and Proceedings. 2018;108:103–108. [Google Scholar]
  • 8.Stalder T., Steudte-Schmiedgen S., Alexander N., Klucken T., Vater A., Wichmann S., et al. Stress-related and basic determinants of hair cortisol in humans: A meta-analysis. Psychoneuroendocrinology. 2017;77:261–274. doi: 10.1016/j.psyneuen.2016.12.017. [DOI] [PubMed] [Google Scholar]
  • 9.Herman J.P., McKlveen J.M., Ghosal S., Kopp B., Wulsin A., Makinson R., et al. Regulation of the hypothalamic-pituitary-adrenocortical stress response. Compr Physiol. 2016;6:603–621. doi: 10.1002/cphy.c150015. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Shonkoff J.P., Garner A.S., Committee on Psychosocial Aspects of Child and Family Health, Committee on Early Childhood, Adoption, and Dependent Care, Section on Developmental and Behavioral Pediatrics The lifelong effects of early childhood adversity and toxic stress. Pediatrics. 2012;129:e232–e246. doi: 10.1542/peds.2011-2663. [DOI] [PubMed] [Google Scholar]
  • 11.Chida Y., Steptoe A. Cortisol awakening response and psychosocial factors: A systematic review and meta-analysis. Biol Psychol. 2009;80:265–278. doi: 10.1016/j.biopsycho.2008.10.004. [DOI] [PubMed] [Google Scholar]
  • 12.Baldwin J.R., Wang B., Karwatowska L., Schoeler T., Tsaligopoulou A., Munafò M.R., Pingault J.B. Childhood maltreatment and mental health problems: A systematic review and meta-analysis of quasi-experimental studies. Am J Psychiatry. 2023;180:117–126. doi: 10.1176/appi.ajp.20220174. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Bunea I.M., Szentágotai-Tătar A., Miu A.C. Early-life adversity and cortisol response to social stress: A meta-analysis. Transl Psychiatry. 2017;7:1274. doi: 10.1038/s41398-017-0032-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Jopling E., LeMoult J. Childhood adversity and the cortisol awakening response in depression: A meta-analysis. J Mood Anxiety Disord. 2023;2 [Google Scholar]
  • 15.Binneman B., Feltner D., Kolluri S., Shi Y., Qiu R., Stiger T. A 6-week randomized, placebo-controlled trial of CP-316,311 (a selective CRH1 antagonist) in the treatment of major depression. Am J Psychiatry. 2008;165:617–620. doi: 10.1176/appi.ajp.2008.07071199. [DOI] [PubMed] [Google Scholar]
  • 16.Coric V., Feldman H.H., Oren D.A., Shekhar A., Pultz J., Dockens R.C., et al. Multicenter, randomized, double-blind, active comparator and placebo-controlled trial of a corticotropin-releasing factor receptor-1 antagonist in generalized anxiety disorder. Depress Anxiety. 2010;27:417–425. doi: 10.1002/da.20695. [DOI] [PubMed] [Google Scholar]
  • 17.Spierling S.R., Zorrilla E.P. Don’t stress about CRF: Assessing the translational failures of CRF1 antagonists. Psychopharmacol (Berl) 2017;234:1467–1481. doi: 10.1007/s00213-017-4556-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Ding Y., Wei Z., Yan H., Guo W. Efficacy of treatments targeting hypothalamic-pituitary-adrenal systems for major depressive disorder: A meta-analysis. Front Pharmacol. 2021;12 doi: 10.3389/fphar.2021.732157. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Menke A. Is the HPA axis as target for depression outdated, or is there a New Hope? Front Psychiatry. 2019;10:101. doi: 10.3389/fpsyt.2019.00101. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Wootton R.E., Jones H.J., Sallis H.M. Mendelian randomisation for psychiatry: How does it work, and what can it tell us? Mol Psychiatry. 2022;27:53–57. doi: 10.1038/s41380-021-01173-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Lawlor D.A., Harbord R.M., Sterne J.A.C., Timpson N., Davey Smith G. Mendelian randomization: Using genes as instruments for making causal inferences in epidemiology. Stat Med. 2008;27:1133–1163. doi: 10.1002/sim.3034. [DOI] [PubMed] [Google Scholar]
  • 22.Zhou X., Qiao N. Association of cortisol levels with neuropsychiatric functions: A Mendelian randomization analysis. Front Endocrinol. 2019;10:564. doi: 10.3389/fendo.2019.00564. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Lee W.H., Larsson S.C., Wood A., Di Angelantonio E., Butterworth A.S., Burgess S., Allara E. Genetically predicted plasma cortisol and common chronic diseases: A Mendelian randomization study. Clin Endocrinol (Oxf) 2024;100:238–244. doi: 10.1111/cen.14966. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Saccaro L.F., Gasparini S., Rutigliano G. Applications of Mendelian randomization in psychiatry: A comprehensive systematic review. Psychiatr Genet. 2022;32:199–213. doi: 10.1097/YPG.0000000000000327. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Werumeus Buning J., van Faassen M., Brummelman P., Dullaart R.P., van den Berg G., van der Klauw M.M., et al. Effects of hydrocortisone on the regulation of blood pressure: Results from a randomized controlled trial. J Clin Endocrinol Metab. 2016;101:3691–3699. doi: 10.1210/jc.2016-2216. [DOI] [PubMed] [Google Scholar]
  • 26.Burgess S., Davey Smith G., Davies N.M., Dudbridge F., Gill D., Glymour M.M., et al. Guidelines for performing Mendelian randomization investigations: Update for summer 2023. Wellcome Open Res. 2019;4:186. doi: 10.12688/wellcomeopenres.15555.1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Henley D.E., Lightman S.L. New insights into corticosteroid-binding globulin and glucocorticoid delivery. Neuroscience. 2011;180:1–8. doi: 10.1016/j.neuroscience.2011.02.053. [DOI] [PubMed] [Google Scholar]
  • 28.Crawford A.A., Bankier S., Altmaier E., Barnes C.L.K., Clark D.W., Ermel R., et al. Variation in the SERPINA6/SERPINA1 locus alters morning plasma cortisol, hepatic corticosteroid binding globulin expression, gene expression in peripheral tissues, and risk of cardiovascular disease. J Hum Genet. 2021;66:625–636. doi: 10.1038/s10038-020-00895-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Pott J., Bae Y.J., Horn K., Teren A., Kühnapfel A., Kirsten H., et al. Genetic association study of eight steroid hormones and implications for sexual dimorphism of coronary artery disease. J Clin Endocrinol Metab. 2019;104:5008–5023. doi: 10.1210/jc.2019-00757. [DOI] [PubMed] [Google Scholar]
  • 30.Willer C.J., Li Y., Abecasis G.R. METAL: Fast and efficient meta-analysis of genomewide association scans. Bioinformatics. 2010;26:2190–2191. doi: 10.1093/bioinformatics/btq340. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Bulik-Sullivan B.K., Loh P.R., Finucane H.K., Ripke S., Yang J., Schizophrenia Working Group of the Psychiatric Genomics Consortium, et al. LD Score regression distinguishes confounding from polygenicity in genome-wide association studies. Nat Genet. 2015;47:291–295. doi: 10.1038/ng.3211. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Bolton J.L., Hayward C., Direk N., Lewis J.G., Hammond G.L., Hill L.A., et al. Genome wide association identifies common variants at the SERPINA6/SERPINA1 locus influencing plasma cortisol and corticosteroid binding globulin. PLOS Genet. 2014;10 doi: 10.1371/journal.pgen.1004474. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Larsson S.C., Lee W.H., Kar S., Burgess S., Allara E. Assessing the role of cortisol in cancer: A wide-ranged Mendelian randomisation study. Br J Cancer. 2021;125:1025–1029. doi: 10.1038/s41416-021-01505-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Utge S., Räikkönen K., Kajantie E., Lipsanen J., Andersson S., Strandberg T., et al. Polygenic risk score of SERPINA6/SERPINA1 associates with diurnal and stress-induced HPA axis activity in children. Psychoneuroendocrinology. 2018;93:1–7. doi: 10.1016/j.psyneuen.2018.04.009. [DOI] [PubMed] [Google Scholar]
  • 35.Sullivan P.F., Agrawal A., Bulik C.M., Andreassen O.A., Børglum A.D., Breen G., Cichon S., et al. Psychiatric genomics: An update and an agenda. Am J Psychiatry. 2018;175:15–27. doi: 10.1176/appi.ajp.2017.17030283. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Sudlow C., Gallacher J., Allen N., Beral V., Burton P., Danesh J., et al. UK Biobank: An open access resource for identifying the causes of a wide range of complex diseases of middle and old age. PLoS Med. 2015;12 doi: 10.1371/journal.pmed.1001779. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Kurki M.I., Karjalainen J., Palta P., Sipilä T.P., Kristiansson K., Donner K., et al. FinnGen: Unique genetic insights from combining isolated population and national health register data. medRxiv. 2022 doi: 10.1101/2022.03.03.22271360. [DOI] [Google Scholar]
  • 38.Otowa T., Hek K., Lee M., Byrne E.M., Mirza S.S., Nivard M.G., et al. Meta-analysis of genome-wide association studies of anxiety disorders. Mol Psychiatry. 2016;21:1391–1399. doi: 10.1038/mp.2015.197. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Purves K.L., Coleman J.R.I., Meier S.M., Rayner C., Davis K.A.S., Cheesman R., et al. A major role for common genetic variation in anxiety disorders. Mol Psychiatry. 2020;25:3292–3303. doi: 10.1038/s41380-019-0559-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Meier S.M., Trontti K., Purves K.L., Als T.D., Grove J., Laine M., et al. Genetic variants associated with anxiety and stress-related disorders: A genome-wide association study and mouse-model study. JAMA Psychiatry. 2019;76:924–932. doi: 10.1001/jamapsychiatry.2019.1119. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Wray N.R., Ripke S., Mattheisen M., Trzaskowski M., Byrne E.M., Abdellaoui A., et al. Genome-wide association analyses identify 44 risk variants and refine the genetic architecture of major depression. Nat Genet. 2018;50:668–681. doi: 10.1038/s41588-018-0090-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Howard D.M., Adams M.J., Clarke T.K., Hafferty J.D., Gibson J., Shirali M., et al. Genome-wide meta-analysis of depression identifies 102 independent variants and highlights the importance of the prefrontal brain regions. Nat Neurosci. 2019;22:343–352. doi: 10.1038/s41593-018-0326-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Nagel M., Jansen P.R., Stringer S., Watanabe K., de Leeuw C.A., Bryois J., et al. Meta-analysis of genome-wide association studies for neuroticism in 449,484 individuals identifies novel genetic loci and pathways. Nat Genet. 2018;50:920–927. doi: 10.1038/s41588-018-0151-7. [DOI] [PubMed] [Google Scholar]
  • 44.Nagel M., Watanabe K., Stringer S., Posthuma D., van der Sluis S. Item-level analyses reveal genetic heterogeneity in neuroticism. Nat Commun. 2018;9:905. doi: 10.1038/s41467-018-03242-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Nagel M., Speed D., van der Sluis S., Østergaard S.D. Genome-wide association study of the sensitivity to environmental stress and adversity neuroticism cluster. Acta Psychiatr Scand. 2020;141:476–478. doi: 10.1111/acps.13155. [DOI] [PubMed] [Google Scholar]
  • 46.International Consortium for Blood Pressure Genome-Wide Association Studies. Ehret G.B., Munroe P.B., Rice K.M., Bochud M., Johnson A.D., et al. Genetic variants in novel pathways influence blood pressure and cardiovascular disease risk. Nature. 2011;478:103–109. doi: 10.1038/nature10405. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Bowden J., Del Greco M.F., Minelli C., Davey Smith G., Sheehan N.A., Thompson J.R. Assessing the suitability of summary data for two-sample Mendelian randomization analyses using MR-Egger regression: The role of the I2 statistic. Int J Epidemiol. 2016;45:1961–1974. doi: 10.1093/ije/dyw220. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Burgess S., Davies N.M., Thompson S.G. Bias due to participant overlap in two-sample Mendelian randomization. Genet Epidemiol. 2016;40:597–608. doi: 10.1002/gepi.21998. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49.Freeman G., Cowling B.J., Schooling C.M. Power and sample size calculations for Mendelian randomization studies using one genetic instrument. Int J Epidemiol. 2013;42:1157–1163. doi: 10.1093/ije/dyt110. [DOI] [PubMed] [Google Scholar]
  • 50.Machiela M.J., Chanock S.J. LDlink: A web-based application for exploring population-specific haplotype structure and linking correlated alleles of possible functional variants. Bioinformatics. 2015;31:3555–3557. doi: 10.1093/bioinformatics/btv402. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51.Bowden J., Davey Smith G., Haycock P.C., Burgess S. Consistent estimation in Mendelian randomization with some invalid instruments using a weighted median estimator. Genet Epidemiol. 2016;40:304–314. doi: 10.1002/gepi.21965. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 52.Bowden J., Davey Smith G., Burgess S. Mendelian randomization with invalid instruments: Effect estimation and bias detection through Egger regression. Int J Epidemiol. 2015;44:512–525. doi: 10.1093/ije/dyv080. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 53.Gkatzionis A., Burgess S., Newcombe P.J. Statistical methods for cis-Mendelian randomization with two-sample summary-level data. Genet Epidemiol. 2023;47:3–25. doi: 10.1002/gepi.22506. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 54.Burgess S., Zuber V., Valdes-Marquez E., Sun B.B., Hopewell J.C. Mendelian randomization with fine-mapped genetic data: Choosing from large numbers of correlated instrumental variables. Genet Epidemiol. 2017;41:714–725. doi: 10.1002/gepi.22077. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 55.Eck S.R., Ardekani C.S., Salvatore M., Luz S., Kim E.D., Rogers C.M., et al. The effects of early life adversity on growth, maturation, and steroid hormones in male and female rats. Eur J Neurosci. 2020;52:2664–2680. doi: 10.1111/ejn.14609. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 56.Woolf B., Zagkos L., Gill D. TwoStepCisMR: A novel method and R package for attenuating bias in cis-Mendelian randomization analyses. Genes (Basel) 2022;13:1541. doi: 10.3390/genes13091541. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 57.Warrier V., Kwong A.S.F., Luo M., Dalvie S., Croft J., Sallis H.M., et al. Gene-environment correlations and causal effects of childhood maltreatment on physical and mental health: A genetically informed approach. Lancet Psychiatry. 2021;8:373–386. doi: 10.1016/S2215-0366(20)30569-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 58.Relton C.L., Davey Smith G. Two-step epigenetic Mendelian randomization: A strategy for establishing the causal role of epigenetic processes in pathways to disease. Int J Epidemiol. 2012;41:161–176. doi: 10.1093/ije/dyr233. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 59.Burgess S., Thompson D.J., Rees J.M.B., Day F.R., Perry J.R., Ong K.K. Dissecting causal pathways using Mendelian randomization with summarized genetic data: Application to age at menarche and risk of breast cancer. Genetics. 2017;207:481–487. doi: 10.1534/genetics.117.300191. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 60.Zuber V., Grinberg N.F., Gill D., Manipur I., Slob E.A.W., Patel A., et al. Combining evidence from Mendelian randomization and colocalization: Review and comparison of approaches. Am J Hum Genet. 2022;109:767–782. doi: 10.1016/j.ajhg.2022.04.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 61.Giambartolomei C., Vukcevic D., Schadt E.E., Franke L., Hingorani A.D., Wallace C., Plagnol V. Bayesian test for colocalisation between pairs of genetic association studies using summary statistics. PLoS Genet. 2014;10 doi: 10.1371/journal.pgen.1004383. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 62.GTEx Consortium The Genotype-Tissue Expression (GTEx) project. Nat Genet. 2013;45:580–585. doi: 10.1038/ng.2653. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 63.Luo J., Zhang B., Cao M., Roberts B.W. The stressful personality: A meta-analytical review of the relation between personality and stress. Pers Soc Psychol Rev. 2023;27:128–194. doi: 10.1177/10888683221104002. [DOI] [PubMed] [Google Scholar]
  • 64.Uffelmann E., Huang Q.Q., Munung N.S., de Vries J., Okada Y., Martin A.R., et al. Genome-wide association studies. Nat Rev Methods Primers. 2021;1:59. [Google Scholar]
  • 65.Adam E.K., Vrshek-Schallhorn S., Kendall A.D., Mineka S., Zinbarg R.E., Craske M.G. Prospective associations between the cortisol awakening response and first onsets of anxiety disorders over a six-year follow-up—2013 Curt Richter Award Winner. Psychoneuroendocrinology. 2014;44:47–59. doi: 10.1016/j.psyneuen.2014.02.014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 66.Carnegie R., Araya R., Ben-Shlomo Y., Glover V., O’Connor T.G., O’Donnell K.J., et al. Cortisol awakening response and subsequent depression: Prospective longitudinal study. Br J Psychiatry. 2014;204:137–143. doi: 10.1192/bjp.bp.113.126250. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 67.Reul J.M., Holsboer F. On the role of corticotropin-releasing hormone receptors in anxiety and depression. Dialogues Clin Neurosci. 2002;4:31–46. doi: 10.31887/DCNS.2002.4.1/jreul. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 68.Young E.A., Abelson J.L., Cameron O.G. Effect of comorbid anxiety disorders on the hypothalamic-pituitary-adrenal axis response to a social stressor in major depression. Biol Psychiatry. 2004;56:113–120. doi: 10.1016/j.biopsych.2004.03.017. [DOI] [PubMed] [Google Scholar]
  • 69.Song L.N., Coghlan M., Gelmann E.P. Antiandrogen effects of mifepristone on coactivator and corepressor interactions with the androgen receptor. Mol Endocrinol. 2004;18:70–85. doi: 10.1210/me.2003-0189. [DOI] [PubMed] [Google Scholar]
  • 70.Wersinger S.R., Ginns E.I., O’Carroll A.M., Lolait S.J., Young W.S., 3rd Vasopressin V1b receptor knockout reduces aggressive behavior in male mice. Mol Psychiatry. 2002;7:975–984. doi: 10.1038/sj.mp.4001195. [DOI] [PubMed] [Google Scholar]
  • 71.Saccaro L.F., Schilliger Z., Dayer A., Perroud N., Piguet C. Inflammation, anxiety, and stress in bipolar disorder and borderline personality disorder: A narrative review. Neurosci Biobehav Rev. 2021;127:184–192. doi: 10.1016/j.neubiorev.2021.04.017. [DOI] [PubMed] [Google Scholar]
  • 72.Bergin D.A., Hurley K., McElvaney N.G., Reeves E.P. Alpha-1 antitrypsin: A potent anti-inflammatory and potential novel therapeutic agent. Arch Immunol Ther Exp (Warsz) 2012;60:81–97. doi: 10.1007/s00005-012-0162-5. [DOI] [PubMed] [Google Scholar]
  • 73.Rajasundaram S., Rahman R.P., Woolf B., Zhao S.S., Gill D. Morning cortisol and circulating inflammatory cytokine levels: A Mendelian randomisation study. Genes. 2022;13:116. doi: 10.3390/genes13010116. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 74.Cumming D.C., Quigley M.E., Yen S.S. Acute suppression of circulating testosterone levels by cortisol in men. J Clin Endocrinol Metab. 1983;57:671–673. doi: 10.1210/jcem-57-3-671. [DOI] [PubMed] [Google Scholar]
  • 75.Walther A., Breidenstein J., Miller R. Association of testosterone treatment with alleviation of depressive symptoms in men: A systematic review and meta-analysis. JAMA Psychiatry. 2019;76:31–40. doi: 10.1001/jamapsychiatry.2018.2734. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 76.Raven P.W., Taylor N.F. Sex differences in the human metabolism of cortisol. Endocr Res. 1996;22:751–755. doi: 10.1080/07435809609043772. [DOI] [PubMed] [Google Scholar]
  • 77.Hartwig F.P., Tilling K., Davey Smith G., Lawlor D.A., Borges M.C. Bias in two-sample Mendelian randomization when using heritable covariable-adjusted summary associations. Int J Epidemiol. 2021;50:1639–1650. doi: 10.1093/ije/dyaa266. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 78.Feczko E., Miranda-Dominguez O., Marr M., Graham A.M., Nigg J.T., Fair D.A. The heterogeneity problem: Approaches to identify psychiatric subtypes. Trends Cogn Sci. 2019;23:584–601. doi: 10.1016/j.tics.2019.03.009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 79.Niwa M., Lee R.S., Tanaka T., Okada K., Kano S.-I., Sawa A. A critical period of vulnerability to adolescent stress: Epigenetic mediators in mesocortical dopaminergic neurons. Hum Mol Genet. 2016;25:1370–1381. doi: 10.1093/hmg/ddw019. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 80.Sackett D.L. Bias in analytic research. J Chronic Dis. 1979;32:51–63. doi: 10.1016/0021-9681(79)90012-2. [DOI] [PubMed] [Google Scholar]
  • 81.Schoorlemmer R.M.M., Peeters G.M.E.E., Van Schoor N.M., Lips P. Relationships between cortisol level, mortality and chronic diseases in older persons. Clin Endocrinol (Oxf) 2009;71:779–786. doi: 10.1111/j.1365-2265.2009.03552.x. [DOI] [PubMed] [Google Scholar]
  • 82.Tian H., Mason A.M., Liu C., Burgess S. Relaxing parametric assumptions for non-linear Mendelian randomization using a doubly-ranked stratification method. PLOS Genet. 2023;19 doi: 10.1371/journal.pgen.1010823. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 83.Gerritsen L., Milaneschi Y., Vinkers C.H., van Hemert A.M., van Velzen L., Schmaal L., Penninx B.W. HPA axis genes, and their interaction with childhood maltreatment, are related to cortisol levels and stress-related phenotypes. Neuropsychopharmacology. 2017;42:2446–2455. doi: 10.1038/npp.2017.118. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 84.Xue A., Jiang L., Zhu Z., Wray N.R., Visscher P.M., Zeng J., Yang J. Genome-wide analyses of behavioural traits are subject to bias by misreports and longitudinal changes. Nat Commun. 2021;12 doi: 10.1038/s41467-020-20237-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 85.Upton T.J., Zavala E., Methlie P., Kämpe O., Tsagarakis S., Øksnes M., et al. High-resolution daily profiles of tissue adrenal steroids by portable automated collection. Sci Transl Med. 2023;15 doi: 10.1126/scitranslmed.adg8464. [DOI] [PubMed] [Google Scholar]
  • 86.Short S.J., Stalder T., Marceau K., Entringer S., Moog N.K., Shirtcliff E.A., et al. Correspondence between hair cortisol concentrations and 30-day integrated daily salivary and weekly urinary cortisol measures. Psychoneuroendocrinology. 2016;71:12–18. doi: 10.1016/j.psyneuen.2016.05.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 87.Rudenstine S., Espinosa A. Latent comorbid depression and anxiety symptoms across sex and race/ethnic subgroupings in a national epidemiologic study. J Psychiatr Res. 2018;104:114–123. doi: 10.1016/j.jpsychires.2018.07.005. [DOI] [PubMed] [Google Scholar]
  • 88.O’Connor D.B., Thayer J.F., Vedhara K. Stress and health: A review of psychobiological processes. Annu Rev Psychol. 2021;72:663–688. doi: 10.1146/annurev-psych-062520-122331. [DOI] [PubMed] [Google Scholar]
  • 89.Fries G.R., Saldana V.A., Finnstein J., Rein T. Molecular pathways of major depressive disorder converge on the synapse. Mol Psychiatry. 2023;28:284–297. doi: 10.1038/s41380-022-01806-1. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Key Resources Table
mmc1.xlsx (20.8KB, xlsx)
Supplement
mmc2.xlsx (44.4KB, xlsx)

Articles from Biological Psychiatry Global Open Science are provided here courtesy of Elsevier

RESOURCES