The Association Between Thyrotropin and Clinically Relevant Depression: A Retrospective Cross-Sectional Study

Abstract

Introduction:

Thyroid dysfunction and mood disorders are chronic health conditions with a significant impact on quality of life. This study aimed to investigate the association between thyrotropin (TSH) and clinically relevant depression (CRD) in patients with and without mood disorders, in a population-based sample.

Methods:

This retrospective cross-sectional study included all consecutive adults (≥18 years) who had TSH and completed the Patient Health Questionnaire (PHQ-9) within 6 months of the index visit, between October 2016 and May 2021, at the University of Utah Health. Data on demographics, hypothyroidism diagnoses, TSH, thyroid hormone replacement (THR), PHQ-9, antidepressant (AD) medications, and mood disorder diagnoses (using the International Classification of Diseases, 10th Revision, Clinical Modification codes; Major depressive disorder single episode—F32, recurrent—F33, persistent mood disorder—F34, bipolar disorder—F30+F31, and mood disorder not otherwise specified—F39) were extracted electronically. CRD was defined as PHQ-9 ≥ 10. t-Test and chi-square test were used to compare continuous and categorical variables, respectively. Logistic regression models were formulated to evaluate the association between TSH and CRD, after adjusting for covariates.

Results:

The cohort included 33,138 patients, mean age 42.41 ± 16.10 years, 80.67% Caucasian, 69.10% females, and mean PHQ-9 score 10.11 ± 6.94. A total of 45.23% (n = 14,989) patients had a diagnosis of mood disorders, and 49.70% had CRD. Patients with mood disorders were more likely to be female, Caucasian, non-Hispanic/Latino, on AD, had hypothyroidism diagnoses, on thyroid medications, had higher mean PHQ-9 scores, and had CRD. TSH level was associated with an increased odds of CRD (odds ratio [OR] = 1.01, confidence interval [CI], 1.01–1.02, p < 0.009) after adjusting for age, sex, body mass index, Charlson Comorbidity Index, and use of THR and AD. Both the low TSH and high TSH groups showed increased odds of CRD, with respective ORs of 1.19 (CI: 1.04–1.37) and 1.26 (CI: 1.13–1.40).

Conclusions:

Thyroid dysfunction is associated with an increase in the odds of depression. Future longitudinal cohort studies are recommended to investigate the association between thyroid function and incident depression.

Introduction

Thyroid dysfunction and mood disorders are chronic health conditions with a significant impact on quality of life. According to the United States National Health and Nutrition Examination Survey III, the prevalence of hypothyroidism in nonpregnant adults is 4.6% (4.3% subclinical hypothyroidism [SCH] and 0.3% overt hypothyroidism [OH]).¹ The estimated 12-month and lifetime prevalence of mood disorders based on the National Comorbidity Survey Replication in the U.S. adult population is 9.7% (11.6% in females and 7.7% in males) and 21.4% (24.9% in females and 17.5% in males), respectively.² Hypothyroidism is 5–8 times more common in women^1,3,4 than men, and mood disorders follow a similar pattern and are 1.5 times more common in women.^5
–7 Thyroid dysfunction is associated with various neuropsychiatric conditions, such as depression, mania, psychosis, and cognitive disorders,^8

–11 and optimal thyroid hormone levels are essential for normal brain function and an adequate response to antidepressants (ADs). The hypothalamic-pituitary-thyroid (HPT) axis plays an important role in the pathophysiology¹² and clinical course of mood disorders.¹³ Abnormalities of HPT axis have been described in patients who have depression, even with normal thyrotropin (TSH) levels. Blunting of TSH response to TSH-releasing hormone is a widely recognized abnormality of the HPT axis. Variations in the circadian rhythm of TSH, including the loss of normal nocturnal TSH surge, have also been described in the literature.

Variable evidence exists regarding TSH in depression depending on their thyroid status. Untreated OH has shown to be associated with an increased risk of depression,¹⁴ with improvement in depressive symptoms after optimal thyroid hormone replacement (THR). The association of SCH and depression is conflicting, with no published recommendations from the Endocrine and Psychiatric Societies for/against THR in nonpregnant adults,^15
–17 except for the 2013 European Thyroid Association’s guideline. This guideline reported that younger patients with SCH may experience mild mood impairment that improves with THR in some cases, but this benefit is not observed in patients over 65 years.¹⁸ In euthyroid patients with depression, particularly those resistant to AD treatment, improved depressive symptoms have been reported when targeting a TSH level of <2.5 mU/L or even <2 mU/L.¹⁹

Several studies have explored the link between thyroid disorders and mood disorders with varying results. Recent renewed interest in this field has led to new meta-analyses and systematic reviews. A 2018 meta-analysis reported that SCH was associated with depression only in adults under 60 years of age.²⁰ Conversely, a 2019 meta-analysis reported that SCH was positively associated with an increased risk of depression, particularly in adults over 50.²¹ These studies underscore the complexity of thyroid dysfunction’s impact on depression. Despite several studies, this association remains an enigma to date.

The goal of this study was to investigate the association between TSH levels and clinically relevant depression (CRD), defined as a Patient Health Questionnaire (PHQ-9) score ≥10, in adults with and without mood disorders.^11,22,23 Furthermore, the study also investigated the association between free thyroxine (fT4) levels and CRD, aiming to strengthen the current evidence and provide a scientific basis for further investigation.

Methods

Overview and study design

This is a retrospective cross-sectional study that included all consecutive adults (≥18 years) who had a TSH and completed the PHQ-9 within 6 months of the index visit, between October 2016 and May 2021, at the University of Utah Health, Utah, USA. An inclusion index date was determined based on the most recent date of the first pair of TSH and PHQ-9 scores that met the inclusion criteria for each patient. Patients with incomplete PHQ-9 scores, PHQ-2 scores alone, and missing TSH data were excluded. In addition, patients with TSH levels over 100 or a body mass index (BMI) over 70 kg/m² were excluded from the cohort as they were potential outliers. The study protocol was approved by the University of Utah Institutional Review Board (IRB #00118610). The authors have adhered to the Principles of Ethical Publishing.²⁴

Setting and participants

The electronic health records (EHRs) of the University of Utah Health were reviewed to identify the study cohort. Patient data were extracted from the University of Utah Health Enterprise Data Warehouse, which contains a longitudinal record of all data since 1993 from inpatient and outpatient EHRs, as well as over 250 ancillary clinical, financial, and laboratory information systems. A standard operating procedure was followed to ensure data quality. As the University of Utah Health clinics are a tertiary referral center, study participants were primarily from Utah and neighboring states.

International Classification of Diseases, 10th Revision, Clinical Modification (ICD-10-CM) codes were used to identify mood disorders (Major depressive disorder [MDD], single episode—F32; recurrent—F33; persistent mood disorder—F34; bipolar disorder—F30, F31; and mood disorder not otherwise specified [NOS]—F39) and hypothyroidism of any etiology (E00.0, E00.1, E00.2, E00.9, E01.0, E01.1, E01.2, E01.8, E02, E03.0, E03.1, E03.2, E03.3, E03.4, E03.8, E03.9, E06.0, E06.1, E06.2, E06.3, E06.4, E06.5, E06.9, E07.0, E07.1, E07.81, E07.89, E07.9, and E89.0).

Data on demographics (age, sex, race/ethnicity, and BMI), Charlson Comorbidity Index (CCI), mood disorder diagnoses using ICD-10-CM codes, PHQ-9 scores, hypothyroidism diagnoses, thyroid laboratory tests (TSH, fT4, total T4 [T4], free T3 [fT3], total T3 [T3], thyroid peroxidase antibody [TPO Ab], thyroid receptor antibodies, thyroid stimulating immunoglobulin), THR (levothyroxine [LT4], liothyronine [T3], or combination), antithyroid drugs (methimazole, propylthiouracil), and medications for mood disorders (ADs, mood-stabilizers, antipsychotics, stimulants, benzodiazepines, Z-drugs/hypnotics, beta-blockers, other anxiolytics, and dopamine agonists) were extracted electronically. Thyroid biochemistry assay utilized was Roche or Abbott for TSH and Roche for fT4. The data were de-identified, including the use of random patient and visit identifiers and obfuscation of all dates present in the dataset by a random number of days between 1 and 181 that was generated separately for each patient.

Outcome measures

Depression symptoms were measured based on the PHQ-9 scores.²³ A PHQ-9 ≥10, which has shown 88% sensitivity and 88% specificity for identifying major depression, was chosen as the primary outcome to deﬁne CRD in this cohort.^22,23 PHQ9-item #4 measures “feeling tired or having little energy” and PHQ9-item #7 measures “trouble concentrating on things, such as reading the newspaper or watching television.”

TSH and fT4 levels were used to measure thyroid function. The normal reference range for TSH was 0.35–4.94 mIU/L, with levels categorized as follows: normal (0.35–4.94 mIU/L), low (<0.35 mIU/L), and high (>4.94 mIU/L). fT4 levels were categorized as normal (0.9–1.7 ng/dL), low (<0.9 ng/dL), and high (>1.7 ng/dL). These reference ranges were used to define SCH, subclinical hyperthyroidism, OH, and overt hyperthyroidism. SCH was defined as TSH >4.94 mIU/L with fT4 within 0.9–1.7 ng/dL, while OH was TSH >4.94 mIU/L with fT4 <0.9 ng/dL. Subclinical hyperthyroidism was TSH <0.35 mIU/L with fT4 within 0.9–1.7 ng/dL, whereas overt hyperthyroidism was TSH <0.35 mIU/L with fT4 >1.7 ng/dL.

Outcome

The primary outcome was CRD. We also assessed the relationship between TSH and PHQ-9, Q#4 (fatigue) and PHQ-9, Q#7 (impaired concentration). A secondary analysis was conducted to assess the relationship between low and high TSH and fT4 groups and CRD.

Statistical analysis

Continuous and categorical data were presented using mean ± standard deviation and frequency (percentage), respectively. Comparison of continuous and categorical variables was carried out using t-tests and chi-square tests, respectively. Logistic regression models were constructed to assess the relationship between TSH and CRD. Unadjusted logistic regression models were first formulated, and then models were adjusted for confounding variables. Odds ratios (ORs) along with corresponding confidence intervals (CIs) were reported. Covariates treated as confounders included age, sex, BMI, CCI, hypothyroidism, THR, and AD use. Stratification was performed to examine subgroup differences based on sex (male vs. female), age groups (<65 years vs. ≥65 years), mood disorders, hypothyroidism, THR, and AD use. In multivariable subgroup models, the respective subgroup variable of interest was excluded as a covariate. Spearman’s correlation was used to assess the relationship between TSH and scores on questions 4 (Q4) and 7 (Q7) of PHQ-9. Recent Mendelian randomization (MR) studies on depression and thyroid diseases have reported inconsistent relationships between depression and hypothyroidism among individuals of European and East Asian ancestry. We conducted a post hoc sensitivity analysis to investigate the association between TSH and CRD among patients of European ancestry.²⁵

Free T4 levels were available for 24.32% of patients (n = 8060). Spearman’s correlation was used to assess the relationship between fT4 and PHQ-9. The associations between TSH categories (low, normal, and high) and fT4 categories (low, normal, and high) with CRD were examined in both unadjusted models and after adjusting for confounders.

Statistical significance was set at p < 0.05 for all analyses. No p value adjustment for multiple comparison testing was done based on prior recommendations.²⁶ The management and analysis of data were conducted using SAS Studio 3.8 (SAS Institute Inc, Cary, North Carolina).

Results

The final study cohort (Fig. 1) consisted of 33,138 adults with a mean age of 42.41 ± 16.10 years, 80.67% of whom were Caucasian, and 69.10% were female. The mean PHQ-9 score was 10.11 ± 6.94 (Table 1). Among the cohort, 45.23% (n = 14,989) had a diagnosis of mood disorders (MDD: 86.05%, bipolar disorder: 10.91%, persistent mood disorder: 1.89%, mood disorder not otherwise specified: 1.15%). AD use was observed in 50.02% of the patients. Hypothyroidism was diagnosed in 10.21%, and 13.39% of the entire cohort was on THR. Normal TSH levels were most prevalent at 91.79%, while high and low TSH levels were observed in 5.26% and 2.95%, respectively. CRD was observed in 49.70% of the cohort. The mean TSH was slightly higher in the CRD group compared with the non-CRD group, at 2.44 ± 4.42 versus 2.39 ± 3.76, respectively, although this difference was not statistically significant (p = 0.26).

FIG. 1.

Flow diagram of the study cohort. PHQ-9, Patient Health Questionnaire; TSH, thyrotropin; BMI, body mass index; PMD, persistent mood disorder; NOS, not otherwise specified.

Table 1.

The Difference in Characteristics Between Patients With and Without Mood Disorders

Cohort characteristics	Overall cohort (N = 33,138)	Mood disorder (n = 14,989)	No mood disorder (N = 18,149)	p value
Age	42.41 (16.10)	42.32 (16.22)	42.49 (16.01)	0.34
Older adults (age ≥ 65yrs)	3512 (10.60%)	1659 (11.07%)	1853 (10.21%)	0.01
Sex (Female)	22,900 (69.10%)	10,816 (72.16%)	12,084 (66.58%)	< 0.001
Race				< 0.001
Caucasian	26,731 (80.67%)	12,784 (85.29%)	13,947 (76.85%)
African American	705 (2.13%)	273 (1.82%)	432 (2.38%)
American Indian and Alaska Native	237 (0.72%)	106 (0.71%)	131 (0.72%)
Asian	841 (2.54%)	228 (1.52%)	613 (3.38%)
Native Hawaiian and Other Pacific Islander	340 (1.03%)	91 (0.61%)	249 (1.37%)
Other	3795 (11.45%)	1346 (8.98%)	2449 (13.49%)
Unknown	489 (1.48%)	161 (1.07%)	328 (1.81%)
Ethnicity				< 0.001
Hispanic/Latino	4582 (13.83%)	1677 (11.19%)	2905 (16.01%)
Not Hispanic/Latino	27,827 (83.97%)	13,053 (87.08%)	14,774 (81.40%)
Unknown	729 (2.20%)	259 (1.73%)	470 (2.59%)
TSH	2.41 (4.10)	2.35 (3.75)	2.46 (4.37)	0.01
PHQ-9	10.11 (6.94)	12.28 (6.76)	8.31 (6.55)	< 0.001
GAD-7 (n = 23,677)	9.25 (6.46)	10.67 (6.27)	7.87 (6.35)	< 0.001
BMI (n = 32,734)	29.85 (7.88)	30.54 (8.21)	29.28 (7.54)	< 0.001
CCI (n = 33,136)	1.12 (2.03)	1.38 (2.27)	0.90 (1.78)	< 0.001
Hypothyroidism diagnoses	3384 (10.21%)	2177 (14.52%)	1207 (6.65%)	< 0.001
Free T4 (n = 8060)	1.26 (0.38)	1.25 (0.38)	1.27 (0.37)	0.007
Total T4 (n = 445)	7.61 (2.31)	7.48 (2.30)	7.77 (2.33)	0.18
Free T3 (n = 1897)	3.20 (1.57)	3.19 (1.75)	3.21 (1.30)	0.73
Total T3 (n = 1117)	120.01 (57.90)	116.77 (55.14)	123.69 (60.73)	0.05
TPO Ab (n = 1639)	115.78 (372.99)	99.15 (358.30)	133.92 (387.79)	0.06
TRAb (n = 343)	2.96 (6.36)	3.27 (7.30)	2.70 (5.45)	0.42
TSI (n = 200)	144.88 (111.01)	140.88 (111.09)	150.41 (111.33)	0.55
Thyroid hormone replacement	4438 (13.39%)	2560 (17.08%)	1878 (10.35%)	< 0.001
Antithyroid drugs	86 (0.26%)	48 (0.32%)	38 (0.21%)	0.05
TSH categories				0.0007
Normal (0.35 ≤TSH ≤4.94)	30,416 (91.79%)	13,705 (91.43%)	16,711 (92.08)
Low (TSH <0.35)	978 (2.95%)	501 (3.34%)	477 (2.63%)
High (TSH >4.94)	1744 (5.26%)	783 (5.22%)	961 (5.30%)
Depression severity				< 0.001
PHQ-9: <10	16,669 (50.30%)	5526 (36.87%)	11,143 (61.40%)
PHQ-9: 10–14	7150 (21.58%)	3640 (24.28%)	3510 (19.34%)
PHQ-9: 15–19	5503 (16.61%)	3277 (21.86%)	2226 (12.27%)
PHQ-9: >19	3816 (11.52%)	2546 (16.99%)	1270 (7.00%)
CRD (PHQ-9 ≥10)	16,469 (49.70%)	9463 (63.13%)	7006 (38.60%)	< 0.001
Mood disorder diagnoses
MDD, single	7630 (23.02%)	7630 (50.90%)	0 (0.00%)
MDD, recurrent	5268 (15.90%)	5268 (35.15%)	0 (0.00%)
Bipolar disorder	1636 (4.94%)	1636 (10.91%)	0 (0.00%)
Persistent mood disorder	283 (0.85%)	283 (1.89%)	0 (0.00%)
Mood disorder NOS	172 (0.52%)	172 (1.15%)	0 (0.00%)
Antidepressants	16,576 (50.02%)	12,365 (82.49%)	4211 (23.20%)	< 0.001
SSRIs	11,385 (34.36%)	8681 (57.92%)	2704 (14.90%)	< 0.001
SNRIs	4283 (12.92%)	3517 (23.46%)	766 (4.22%)	< 0.001
TCAs	1836 (5.54%)	1354 (9.03%)	482 (2.66%)	< 0.001
MAOIs	20 (0.06%)	18 (0.12%)	2 (0.01%)	< 0.001
Others	6867 (20.72%)	5651 (37.70%)	1216 (6.70%)	< 0.001
Mood stabilizing anticonvulsants	8126 (24.52%)	6111 (40.77%)	2015 (11.10%)	< 0.001
Lithium	578 (1.74%)	521 (3.48%)	57 (0.31%)	< 0.001
First-generation antipsychotics	2823 (8.52%)	2081 (13.88%)	742 (4.09%)	< 0.001
Second-generation antipsychotics	3645 (11.00%)	3068 (20.47%)	577 (3.18%)	< 0.001
Stimulants	1984 (5.99%)	1445 (9.64%)	539 (2.97%)	< 0.001
Benzodiazepines	4747 (14.32%)	3593 (23.97%)	1154 (6.36%)	< 0.001
Z-drugs	1599 (4.83%)	1211 (8.08%)	388 (2.14%)	< 0.001
Beta-blockers	3051 (9.21%)	2064 (13.77%)	987 (5.44%)	< 0.001
Other anxiolytics	853 (2.57%)	671 (4.48%)	182 (1.00%)	< 0.001
Dopamine agonists	273 (0.82%)	212 (1.41%)	61 (0.34%)	< 0.001

Continuous data are reported as mean and standard deviation (SD). Categorical data are reported as number and percentage.

TSH, thyrotropin; PHQ-9, Patient Health Questionnaire; T4, thyroxine; CRD, clinically relevant depression; GAD-7, General Anxiety Disorder-7; MAOIs, monoamine oxidase inhibitors; MDD, major depressive disorder; SSRIs, selective serotonin reuptake inhibitors; SNRIs, serotonin and norepinephrine reuptake inhibitors; TCAs, tricyclic antidepressants; TPO Ab, thyroid peroxidase antibody; TRAb, thyroid receptor antibody; TSI, thyroid stimulating immunoglobulin; BMI, body mass index; CCI, Charlson Comorbidity Index; NOS, not otherwise specified.

Table 1 reports the differences in demographics and clinical characteristics between patients with and without mood disorders. Compared with patients without mood disorders, patients with mood disorders were more likely to be female (72.16% vs. 66.58%), Caucasian (85.29% vs. 76.85%), non-Hispanic/Latino (87.08% vs. 81.4%), on ADs (82.49% vs. 23.20%), had a diagnosis of hypothyroidism (14.52% vs. 6.65%), were on THR (17.08% vs. 10.35%), and had a higher mean PHQ-9 score (12.28 ± 6.76 vs. 8.31 ± 6.55). Patients with mood disorders had a significantly higher prevalence of CRD (63.13% vs. 38.60%, p < 0.001).

Association between TSH and CRD

Overall, increasing TSH level was associated with an increased odds of CRD (OR = 1.01, CI:, 1.01–1.02, p < 0.009) after adjusting for covariates (see Table 2). This association was consistent across the age groups. In females, higher TSH levels were linked to increased odds of CRD (OR = 1.02, CI: 1.01–1.02, p < 0.0001), whereas no significant association was found in males (OR = 1.00, CI: 0.98–1.01, p = 0.41). For patients with hypothyroidism, higher TSH levels increased CRD odds (OR = 1.02, CI: 1.01–1.03, p < 0.0001), unlike those without the diagnosis (OR = 1.01, CI: 1.00–1.01, p = 0.22). TSH levels were associated with increased CRD odds in patients on THR (OR = 1.02, CI: 1.01–1.03, p < 0.0001), but not in those not on THR (OR = 1.00, CI: 1.00–1.01, p = 0.88). This increase was observed in patients with and without mood disorders (OR = 1.01, CI: 1.00–1.02, p = 0.02; OR = 1.01, CI: 1.01–1.02, p = 0.0007, respectively). Finally, TSH levels were associated with increased CRD odds in patients not on ADs (OR = 1.01, CI: 1.01–1.02, p = 0.0002), but not in those prescribed ADs (OR = 1.01, CI: 1.00–1.02, p = 0.20). Sensitivity analysis restricted to patients of European ancestry (White population) yielded results consistent with the entire cohort in both unadjusted (OR = 1.00, CI:, 1.00–1.01, p = 0.229) and adjusted models (OR = 1.01, CI:, 1.01–1.02, p = 0.0004).

Table 2.

Relationship Between Thyrotropin and Clinically Relevant Depression

	Unadjusted			Adjusted^a
	Odds ratio	CI	p value	Odds ratio	CI	p value
Entire Cohort	1.00	1.00, 1.01	0.07	1.01	1.01, 1.02	<0.009
Sex
Female	1.01	1.00, 1.02	0.006	1.02	1.01, 1.02	<0.0001
Male	0.99	0.98, 1.00	0.02	1.00	0.98, 1.01	0.41
Age
	1.00	1.00, 1.01	0.1847	1.01	1.00, 1.02	0.002
≥65 years	1.02	1.00, 1.04	0.0145	1.02	1.00, 1.04	0.03
Hypothyroidism diagnosis^b
Yes	1.02	1.01, 1.02	0.0003	1.02	1.01, 1.03	<0.0001
No	1.00	0.99, 1.01	0.99	1.01	1.00, 1.01	0.22
Thyroid hormone replacement
Yes	1.02	1.01, 1.02	<0.0001	1.02	1.01, 1.03	<0.0001
No	0.99	0.98, 1.00	0.0188	1.00	0.99, 1.01	0.88
Mood disorder diagnosis
Yes	1.00	0.99, 1.01	0.4888	1.01	1.00, 1.02	0.02
No	1.01	1.00, 1.01	0.086	1.01	1.01, 1.02	0.0007
Use of antidepressants
Yes	1.00	0.99, 1.01	0.87	1.006	1.00, 1.02	0.20
No	1.01	1.00, 1.01	0.04	1.01	1.01, 1.02	0.0002

Adjusted for age, sex, BMI, Charlson score, hypothyroidism, thyroid hormone replacement, and antidepressant use.

Adjusted for age, sex, BMI, Charlson score, thyroid hormone replacement, and antidepressant use.

CI, confidence interval.

When dividing TSH into subgroups (low, normal, and high), both the low TSH and high TSH groups showed increased odds of CRD compared to the euthyroid group (after adjusting for the covariates—age, sex, ADs, hypothyroidism, THR, BMI, and CCI), with respective ORs of 1.19 (CI: 1.04–1.37) and 1.26 (CI: 1.13–1.40), Supplementary Table S1.

TSH and PHQ-9 items #4 and #7

We observed a weak negative correlation between TSH and both Q4 (fatigue) and Q7 (impaired concentration) of PHQ-9, with correlation coefficients of r = −0.013 (p = 0.014) for question 4 and r = −0.046 (p < 0.001) for question 7. TSH level was associated with an increase in the odds of fatigue (PHQ-9, item #4, OR = 1.01, CI:, 1.00–1.01, p = 0.0004), but not of impaired concentration (PHQ-9, item #7, OR = 1.00, CI:, 1.00–1.01, p = 0.59).

Thyroid disorder categories and CRD

In the subcohort (n = 8060) where both TSH and fT4 were available, 3986 (49.45%) had CRD. Patients with either low TSH and low fT4, high TSH and high fT4, or normal TSH and low/high fT4 were excluded for this analysis. In the adjusted models, the SCH group was associated with an increased odds of CRD in both the entire subcohort (OR = 1.02, CI:, 1.00–1.03, p = 0.02) and the mood disorder cohort (OR = 1.03, CI:, 1.00–1.05, p = 0.03). Other thyroid disorders were not significantly associated with CRD (all p > 0.05), Table 3.

Table 3.

Relationship Between Thyrotropin Disorder Categories and Clinically Relevant Depression^a

	Unadjusted			Adjusted^b
	Odds ratio	CI	p value	Odds ratio	CI	p value
Entire cohort
SC hypothyroidism	1.02	1.00, 1.03	0.03	1.02	1.00, 1.03	0.02
SC hyperthyroidism	0.86	0.18, 4.22	0.85	0.89	0.17, 4.75	0.89
Overt hypothyroidism	1.00	0.99, 1.01	0.62	1.00	0.99, 1.02	0.58
Overt hyperthyroidism^c	3.61	0.23, 56.35	0.36	5.99	0.30, 117.92	0.24
Mood disorder
SC hypothyroidism	1.02	1.00, 1.05	0.049	1.03	1.00, 1.05	0.03
SC hyperthyroidism	3.23	0.33, 31.19	0.31	2.71	0.26, 28.82	0.41
Overt hypothyroidism	1.01	0.99, 1.03	0.42	1.00	0.98, 1.03	0.89
Overt hyperthyroidism^a	10.54	0.21, 522.72	0.24	13.27	0.22, 820.43	0.22
No mood disorder
SC hypothyroidism	1.01	0.99, 1.03	0.39	1.01	0.99, 1.03	0.32
SC hyperthyroidism	0.26	0.02, 2.80	0.27	0.26	0.02, 3.16	0.29
Overt hypothyroidism	1.00	0.99, 1.02	0.67	1.01	0.99, 1.02	0.51
Overt hyperthyroidism^a	0.50	0.01, 31.87	0.74	2.39	0.02, 249.63	0.71

Although 8060 patients had fT4 levels, we only included patients with the patterns of high TSH/low fT4 or low TSH/high fT4 (n = 7526) for this analysis.

Adjusted for age, sex, BMI, Charlson score, thyroid hormone replacement, mood disorders, and antidepressant use.

Confidence intervals are very wide because sample size is small.

SC, subclinical.

Free T4 and CRD

Overall, the fT4 level was negatively associated with CRD (OR = 0.83, CI: = 0.73–0.93). However, after adjusting for covariates, this association was no longer significant (OR = 0.90, CI: = 0.80–1.01, p = 0.08), Table 4. When dividing fT4 data into subgroups (low, normal, and high), only the low fT4 group showed increased odds of CRD compared with the normal fT4 group after adjusting for covariates (OR = 1.40, CI: = 1.14–1.72, p = 0.001), Supplementary Table S2. We observed a weak negative correlation between fT4 and PHQ-9, with a correlation coefficient of r = −0.053 (p value <0.001).

Table 4.

Relationship Between Free Thyroxine and Clinically Relevant Depression

	Unadjusted			Adjusted^a
	Odds ratio	CI	p value	Odds ratio	CI	p value
Entire cohort	0.83	0.73, 0.93	0.002	0.90	0.80, 1.01	0.08
Sex
Female	0.81	0.71, 0.93	0.002	0.88	0.77, 1.00	0.05
Male	0.90	0.68, 1.16	0.38	0.94	0.71, 1.23	0.64
Age
<65 years	0.83	0.73, 0.94	0.003	0.89	0.78, 1.01	0.07
≥65 years	0.87	0.62, 1.22	0.42	0.90	0.63, 1.28	0.56
Hypothyroidism diagnosis^b
Yes	0.83	0.68, 1.02	0.081	0.898	0.73, 1.11	0.31
No	0.83	0.72, 0.96	0.01	0.88	0.76, 1.02	0.08
Thyroid hormone replacement
Yes	0.82	0.68, 0.98	0.03	0.88	0.73, 1.06	0.17
No	0.85	0.72, 0.99	0.04	0.89	0.76, 1.04	0.14
Mood disorder diagnosis
Yes	0.88	0.75, 1.03	0.1090	0.92	0.79, 1.09	0.33
No	0.83	0.69, 0.99	0.04	0.88	0.73, 1.05	0.16
Use of antidepressants
Yes	0.89	0.76, 1.04	0.15	0.93	0.79, 1.09	0.38
No	0.81	0.68, 0.98	0.03	0.85	0.70, 1.02	0.08

Adjusted for age, sex, BMI, Charlson score, hypothyroidism, thyroid hormone replacement, and antidepressant use.

Adjusted for age, sex, BMI, Charlson score, thyroid hormone replacement, and antidepressant use.

Discussion

The lifetime prevalence of depression in Utah is higher than the national average, at 26.1% compared with 21.4%, respectively.²⁷ Our cohort had a much higher prevalence (49.7%) of depression compared with the general population in Utah. This could reflect the study cohort as TSH is often conducted as a part of depression workup. In this retrospective cross-sectional study, we found that increasing TSH levels were associated with an increased odds of CRD. Although the association is significant, it is small as follows: a 1-point increase in TSH corresponds to a 1.2% increase in the odds of developing CRD. This was observed in females, all age groups, in those with a diagnosis of hypothyroidism, on THR, patients with or without mood disorders, and those not on ADs. TSH level was associated with an increase in the odds of fatigue as well, a common symptom of depression and hypothyroidism. The finding of an association between low TSH group and increased odds of CRD aligns with recent studies that have reported a similar relationship.^11,28

We replicated previously reported associations between SCH and increased odds of CRD. A cross-sectional study (n = 137) from Brazil used the Beck Depression Scale and reported a depression prevalence 2.3 times higher in individuals with SCH compared with the euthyroid group (45.6% vs. 20.9%, p = 0.006).²⁹ A 2014 Italian case–control study (n = 246) used the Hamilton Depression Scale (HAM-D) and the Montgomery-Asberg Depression Rating Scale (MADRS). They reported a higher prevalence of depression of 63.4% (HAM-D) and 64.2% (MADRS) in patients with SCH and a lower prevalence of depression of 27.6% (HAM-D) and 29.3% (MADRS) in controls without thyroid disease. For both scales, there was a significant difference between the groups (CI:, 51.5%–75.5%, p < 0.00).³⁰ A 2016 Chinese case–control study (n = 216) used HAM-D and reported a significant increase in depression among SCH patients compared with euthyroid patients after radioactive iodine therapy for Graves’ hyperthyroidism (OR = 3.15, CI: 1.20–6.97, p = 0.02).³¹ While similar results have been reported in other studies,^32

–35 some studies have found no association.^36

–42 A recent study utilized data from various European and East Asian databases to conduct a two-sample bidirectional MR analysis, examining the relationship between MDD and thyroid diseases in individuals of European and East Asian ancestries. They reported a significant causal association between MDD and the risk of hypothyroidism and thyroiditis, with alcohol use and ADs serving as potential mediators. This association was observed only in individuals of European ancestry.²⁵ However, they found no causal association between normal TSH with MDD risk. In a meta-analysis by Bode et al.,⁴³ the authors reported a weak association between SCH and depression, with an OR of 1.13 (CI:, 1.01–1.28), showing more conservative estimates in sensitivity analysis. In our cross-sectional study, SCH was associated with increased odds of CRD, although the OR was modest (1.02), possibly due to differences in study population, design, and the use of a self-reported PHQ-9 scale as an outcome measure. Unfortunately, we did not have TPO Ab data available for the entire cohort, as it was only collected in 4.9% of participants. In addition, fT4 data were available for only 24% of the cohort, further limiting the generalizability of our findings among patients with SCH. Findings for individuals of European ancestry (White population) were consistent with those observed in the entire cohort. However, we lacked exclusive data on individuals of East Asian ancestry, and information on alcohol intake was not available. The difference in findings may be due to variations in study design and heterogeneity of the study populations.

The association between thyroid dysfunction and depression has been well known for decades. Psychiatrists have been prescribing thyroid hormones, both LT4 and T3, to treat depression in patients with mood disorders.^44,45 Supraphysiologic doses of LT4 (maximum dose: 500 mcg/day) and T3 (maximum dose: 62.5 mcg/day) have been investigated in RCTs and observational studies to be useful in treatment-resistant depression along with ADs, when first-line medications have been unsuccessful.⁴⁶ Based on the above data, THR has been proposed to improve depression symptoms, but several studies have published conflicting results. A recent meta-analysis from 2018 reported a null effect of LT4 on depression in patients with SCH.²⁰ Another meta-analysis from 2018 did not find a beneficial effect of THR to improve general quality of life and thyroid-related symptoms in nonpregnant adults with SCH.⁴⁷ The findings remain equivocal at the moment and need further investigation. In our cohort, TSH levels were associated with an increased odds of CRD among patients on THR. This may be attributed to underlying thyroid dysfunction.

We conducted a subgroup analysis to investigate the association of fT4 levels with CRD. We found a borderline association suggesting that lower fT4 levels were associated with higher odds of CRD. This is consistent with previous studies that reported low fT4 levels being associated with depression.^11,48 The weak association could be due to the small sample size of the subcohort.

Several experimental studies have been conducted to investigate the brain serotonin system, thyroid dysfunction, and the effects of THR on the brain.⁴⁹ Experimentally induced hypothyroidism has been found to result in increased turnover of brainstem serotonin (5-hydroxytryptamine [5-HT]), accompanied by a decrease in cortical 5-HT concentrations and 5-HT2A receptor density. Furthermore, evidence indicates that thyroid hormones elevate cortical 5-HT concentrations and desensitize autoinhibitory raphe 5-HT1A receptors, leading to the disinhibition of cortical and hippocampal 5-HT release. In addition, thyroid hormones may enhance cortical 5-HT2 receptor sensitivity. These collective findings underscore the modulatory impact of thyroid hormones on the brain 5-HT system, providing support for their AD effects in mood disorders. In human subjects, neuroendocrine challenge studies conducted on hypothyroid patients have revealed a diminished 5-HT responsiveness, a condition that proves reversible with THR.

Despite experimental studies in animals and neuroendocrine challenge and functional imaging studies⁵⁰ in humans showing a positive association between hypothyroidism and depression, observational studies in humans have shown variable results. This is most likely due to differences in several factors, including (1) age groups, (2) race and ethnicity, (3) socioeconomic and educational status, (4) depression evaluation scales, (5) TSH assay and diurnal/seasonal variation, (6) diagnostic cutoffs for SCH, and (7) study methodology. In our cross-sectional study, we found an association between thyroid dysfunction as follows: low (OR, 1.19, CI:, 1.04–1.37) and high TSH (OR, 1.26, CI:, 1.13–1.40) with CRD. Our findings suggest that patients with high as well as low TSH should be considered for depression screening, with additional evidence needed to support such a decision coming from future research. Personalized health care and appropriate management in collaboration with pertinent specialists are recommended.

Strengths and limitations

Our data’s primary strengths lie in its substantial sample size and the utilization of a population-based study design. We used ICD-10 codes, which may not accurately capture the diagnoses of hypothyroidism and mood disorders from EHR. We used the PHQ-9, a widely recognized and validated scale for screening and assessing depression severity in clinical settings. However, a notable limitation of our study is its cross-sectional design, which restricts our ability to explore the dynamic, time-dependent relationship between TSH and depression. Despite our efforts to adjust for potential confounders, the potential for unmeasured confounding remains. Comparison study results are from different ethnic groups in which genetics and cultural differences may have a significant effect on diagnosis of mood disorders. Hence, these results have to be interpreted with caution. Hormonal changes in women during their peri-menopausal period may cause multiple symptoms, thus, increasing the PHQ-9 score. We were unable to account for this as accurate data on menopausal status and/or hormone replacement therapy are challenging to obtain in our electronic data extraction software. In euthyroid sick syndrome (ESS), TSH levels may transiently elevate and normalize within 4–6 weeks. This could underestimate some of the association. It is difficult to identify such cases within cross-sectional data. A longitudinal study could help identify cases of ESS. TSH testing is frequently conducted as part of routine depression screening, which may explain the high prevalence of CRD (49.7%) in our cohort. However, this study cannot establish the cause of the elevated prevalence of CRD within this sample. fT4 data were only available for 24.3% of the cohort, thus finding from this subcohort would be considered hypothesis generating. T3 and T4 may be associated with the severity of mood disorders; however, in our cohort, fT3, total T3, and total T4 data were available for only 1897 (5.7%), 1117 (3.4%), and 445 (1.3%) subjects, respectively. Therefore, we were unable to explore the relationship between T3, total T4, and depression. The relationship between these variables needs further investigation.

Conclusions

In this large retrospective cross-sectional study, thyroid dysfunction was associated with depression. Increasing TSH levels were associated with a small increase in the odds of CRD, especially among women, and patients who were not prescribed ADs or were on THR. Future longitudinal cohort studies are recommended to investigate the association between TSH and CRD, with an emphasis on sex-based differences.

Footnotes

Authors’ Contributions

B.S.: Conceptualization, methodology, data curation, writing—original draft, review, and editing. A.V.B.: Formal analysis. M.N.: Software (electronic data extraction). V.S.: Conceptualization, methodology, writing—original draft, review, and editing. All authors have given approval for the final version of the article.

Author Disclosure Statement

B.S. reports research grant support from Mayo Clinic, the National Network of Depression Centers (NNDC), and Breakthrough Discoveries for Thriving with Bipolar Disorder (BD2). He is a KL2 Mentored Career Development Program scholar, supported by CTSA Grant Number KL2TR002379 from the National Center for Advancing Translational Science (NCATS). He has received honoraria (to Mayo Clinic) from Elsevier for editing a Clinical Overview on Treatment-Resistant Depression. A.V.B: No disclosures. M.N.: No disclosures. V.S. is a consultant on osteoporosis for Blueprint Medicines (2024–2025). He was the consultant on osteoporosis for National Academies of Sciences, Engineering, and Medicine (2024). He received a patient-education research grant from Radius Health, Inc. (2019–2022).

Funding Information

This publication was supported by CTSA Grant Number KL2 TR002379 from the NCATS. Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the NIH.

Supplementary Material

Supplementary Table S1

Supplementary Table S2

References

Hollowell

, Staehling

, Flanders

, et al. Serum TSH, T(4), and thyroid antibodies in the United States population (1988 to 1994): National Health and Nutrition Examination Survey (NHANES III). J Clin Endocrinol Metab, 2002; 87(2):489–499; doi: 10.1210/jcem.87.2.8182

Harvard Medical School. National Comorbidity Survey (NCS). 2007. Available from: https://www.hcp.med.harvard.edu/ncs/index.php [Last accessed: June 17, 2021 ].

Aoki

, Belin

, Clickner

, et al. Serum TSH and total T4 in the United States population and their association with participant characteristics: National Health and Nutrition Examination Survey (NHANES 1999-2002). Thyroid, 2007; 17(12):1211–1223; doi: 10.1089/thy.2006.0235

Kajantie

, Phillips

, Osmond

, et al. Spontaneous hypothyroidism in adult women is predicted by small body size at birth and during childhood. J Clin Endocrinol Metab, 2006; 91(12):4953–4956; doi: 10.1210/jc.2006-1093

Regier

, Narrow

, Rae

, et al. The de facto US mental and addictive disorders service system. Epidemiologic catchment area prospective 1-year prevalence rates of disorders and services. Arch Gen Psychiatry, 1993; 50(2):85–94; doi: 10.1001/archpsyc.1993.01820140007001

Patten

, Wang

, Williams

, et al. Descriptive epidemiology of major depression in Canada. Can J Psychiatry, 2006; 51(2):84–90; doi: 10.1177/070674370605100204

Seedat

, Scott

, Angermeyer

, et al. Cross-national associations between gender and mental disorders in the World Health Organization World Mental Health Surveys. Arch Gen Psychiatry, 2009; 66(7):785–795; doi: 10.1001/archgenpsychiatry.2009.36

Davis

, Tremont

. Neuropsychiatric aspects of hypothyroidism and treatment reversibility. Minerva Endocrinol, 2007; 32(1):49–65.

Brownlie

, Rae

, Walshe

, et al. Psychoses associated with thyrotoxicosis - ‘thyrotoxic psychosis.’ A report of 18 cases, with statistical analysis of incidence. Eur J Endocrinol, 2000; 142(5):438–444; doi: 10.1530/eje.0.1420438

10.

Ceresini

, Lauretani

, Maggio

, et al. Thyroid function abnormalities and cognitive impairment in elderly people: Results of the Invecchiare in Chianti study. J Am Geriatr Soc, 2009; 57(1):89–93; doi: 10.1111/j.1532-5415.2008.02080.x

11.

Kumar

, LeMahieu

, Stan

, et al. The association between thyroid stimulating hormone and depression: A historical cohort study. Mayo Clin Proc, 2023; 98(7):1009–1020; doi: 10.1016/j.mayocp.2022.12.020

12.

Bauer

, Whybrow

. Thyroid hormone, neural tissue and mood modulation. World J Biol Psychiatry, 2001; 2(2):59–69; doi: 10.3109/15622970109027495

13.

Jackson

. The thyroid axis and depression. Thyroid, 1998; 8(10):951–956; doi: 10.1089/thy.1998.8.951

14.

Ittermann

, Volzke

, Baumeister

, et al. Diagnosed thyroid disorders are associated with depression and anxiety. Soc Psychiatry Psychiatr Epidemiol, 2015; 50(9):1417–1425; doi: 10.1007/s00127-015-1043-0

15.

Bekkering

, Agoritsas

, Lytvyn

, et al. Thyroid hormones treatment for subclinical hypothyroidism: A clinical practice guideline. BMJ, 2019; 365:l2006; doi: 10.1136/bmj.l2006

16.

Garber

, Cobin

, Gharib

, et al; American Association of Clinical Endocrinologists and American Thyroid Association Taskforce On Hypothyroidism in Adults. Clinical practice guidelines for hypothyroidism in adults: Cosponsored by the American Association of Clinical Endocrinologists and the American Thyroid Association. Thyroid, 2012; 22(12):1200–1235; doi: 10.1089/thy.2012.0205

17.

Jonklaas

, Bianco

, Bauer

, et al; American Thyroid Association Task Force on Thyroid Hormone Replacement. Guidelines for the treatment of hypothyroidism: Prepared by the American Thyroid Association Task Force on thyroid hormone replacement. Thyroid, 2014; 24(12):1670–1751; doi: 10.1089/thy.2014.0028

18.

Pearce

, Brabant

, Duntas

, et al. 2013 ETA guideline: Management of subclinical hypothyroidism. Eur Thyroid J, 2013; 2(4):215–228; doi: 10.1159/000356507

19.

Cohen

, Sommer

, Vuckovic

. Antidepressant-resistant depression in patients with comorbid subclinical hypothyroidism or high-normal TSH levels. Am J Psychiatry, 2018; 175(7):598–604; doi: 10.1176/appi.ajp.2017.17080949

20.

Zhao

, Chen

, Zhao

, et al. Subclinical hypothyroidism and depression: A meta-analysis. Transl Psychiatry, 2018; 8(1):239; doi: 10.1038/s41398-018-0283-7

21.

Tang

, Wang

, Yang

, et al. Subclinical hypothyroidism and depression: A systematic review and meta-analysis. Front Endocrinol (Lausanne), 2019; 10:340; doi: 10.3389/fendo.2019.00340

22.

Bakian

, Huber

, Scholl

, et al. Dietary creatine intake and depression risk among U.S. adults. Transl Psychiatry, 2020; 10(1):52; doi: 10.1038/s41398-020-0741-x

23.

Kroenke

, Spitzer

, Williams

. The PHQ-9: Validity of a brief depression severity measure. J Gen Intern Med, 2001; 16(9):606–613; doi: 10.1046/j.1525-1497.2001.016009606.x

24.

Shewan

, Coats

. Adherence to ethical standards in publishing scientific articles: A statement from the International Journal of Cardiology. Int J Cardiol, 2012; 161(3):124–125; doi: 10.1016/j.ijcard.2012.10.021

25.

Zhang

, Lu

, Luo

, et al. The causal relationship between major depression disorder and thyroid diseases: A Mendelian randomization study and mediation analysis. J Affect Disord, 2024; 359:287–299; doi: 10.1016/j.jad.2024.05.097

26.

Rothman

. No adjustments are needed for multiple comparisons. Epidemiology, 1990; 1(1):43–46.

27.

Health Indicator Report of Depression: Adult Prevalence. Utah Department of Health and Human Services, Division of Data, Systems and Evaluation, Indicator-Based Information System for Public Health. July 12, 2024. Available from: http://ibis.health.utah.gov/

28.

Varella

, Bensenor

, Janovsky

, et al. Thyroid-stimulating hormone levels and incident depression: Results from the ELSA-Brasil study. Clin Endocrinol (Oxf), 2021; 94(5):858–865; doi: 10.1111/cen.14407

29.

Almeida

, Brasil

, Costa

, et al. Subclinical hypothyroidism: Psychiatric disorders and symptoms. Braz J Psychiatry, 2007; 29(2):157–159; doi: 10.1590/s1516-44462007000200013

30.

Demartini

, Ranieri

, Masu

, et al. Depressive symptoms and major depressive disorder in patients affected by subclinical hypothyroidism: A cross-sectional study. J Nerv Ment Dis, 2014; 202(8):603–607; doi: 10.1097/NMD.0000000000000168

31.

, Tian

, Yuan

, et al. Subclinical hypothyroidism after 131i-treatment of Graves’ disease: A risk factor for depression? PLoS One, 2016; 11(5):e0154846; doi: 10.1371/journal.pone.0154846

32.

Manciet

, Dartigues

, Decamps

, et al. The PAQUID survey and correlates of subclinical hypothyroidism in elderly community residents in the southwest of France. Age Ageing, 1995; 24(3):235–241; doi: 10.1093/ageing/24.3.235

33.

Chueire

, Romaldini

, Ward

. Subclinical hypothyroidism increases the risk for depression in the elderly. Arch Gerontol Geriatr, 2007; 44(1):21–28; doi: 10.1016/j.archger.2006.02.001

34.

Chueire

, Silva

, Perotta

, et al. High serum TSH levels are associated with depression in the elderly. Arch Gerontol Geriatr, 2003; 36(3):281–288; doi: 10.1016/s0167-4943(02)00173-5

35.

Larisch

, Kley

, Nikolaus

, et al. Depression and anxiety in different thyroid function states. Horm Metab Res, 2004; 36(9):650–653; doi: 10.1055/s-2004-825925

36.

Fjaellegaard

, Kvetny

, Allerup

, et al. Well-being and depression in individuals with subclinical hypothyroidism and thyroid autoimmunity - a general population study. Nord J Psychiatry, 2015; 69(1):73–78; doi: 10.3109/08039488.2014.929741

37.

Roberts

, Pattison

, Roalfe

, et al. Is subclinical thyroid dysfunction in the elderly associated with depression or cognitive dysfunction? Ann Intern Med, 2006; 145(8):573–581; doi: 10.7326/0003-4819-145-8-200610170-00006

38.

Engum

, Bjoro

, Mykletun

, et al. An association between depression, anxiety and thyroid function–a clinical fact or an artefact? Acta Psychiatr Scand, 2002; 106(1):27–34; doi: 10.1034/j.1600-0447.2002.01250.x

39.

Gussekloo

, van Exel

, de Craen

, et al. Thyroid status, disability and cognitive function, and survival in old age. JAMA, 2004; 292(21):2591–2599; doi: 10.1001/jama.292.21.2591

40.

Kim

, Zhang

, Chang

, et al. Subclinical hypothyroidism and incident depression in young and middle-age adults. J Clin Endocrinol Metab, 2018; 103(5):1827–1833; doi: 10.1210/jc.2017-01247

41.

Park

, Lee

, et al. Subclinical hypothyroidism (SCH) is not associated with metabolic derangement, cognitive impairment, depression or poor quality of life (QoL) in elderly subjects. Arch Gerontol Geriatr, 2010; 50(3):e68–e73; doi: 10.1016/j.archger.2009.05.015

42.

Jorde

, Waterloo

, Storhaug

, et al. Neuropsychological function and symptoms in subjects with subclinical hypothyroidism and the effect of thyroxine treatment. J Clin Endocrinol Metab, 2006; 91(1):145–153; doi: 10.1210/jc.2005-1775

43.

Bode

, Ivens

, Bschor

, et al. Association of hypothyroidism and clinical depression: A systematic review and meta-analysis. JAMA Psychiatry, 2021; 78(12):1375–1383; doi: 10.1001/jamapsychiatry.2021.2506

44.

Nunez

, Joseph

, Pahwa

, et al. Augmentation strategies for treatment resistant major depression: A systematic review and network meta-analysis. J Affect Disord, 2022; 302:385–400; doi: 10.1016/j.jad.2021.12.134

45.

Aronson

, Offman

, Joffe

, et al. Triiodothyronine augmentation in the treatment of refractory depression. A meta-analysis. Arch Gen Psychiatry, 1996; 53(9):842–848; doi: 10.1001/archpsyc.1996.01830090090013

46.

Singh

, Sundaresh

. Thyroid hormone use in mood disorders: Revisiting the evidence. J Clin Psychiatry, 2022; 83(5); doi: 10.4088/JCP.22ac14590

47.

Feller

, Snel

, Moutzouri

, et al. Association of thyroid hormone therapy with quality of life and thyroid-related symptoms in patients with subclinical hypothyroidism: A systematic review and meta-analysis. JAMA, 2018; 320(13):1349–1359; doi: 10.1001/jama.2018.13770

48.

Odawara

, Shimbo

, Yamauchi

, et al. Association of low-normal free T4 levels with future major depression development. J Endocr Soc, 2023; 7(8):bvad096; doi: 10.1210/jendso/bvad096

49.

Bauer

, Heinz

, Whybrow

. Thyroid hormones, serotonin and mood: Of synergy and significance in the adult brain. Mol Psychiatry, 2002; 7(2):140–156; doi: 10.1038/sj.mp.4000963

50.

Constant

, de Volder

, Ivanoiu

, et al. Cerebral blood flow and glucose metabolism in hypothyroidism: A positron emission tomography study. J Clin Endocrinol Metab, 2001; 86(8):3864–3870; doi: 10.1210/jcem.86.8.7749