Thyroid Abnormalities Linked to Bipolar Disorder in Large Retrospective Study

The butterfly-shaped gland at the base of the neck has long been suspected of whispering into the mind. Now a large retrospective analysis from Nanjing, China, has added fresh weight to that suspicion, reporting that people with bipolar disorder show measurably different thyroid structure and function compared with psychiatrically healthy individuals. The study, published in BMC Psychiatry by a team at Affiliated Nanjing Brain Hospital of Nanjing Medical University, examined clinical records from 593 participants and found that bipolar patients carried a higher burden of thyroid nodules, shifted patterns of thyroid hormones, and more frequently tested positive for thyroid autoantibodies than matched controls.

Bipolar disorder is one of the most disabling psychiatric conditions worldwide, characterized by alternating episodes of mania or hypomania and depression. Its causes remain only partially understood, but decades of research have pointed to the hypothalamic–pituitary–thyroid axis, the hormonal circuit that links the brain’s endocrine command centers to the thyroid gland, as a plausible contributor. Thyroid hormones regulate cerebral metabolism, neurotransmitter signaling, and neurodevelopment, and overt thyroid disease is well known to produce mood disturbances of its own. What has remained murky is whether subtler thyroid deviations, detectable in routine clinical testing, reliably accompany bipolar illness and whether they track the severity of manic or depressive symptoms.

To probe that question, the researchers combed through records collected between January 2022 and December 2024, identifying 398 patients diagnosed with bipolar disorder and 195 healthy controls. Thyroid structure was assessed with ultrasound, documenting the presence of nodules and gland size, while function was evaluated through a panel of nine serum parameters. These included total and free forms of the two principal thyroid hormones, triiodothyronine and thyroxine, thyroid stimulating hormone, thyroglobulin, and a battery of autoantibodies targeting thyroglobulin, thyroid peroxidase, and the TSH receptor. Clinical symptoms were quantified with standardized rating scales, and the team applied logistic regression to test associations between thyroid measures and bipolar diagnosis, along with Spearman correlation to explore links with symptom scores.

The results painted a consistent picture of thyroid deviation in the bipolar group. Compared with controls, patients showed a higher prevalence of structural abnormalities on ultrasound, lower levels of total triiodothyronine and total thyroxine, elevated thyroid stimulating hormone, and higher rates of positivity for both thyroglobulin antibodies and thyroid peroxidase antibodies. Taken together, the pattern hints at a gland working under strain: slightly underactive in its hormone output, pushed harder by the pituitary’s signal, and more often under immunological attack. Such findings echo earlier clinical observations that subclinical hypothyroidism and autoimmune thyroiditis appear more frequently in mood disorders than in the general population.

Yet the study’s multivariate analysis complicated the story in instructive ways. When the researchers adjusted for multiple variables simultaneously, several associations shifted or lost statistical significance. In the adjusted models, the presence of thyroid nodules, normal or elevated levels of total triiodothyronine, total thyroxine, and thyroglobulin, and positive thyroglobulin antibodies appeared to have a protective effect on the outcomes examined, though some of these effects did not survive full scrutiny. This attenuation is a familiar hazard of cross-sectional endocrine research: thyroid parameters are interdependent, influenced by age, sex, iodine intake, medication, and body weight, and isolating the independent contribution of any single measure demands large, carefully stratified samples.

Perhaps the most sobering finding came from the correlation analysis. Despite the clear group-level differences in thyroid structure and function, the researchers found no significant relationship between any thyroid parameter and the severity of manic or depressive symptoms as measured by the standardized scales. In other words, knowing a patient’s TSH level or antibody status did not predict how ill they were at the time of assessment. This absence of a dose-response relationship underscores that whatever role the thyroid plays in bipolar disorder is unlikely to be a simple, linear one, where more hormone abnormality equals more mood disturbance.

The authors are candid about this complexity. Their conclusion emphasizes that thyroid structure and function abnormalities were associated with bipolar disorder in this retrospective analysis, but that several associations weakened after multivariable adjustment and that no significant cross-sectional correlation with symptom severity emerged. They call for further mechanistic investigation to untangle whether thyroid changes are a cause of the illness, a consequence of it or its treatment, or a correlated marker of shared underlying biology. Lithium, the cornerstone mood stabilizer in bipolar therapy, is well documented to impair thyroid function, and the retrospective design could not fully disentangle medication effects from disease effects.

Methodologically, the study has both strengths and limits worth weighing. Its sample of nearly 600 participants with a nine-parameter hormonal panel and ultrasound imaging is unusually comprehensive for this literature, and the use of standardized symptom scales and multivariate statistics reflects a rigorous approach. On the other hand, a retrospective cross-sectional design captures only a single moment in time, cannot establish causality, and relies on the quality of recorded clinical data. The ethics review at Nanjing Brain Hospital approved the work with a waiver of written informed consent given the anonymized nature of the records, and the analysis was conducted in accordance with the Declaration of Helsinki.

For clinicians, the practical message is nuanced. The findings do not support using thyroid testing to gauge mood symptom severity, but they do reinforce the value of monitoring thyroid health in bipolar patients, many of whom take medications with endocrine side effects. The elevated rates of thyroid autoantibody positivity in particular suggest that autoimmune thyroid disease deserves attention in this population, since untreated thyroid dysfunction can itself destabilize mood and complicate psychiatric management.

For researchers, the study adds to a growing body of evidence that the boundary between endocrinology and psychiatry is porous. Future work, the authors suggest, should move beyond cross-sectional snapshots toward longitudinal designs that track thyroid parameters and mood over time, ideally with mechanistic studies of how thyroid hormones influence neural circuits governing emotion regulation. Until then, the thyroid remains a compelling but enigmatic character in the biology of bipolar disorder, clearly present in the picture, yet not yet deciphered in the plot.

Subject of Research: Association between thyroid structural and functional abnormalities and bipolar disorder

Article Title: Clinical characteristics of thyroid structure and functional abnormalities in bipolar disorder and their association with mood symptoms: a retrospective cross-sectional study

Article References: Song, J., Liu, H., Wu, D., & Gong, K. (2026). Clinical characteristics of thyroid structure and functional abnormalities in bipolar disorder and their association with mood symptoms: a retrospective cross-sectional study. BMC Psychiatry. https://doi.org/10.1186/s12888-026-08679-6

Image Credits: AI Generated

DOI: 10.1186/s12888-026-08679-6

Keywords: bipolar disorder, thyroid function, thyroid nodules, TSH, thyroid autoantibodies, hypothalamic-pituitary-thyroid axis, retrospective study, psychiatry, neuroendocrinology, mood symptoms, Clinical, characteristics