Thyroid trouble may not be the anxiety driver it's made out to be

It is common to hear anxiety blamed on the thyroid, and earlier research seemed to back that up: studies comparing patients with healthy controls suggested that an underactive thyroid, an overactive thyroid and thyroid autoimmunity might all raise anxiety risk. The problem is that those clinic-based samples are prone to selection bias, because the people who end up in them are not representative of the wider population. Population-based studies, which draw on more representative samples, have produced conflicting results. This review set out to settle whether a genuine association exists once the more reliable study designs are pooled together.

The researchers searched three major databases from inception to March 2026 for population-based studies that used laboratory tests or formal diagnostic codes to identify thyroid dysfunction or autoimmunity, paired with anxiety outcomes measured by validated criteria or rating scales. Two reviewers selected studies independently, and study quality was rated with an established scale. From 2,548 screened articles, 12 studies covering 676,658 participants qualified: six examined hypothyroidism, seven hyperthyroidism, and four thyroid autoimmunity. Results were pooled using models that allow for genuine differences between studies, and the analysis was pre-registered before it began.

The pooled estimates were small and mostly failed to reach statistical significance. For hypothyroidism the odds of anxiety were 1.13 times higher, a result whose uncertainty range comfortably included no effect at all. Thyroid peroxidase antibodies, a marker of autoimmunity, showed odds of 1.47 with a similarly wide range that could not rule out no association. Only hyperthyroidism produced a statistically significant result, with odds of 1.38, and even that was modest. The included studies disagreed with each other substantially, and sensitivity and subgroup analyses, along with studies that could only be summarised descriptively, showed the same overall pattern.

This result complicates the older case-control literature, which painted a stronger picture of thyroid conditions driving anxiety. The authors conclude the evidence is genuinely inconclusive: small positive associations seem likely, but the data are equally compatible with moderate links or even small protective effects. Part of the problem is that both thyroid pathology and anxiety were defined differently across studies, and the number of eligible studies was small. Whether a meaningful causal relationship exists, and in which direction, remains an open question that better standardised population studies will need to answer.

If you have a thyroid condition and also struggle with anxiety, this is a reason to be cautious about assuming one is causing the other, because the best population-level evidence available does not clearly support that link. The one signal that did reach significance concerned an overactive thyroid, and even there the increase in odds was small. Equally, the analysis cannot rule out a modest association, so it does not prove the two are unrelated. The honest reading is that the connection between thyroid function and anxiety is far less certain than clinic folklore suggests.