Evaluation of the Association Between St. John's Wort
Study Objective. To evaluate whether St. John's wort is associated with an elevation in thyroid-stimulating hormone (TSH) levels.
Design. Retrospective case-control study.
Setting. Hamilton, Ontario; September 1998-May 1999.
Subjects. Thirty-seven subjects with elevated TSH levels and 37 age- (± 6 yrs) and gender-matched subjects with normal TSH levels who resided in the same region.
Measurements and Main Results. Exposure to St. John's wort during the 3-6 months before TSH measurement was assessed through a telephone interview. A total of 4 of 37 subjects with elevated TSH levels and 2 of 37 subjects with normal TSH levels had taken St. John's wort in that time period. The odds ratio for elevated TSH levels associated with taking St. John's wort was 2.12 (95% confidence interval [CI] 0.36-12.36).
Conclusion. This study suggests a probable association between St. John's wort and elevated TSH levels. However, because of the wide CIs surrounding the point estimate (which crosses unity), further investigation, including a large, prospective cohort study, is warranted.

St. John's wort has been effective in the treatment of mild-to-moderate depression in a number of randomized controlled trials. However, these studies were generally short in duration, with follow-up periods of 2-12 weeks, which may not permit adequate evaluation of long-term adverse effects.

Immediate adverse effects associated with St. John's wort include dry mouth, constipation, gastrointestinal symptoms, confusion, and photosensitivity. However, little is known about adverse events associated with long-term therapy with this herb, which has led to concerns about its safety. Recent case reports have addressed drug interactions with St. John's wort.

St. John's wort can cause symptoms of a serotonin syndrome, and evidence suggests that the herb inhibits serotonin reuptake. Some selective serotonin reuptake inhibitors (SSRIs), such as sertraline, have been associated with changes in thyroid hormone levels.

Two unpublished case reports of patients who received thyroid hormone replacement therapy after partial lobectomy for thyroid cancer indicate that St. John's wort may interfere with the hypothalamic-pituitary-thyroid axis (personal communication, L. Gulenchyn, 1999). In both patients, thyroid-stimulating hormone (TSH) levels became elevated, although thyroxine (T₄) levels remained normal, during exposure to St. John's wort. After discontinuation of the herb, TSH levels subsequently returned to the normal range (0.5-5.0 µU/ml). In one patient, TSH dropped from 18.63 to 5.37 µU/ml 1 month after discontinuing St. John's wort, and subsequently to 1.32 µU/ml 6 months later. In the second patient, TSH decreased from 14.76 to 3.23 µU/ml 1 month after discontinuation of the herb. Both patients experienced abnormal elevated TSH levels on at least two consecutive occasions during exposure to St. John's wort.

In light of evidence that SSRIs might cause thyroid dysfunction and the two case reports mentioned above regarding elevated TSH associated with St. John's wort, we designed a case-control study to assess whether an association exists between St. John's wort and elevated TSH (relative hypothyroidism).