to the editor: Thank you for the article on thyroiditis in the May 15, 2006, issue of American Family Physician (AFP).¹ I would like to comment on Figure 1 in the article, an algorithm for the diagnosis of suspected thyroiditis. The authors suggest that the question “Is the patient postpartum?” should be answered at a key decision point. Based on a “yes” answer to that question, combined with low thyroid-stimulating hormone and high radioactive iodine uptake levels, the suggested diagnosis is Graves' disease.

The implication here is that Graves' disease only affects postpartum women. As noted in another feature in the same issue, Graves' affects adult women primarily, with a female-to-male incidence ratio of at least 5:1, and may occur sporadically in adolescent females.² Based on the autoimmune nature of Graves' disease, pregnancy could certainly play a role. However, other etiologies may include genetic susceptibility, stress, and high iodine intake. A recent prospective cohort study demonstrated smoking to be an independent risk factor for Graves' in women.³ A previous issue of AFP offered an alternate algorithm for diagnosing hyperthyroidism.⁴ Family physicians should not exclude a diagnosis of Graves' disease in a patient with hyperthyroidism simply because she (or he) is not postpartum.

in reply: Graves' disease most commonly affects women of reproductive age, and the majority will have the onset of the condition when they are not pregnant. Some patients with preexisting Graves' disease will improve during pregnancy, only to experience an exacerbation in the postpartum period when the immunosuppressive effects of pregnancy are gone. Other women will have the initial onset of Graves' disease in the postpartum period. When a woman presents with hyperthyroidism during the year following a pregnancy, it is important to differentiate between Graves' disease and the more common postpartum thyroiditis because the therapies for each condition differ. For this reason, we emphasized the postpartum period in Figure 1 of our article, which was an algorithm for the diagnosis of suspected thyroiditis.¹ The text of our article provides clinical and imaging clues to allow the physician to make the correct diagnosis.¹