Postpartum thyroid dysfunction and postpartum depression:: are they two linked disorders?

OBJECTIVE Postpartum has been considered as a period of risk for developing postpartum depression (PD) by some but not all authors, and this PD has been linked with postpartum thyroid dysfunction (PPTD). The major aim of this study was to evaluate the relation between the presence of PPTD and PD. DESIGN and PATIENTS Six hundred and forty-one healthy Caucasian women recruited between their 36th week of pregnancy and fourth day postpartum underwent clinical and laboratory evaluation and were checked again at 1 (n = 605), 3 (n = 552), 6 (n = 574), 9 (n = 431), and 12 (n = 444) months postpartum. MEASUREMENTS At baseline and at each clinical evaluation, Beck Depression Inventory (BDI) was administered to screen PD. The definitive diagnoses of PD was performed by a psychiatrist according to the DSM-III-R criteria. At each visit, we determined serum free T-4 and TSH concentrations. Thyroperoxidase and thyroglobulin antibodies were determined only in patients with abnormal hormone concentrations. Postpartum thyroiditis (PPT) was considered to be present in women with overt or subclinical transient hyperthyroidism between 1 and 3 months postpartum and/or overt or subclinical hypothyroidism between 3 and 6 months postpartum. RESULTS Fifty-six women developed postpartum thyroid dysfunction (PPTD), corresponding to an incidence rate of 11%: 45 with PPT [incidence rate 7.8%; confidence interval (CI) 5.6-10%], eight with Graves' disease (incidence rate 1.5%; Cl 0.5-2.5%) and three with nonpalpable toxic thyroid adenoma (incidence rate 0.5%; Cl 0-1.5%). Five hundred and eighty of the evaluated women (incidence rate 90.5%; Cl 95% 88.2-92.8) presented BDI scores below 21 and therefore the PD diagnoses was excluded. In 50 cases (incidence rate 7.8%; Cl 95% 5.7-9.8), we detected a BDI score over 21 in some evaluations, but the PD diagnosis was not confirmed. Another 11 (incidence rate 1.7%; Cl 95% 0.7-2.7) were diagnosed as having PD and required psychiatric treatment. None of the PPTD was diagnosed as having PD. The BDI scores frequency over 21 was similar between healthy women and those with PPTD. Patients with a previous history of depression developed PD more often (P < 0.0001). One hundred and ninety women breast fed their babies for more than 2 months, without observing a higher PD rate or BDI scores over 21 (P = 0.5). CONCLUSIONS We found a general PD incidence rate of 1.7% in our group of patients. This figure is not higher in women with hormone abnormalities caused by PPTD. Women with a past history of depression present a higher risk of PD while those who breast fed did not have an increased risk.