Incidence of thyroid disease associated with oral contraceptives.

Incidence of thyroid disease associated with oral contraceptives Cell-mediated irnmunosuppression in pregnancy and in women taking oral contraceptives has been described.1 2 Temporary remission of autoimmune thyroid disease with postpartum recurrence has also been reported associated with the gradual reduction of serum antithyroid antibody titres as the pregnancy progressed, followed by a postpartum increase reaching a peak three to four months after delivery.3 Study and results Thyroid diseases were analysed as part of a continuing major prospective survey of oral contraceptives. The study began in 1968. About 23 000 women currently taking a contraceptive pill and a similar number of controls who had never taken an oral contraceptive were recruited by 1400 general practitioners over a period of 14 months. Observations of subsequent morbidity and other relevant data of this population formed the basis for our analyses. The design of the study and the interpretation of the findings have been described elsewhere.4 The main findings are shown in the table. Because of the close interrelationship of thyroid disorders we analysed only the frequency of reporting of the first diagnosis during the study of any thyroid disease in one patient. Cases first diagnosed during pregnancy were excluded together with the associated periods of observations. Patients were categorised according to their contraceptive status at the time of diagnosis as (1) takers (those currently taking an oral contraceptive who had had no break in use longer than one cycle); (2) ex-takers (those who had formerly used oral contraceptives); and (3) controls (those who had never used oral contraceptives). All three clinical groups of thyroid disease were reported less frequently in takers than in controls. In patients with euthyroid swelling the rate in takers was 69 'O of that in controls, and the proportions for thyrotoxicosis and myxoedema were 71 %/1 and 57 ON respectively. Although these differences taken separately do not reach statistical significance, the frequency of reporting in takers was consistently materially less than in controls. When the subcategories are combined the risk ratio of 0-68 becomes highly significant (P< 001; 950% confidence limits 0-52 to 0 85). An important finding was that the rates for thyroid disease in ex-takers did not differ from controls. The frequency of reporting of thyroid disease showed no correlation with age, parity, cigarette consumption, or social class in any of the contraceptive categories, and reporting rates did not correlate with oestrogen or progestogen dosage or with …