Mortality in patients who have their antihypertensive therapy changed

Medical records were examined for 1935 patients who presented sequentially to a hypertension clinic between 1971 and 1981. Patients were classified according to whether they were on a beta-blocker, methyldopa, a potassium-losing diuretic, or whether they had discontinued any of these treatments. Age-standardized mortality rates were calculated and the relative risks of stopping compared with non-stopping were computed. Those stopping a beta-blocker had a significantly higher mortality in the following year than those who continued, both in men [relative risk (RR) = 5.91, 95% confidence interval (Cl) 2.78-12.56] and women (RR = 5.67, 95% Cl 1.75-18.41). Moreover, women also had a significantly higher mortality when stopping methyldopa, compared with those who continued on the drug (RR = 4.91, 95% Cl 1.82-13.20). However, analysis of data from the years following withdrawal indicated that a high RR was not limited to the first year after the withdrawal of beta-blockers, but was still apparent in the fourth year after stopping. This indicates that the high mortality was not an early function of withdrawal. The high initial RR of mortality in women stopping methyldopa was followed by a substantial decrease in risk over the later years of follow-up. The high mortality in patients stopping particular antihypertensive drugs was not explained by known cardiovascular risk factors.