Objective: To investigate whether there are definable subgroups of patients with essential hypertension who respond specifically to particular antihypertensive drugs.
Design: Randomized cross-over comparison of the antihypertensive effect of 50 mg atenolol per day, 10 mg lisinopril per day and 20 mg nifedipine retard twice a day. Ambulatory blood pressure monitoring was used to assess the blood pressure level both for recruitment and at the end of each treatment period. The treatment periods lasted 4 weeks and were preceded by 4 weeks of placebo.
Patients: Seventy-two untreated hypertensive patients with a mean age of 52 (SD 8.4) years were recruited from six general practices and from the hospital outpatient clinic. Sixty-eight patients completed the trial.
Main outcome measures: To assess the within-patient correlations among the blood pressure responses to each drug and explore the possible role of simple characteristics, such as the initial blood pressure, plasma renin concentration and age, in identifying the responders to a particular drug.
Results: Systolic/diastolic blood pressure fell significantly with each agent (P < 0.001): atenolol reduced it by 16.3 +/- 13.3/9.9 +/- 8.8, lisinopril by 14.8 +/- 15.0/9.4 +/- 9.1 and nifedipine by 11.6 +/- 12.3/6.7 +/- 8.3 mmHg. There was a low degree of correlation between the changes in blood pressure with the three drugs in individual patients. With each drug there was a small percentage (8.9-14.7%) of non-responders. The initial level of systolic blood pressure was weakly correlated with the antihypertensive effect of nifedipine (r = 0.47, P < 0.001) and plasma renin concentration was related to the effect of atenolol (r = 0.32, P < 0.01). Age did not predict the blood pressure response to any agent.
Conclusions: The low level of the correlation between the blood pressure changes with the three drugs suggests that different mechanisms may be involved in the aetiology of essential hypertension. Plasma renin concentration and the initial level of systolic blood pressure contribute only weakly to the identification of responders to the three drugs.