Different Concepts in First-Line Treatment of Essential Hypertension
Comparison of a Low-Dose Reserpine-Thiazide Combination With Nitrendipine Monotherapy
Low-dose combination therapy has been proposed as a rational first-line approach to hypertension treatment. We compared the efficacy and tolerability of the fixed combination of reserpine (0.1 mg) plus the thiazide clopamid (5 mg) with its single components and the calcium-antagonist nitrendipine (20 mg) in a randomized, double-blind, parallel study of 273 hypertensive patients with diastolic blood pressure (BP) between 100 and 114 mm Hg. The four groups did not differ regarding baseline characteristics (mean age, 58 years; 51% men; mean BP after a 2-week placebo period, 158 to 160/103 to 104 mm Hg). After 6 weeks of treatment with one capsule daily, mean reductions in sitting BP from baseline at 24 hours after dosing in the reserpine-clopamid combination, reserpine, clopamid, and nitrendipine groups were −23.0/−17.1, −14.0/−11.7, −13.6/−11.9, and −11.6/−12.3 mm Hg, respectively (2P<.01). The corresponding normalization rates (diastolic BP <90 mm Hg) were 55%, 40%, 36%, and 33% (2P=.11). All patients whose BP had not been normalized at this point received two capsules of the respective medication once daily from weeks 7 to 12. At week 12, mean BP reductions were −25.7/−18.1, −14.6/−12.2, −17.7/−13.4, and −14.9/−15.3 mm Hg in the four groups, respectively (2P<.01). The respective normalization rates were 69%, 35%, 39%, and 45% (2P<.0001). Linear regression modeling indicated that reserpine and clopamid combined acted more than additively. As regards tolerability, adverse experiences were observed in 27%, 28%, 29%, and 48% of patients, respectively (2P<.05). The respective rates of premature discontinuation because of adverse effects were 3%, 3%, 7%, and 13% (2P=.06). In conclusion, a low-dose combination of reserpine and clopamid lowered BP significantly more than both the components alone and nitrendipine. Moreover, the combination was tolerated as well as its components and significantly better than nitrendipine. Thus, the use of this low-dose reserpine-thiazide combination appears to be a rational alternative to conventional monotherapy in the first-line treatment of hypertension.
Fixed-dose combinations have usually been reserved for patients who do not respond adequately to monotherapy, and these combinations generally contain the full conventional doses of each drug. Therefore, fixed-dose combinations as initial therapy have traditionally been discouraged.1 2 Reasons frequently cited include the inability to titrate individual components, difficulty in separating individual toxicities, and unnecessary exposure of patients to superfluous therapy.3 However, the use of low doses of two agents with documented additive antihypertensive effects in first-line treatment is an entirely different approach.4 It is obvious that small doses of drugs in combination could minimize the dose-dependent type A side effects that typically occur when doses are titrated to higher levels5 ; therefore, suitable combinations may indeed offer a favorable benefit-risk ratio. Direct clinical comparisons between low-dose combinations and monotherapy in usual doses could clarify the role of the low-dose combinations; however, such studies are rare.
A typical representative of low-dose combinations, namely, reserpine combined with a thiazide, continues to be frequently prescribed in some countries, such as Germany, Switzerland, and South Africa. Both agents have been used in several large-scale intervention studies, such as the Veterans Administration Cooperative Study Group on Antihypertensive Agents (VACS),6 the Multiple Risk Factor Intervention Trial (MRFIT),7 the Systolic Hypertension in the Elderly Program (SHEP),8 9 and others.10 11 In a recent comparison of antihypertensive drugs in elderly patients, reserpine was found to be equal or superior to the other drugs in terms of effectiveness and tolerability.12 Even more important, treatment with a diuretic in combination with reserpine has been shown to reduce vascular morbidity and mortality in hypertensive patients, whereas the role of newer agents, such as calcium antagonists and angiotensin-converting enzyme inhibitors, still has to be determined in this respect.
The continued use of low-dose reserpine-thiazide combinations in first-line therapy in practice would imply that they still can be regarded as an effective and rational alternative to “modern” monotherapies. However, to date, no studies have been conducted to prove the additive effect of the components; furthermore, direct comparisons with the “newer” antihypertensive agents have yet to be made. Therefore, in this double-blind, parallel-group study, we compared a fixed low-dose combination of reserpine and the thiazide clopamid with its components and nitrendipine, an efficacious and well-tolerated once-daily calcium antagonist,13 14 in terms of efficacy and side effects.
This study was a multicenter, randomized, active-controlled, four-arm, parallel-group trial of the ambulatory antihypertensive efficacy and tolerability of the following medications: a fixed low-dose combination of 0.1 mg reserpine and 5 mg clopamid (1 to 2 capsules once daily), reserpine (0.1 to 0.2 mg), clopamid (5 to 10 mg), and nitrendipine (20 to 40 mg).
Before entering the study, patients discontinued any antihypertensive medication for at least 4 weeks. After a 2-week, single-blind placebo run-in phase, patients were randomized to one of the four treatment groups. The study evaluated 12 weeks of once-daily treatment with the above medications. In patients that did not normalize blood pressure (BP) (diastolic BP [DBP] <90 mm Hg) after 6 weeks of therapy, the dose of the respective medication was doubled from weeks 7 to 12 (Fig 1⇓).
All of the participants had mild to moderate essential hypertension. To qualify for entry into the double-blind phase of the trial, they were to have a DBP of 100 to 114 mm Hg after relaxing in a sitting position for at least 10 minutes. This sitting BP value was determined as the median of at least five valid consecutive measurements with a semiautomated cuff-oscillometric sphygmomanometer (boso-Oscillomat, Bosch+Sohn GmbH & Co). An identical version of this device has been compared with the “gold standard” Hawksley random-zero sphygmomanometer15 and validated according to the criteria of the Association for the Advancement of Medical Instrumentation.16
Participants were at least 18 years old; pregnant or lactating women and those of childbearing potential were excluded. Other criteria for exclusion were participation in another clinical trial within the previous 3 months; drug or alcohol abuse; mental impairment; history of allergy; or a previous adverse event from exposure to either reserpine, thiazide diuretics/sulfonamides, or dihydropyridines. Also excluded were individuals with secondary hypertension, cerebrovascular insult less than 6 weeks previously or known cerebral blood flow disturbances (transient ischemic attack, etc), unstable angina or myocardial infarction in the previous 3 months, heart failure not controlled by digitalis, hemodynamically relevant heart valve disease, colitis or severe gastroenteritis, or a history of depression. Additionally, patients with abnormalities on prestudy laboratory screening were also excluded, especially those with severely impaired renal function (serum creatinine >1.8 mg/dL), hepatic impairment (transaminases >200% above reference limit), severe disorders of lipid metabolism (total cholesterol or triglycerides >350 mg/dL), hypokalemia (<3.5 mmol/L), hyponatremia (<130 mmol/L), hypercalcemia (>2.7 mmol/L), or hyperuricemia (>8.0 mmol/L). Patients were also excluded if they used major psychotropic or antidepressant drugs or nonsteroidal anti-inflammatory agents, steroid hormones, or any agent that could lower BP. The use of digitalis and nitroglycerin was allowed if the dose was maintained throughout the course of the study. Patients were withdrawn from the study if the BP-lowering effect of the test medication was either too strong or insufficient. Patients were also excluded from the study if their compliance was poor, ie, if they had not taken more than three capsules during the placebo phase.
The study design was approved by the local ethics review committee, and participants gave their written informed consent to take part. The study was conducted on outpatients in 18 centers throughout Germany. Patients were reviewed 2 weeks after the single-blind run-in phase and every 3 weeks during the double-blind treatment phase. All trial medication was encapsulated and identical in shape, color, and taste. One or two capsules, respectively, were to be taken in the morning before breakfast, with the exception of the days of clinic visits, when the medication was to be taken after BP measurement.
Sitting BP was measured in accordance with the American Society of Hypertension recommendations17 at all clinic visits approximately 24 hours after the last dose of study medication, ie, between 7 and 9 am, if possible, by the same investigator. The examination started with the patient resting 10 minutes in the sitting position; this was followed by at least five valid automatic BP readings, with 2 minutes between readings. The printout of the measurements was attached to the case report forms. Pulse was counted for 30 seconds after the last BP reading. Patient compliance was checked by means of pill counting during every visit. Standard fasting laboratory tests were performed at the beginning of the single-blind placebo treatment phase and after 6 and 12 weeks of double-blind treatment. A complete physical examination including a 12-lead electrocardiogram was performed at baseline and after 12 weeks of double-blind therapy.
Adverse events were assessed at each visit by recording spontaneous reports and by questioning in a standardized, general manner. Adverse events were graded by the investigator as mild, moderate, or severe, and their relationship to study medication was characterized as “none,” “possibly,” and “definitely.”
Statistical Methods and Power
Comparisons between treatment groups at entry were made with the χ2 test for nominal data, the Kruskal-Wallis test for ordinal data, and ANOVA for continuous data. The primary analyses in this study followed the intention-to-treat principle, which requires all patients who are properly randomized to be included in the analysis regardless of eligibility status. The primary efficacy parameter was the change from baseline in sitting systolic BP (SBP) and DBP at the end of the 6-week treatment on uniform dosage. The working hypothesis stated that a more pronounced reduction in sitting SBP and DBP would occur in the combination group versus the respective components as well as, or along with, a better tolerability compared with the group treated with nitrendipine after week 12. To confirm this assumption, we tested the null hypothesis by means of Dunnett's multiple comparison test. The error term was obtained by a two-factor ANOVA model, with treatment and center as the main factors. One-sided critical regions of the tests were considered in comparing the combination group with its individual components. Two-sided tests were performed in comparing the combination with the nitrendipine group. Categories of antihypertensive response were analyzed with a cumulative logit model for the proportion of patients within each category of response. The incidences of side effects were compared by means of Fisher's exact test. The significance level was set at 5%. To control the experiment error rate at this level, we applied the Bonferroni-Holm procedure to the ranked probability values. To predict the changes from baseline of mean arterial BP for different baseline values, we performed a linear regression analysis. Baseline BP served as the regressor variable. The intercept and slope of the straight lines were estimated for all treatment groups. We designed the study to provide an 80% power to detect a difference in changes from baseline BP of at least 4 mm Hg between the combination group and any active treatment group with an SD of 9 and a significance level of .05. Thus, a total of 63 patients per treatment group was the target.
Patient Characteristics and Accounting
In 18 centers, a total of 273 white patients were recruited into the double-blind phase of the study, ie, 67 to 69 patients per treatment group (Table 1⇓). On entry into the double-blind part of the study, the four treatment groups were well balanced with regard to age, sex, height, weight, smoking habits, and duration of hypertension and pretreatment. At baseline, BP was essentially the same in all groups, with mean SBP ranging from 158.1 to 159.9 mm Hg and mean DBP from 103.3 to 104.0 mm Hg.
During the study, 23 patients (8%) discontinued the study prematurely. Of those, 3 were withdrawn from the study because they did not cooperate, 17 because of clinical or laboratory adverse events, 1 because of ineffective therapy, and 2 because of other reasons.
The degree of patient compliance was calculated as 100×(Capsules Distributed−Capsules Returned)/Days Since Last Visit. Compliance was not less than 95.4% in any study group at any visit.
The efficacy analysis was performed in an “intent-to-treat” manner in all 273 randomized patients at baseline, 259 patients after week 6, and 253 patients after week 12. Fig 2⇓ summarizes the mean change from baseline in trough SBP and DBP per week in the various treatment groups. It is evident that most of the fall in BP occurred by week 3; from week 3 to 6, only a moderate additional decrease was observed. The hypotensive effects from baseline were highly significant (P<.001) for all drugs.
The fall in DBP was similar in the clopamid, reserpine, and nitrendipine groups (range, −11.7 to −12.3 mm Hg), but it was considerably greater in the reserpine-clopamid combination group (−17.1 mm Hg). This difference was highly significant when the combination was compared with each of the monotherapy groups (P<.001) and the nitrendipine group (2P=.002). The stronger antihypertensive effect of the reserpine-clopamid combination was also reflected in the higher normalization rate (55%), defined as trough DBP less than 90 mm Hg after week 6 compared with the other treatments (33% to 40%, 2P=.11). Doubling of the respective medication in the non-normalizers from weeks 7 to 12 yielded a further small BP decrease in all treatment groups, finally resulting in normalization rates of 69% for the combination, 35% for reserpine monotherapy, 39% for clopamid monotherapy, and 45% for nitrendipine (2P<.0001) (Table 2⇓).
For SBP, the decrease after week 6 in the reserpine-clopamid combination group was clinically and statistically significantly stronger (−23.0 mm Hg) than in the other treatment groups (−11.6 mm Hg for nitrendipine to −14.0 for reserpine; P<.001). Dose titration in non-normalizers resulted in a modest further SBP decrease in all groups except for the reserpine monotherapy group.
Synergistic Effect of the Components
A linear regression formula provides a raw approximation of whether the components act in an additive manner. The straight regression lines for changes from baseline in mean arterial BP are plotted in Fig 3⇓; estimates and slopes are given in Table 3⇓. The slope of the combination group was steeper than the slopes of the monotherapy groups, implying that the most pronounced hypotensive effect will be observed in this treatment group. For instance, a mean arterial BP of 145 mm Hg at baseline will be lowered by 32.6 mm Hg during reserpine-clopamid combination therapy, in contrast to 19.6 and 18.0 mm Hg on reserpine and clopamid monotherapy, respectively. According to this model, a total decrease of 37.6 mm Hg would be predicted if the effect of both components were additive in a synchronized manner. Yet this approach does not take into account the well-known interdependence of baseline BP value and the magnitude of the observed hypotensive effect. Therefore, the baseline of the component with a later onset of action has to be adjusted for the already lowered pressure based on the mathematical estimation. In the theoretical case of reserpine being the fast-acting component, an additional clopamid effect of −11.8 mm Hg would be expected; otherwise, the additional effect of reserpine would be −12.5 mm Hg. Addition of the full effect of the rapid component to the additional effect of the slower component results in a total BP-lowering effect of 30.5 to 31.4 mm Hg, which is less than the clinical effect of the combination (−32.6 mm Hg). Thus, it appears that the components act slightly more than additively in combination.
Mean baseline heart rates were 74.9, 75.6, 75.2, and 73.8 beats per minute for the reserpine-clopamid combination, reserpine, clopamid, and nitrendipine groups, respectively. Despite the above changes in BP, heart rates remained constant in the clopamid and nitrendipine groups and fell in the reserpine and reserpine-clopamid groups by 5.6 and 5.3 beats per minute, respectively, by the end of the study.
Clinical and Laboratory Safety
We determined the safety of the different medications by evaluating the incidence of clinical and laboratory adverse experiences and mean changes in body weight, electrocardiographic parameters, and laboratory test results. The percentages of patients with one or more adverse experiences during the course of the study were almost the same in the combination group and reserpine and clopamid monotherapy groups (range, 27% to 29%). Nitrendipine was tolerated to a lesser extent than the other study medications; the rate of adverse experiences (48%) in this group was significantly higher (P=.01) (Table 4⇓). Adverse events with nitrendipine were classified as “severe” in 20% as opposed to 5% to 8% in the other treatment groups.
Table 5⇓ gives the most common clinical adverse experiences during the study, with a patient incidence of 3% or higher in any one treatment group. Since the study was conducted during autumn and winter, some 5% of the patients suffered from upper respiratory tract infections such as the common cold and bronchitis. In the reserpine-clopamid group, 3 patients (4%) expressed complaints concerning fatigue, whereas other clinical adverse experiences occurred only once. In the nitrendipine group, flush was experienced by 10 patients (15%), headache by 7 (10%), and peripheral edema by 4 (6%). The numbers of patients who withdrew from the study because of adverse experiences were 2 (3%) each in the reserpine-clopamid (fatigue and hypokalemia, respectively) and reserpine groups, 5 in the clopamid group, and 9 in the nitrendipine group (Table 4⇑). Two adverse events were considered to be serious. In the clopamid group, uterine bleeding after 6 weeks of treatment was observed and determined to be possibly drug related. The other serious adverse event was a tentative diagnosis of gastrointestinal bleeding with tarry stool in the nitrendipine group after 12 weeks of therapy; this event was also rated as possibly drug related. Both patients were discontinued from the study medication.
Table 6⇓ shows the laboratory data for the treatment groups. In the reserpine-clopamid group, metabolic changes were small: mean total cholesterol remained unchanged; triglycerides increased by 5.6% and glucose by 3.4%. Uric acid increased moderately by 10.9%. In seven cases, this increase in uric acid was considered a laboratory adverse experience. Regarding potassium, in the combination group, the mean decrease of −0.2 mEq/L was smaller than in the diuretic group (−0.4 mEq/L) and slightly higher than in the nitrendipine group (−0.1 mEq/L). Two patients were withdrawn from the study because of hypokalemia, one in the clopamid group and one in the reserpine-clopamid group.
Body weight and electrocardiographic measurements did not change significantly from baseline after 12 weeks of treatment in any of the groups.
The percentage of patients whose DBP was normalized and who remained free of any adverse event was 49% in the reserpine-clopamid group, 19% in the reserpine group, 20% in the clopamid group, and 12% in the nitrendipine group (2P<.0001).
The rationale for using free or fixed combinations in drug therapy has been extensively discussed,2 3 20 21 but clinical studies generally have investigated full-dose combinations or combinations in an add-on or second-line approach. To our knowledge, only one study has compared a combination of low-dose components with usual doses of monotherapy in a first-line approach: A combination of the β-blocker bisoprolol and the diuretic hydrochlorothiazide was reported to be as effective as full-dose amlodipine and more effective than enalapril, with combination therapy being associated with a lower rate of adverse experiences.21
For our investigation, we chose reserpine and the thiazide clopamid as prototypes of the low-dose combination approach. For the first time, the magnitude of the antihypertensive effect that both components contributed to the overall effect was determined. Furthermore, this was the first direct comparison of the “old” reserpine diuretic combination with a “modern” calcium antagonist.
This study was not placebo controlled during the active treatment phase because a placebo arm is not necessary for assessment of the contribution of each single component to the overall BP-lowering effect. Furthermore, this active-controlled concept is the one of choice22 for the other medical question addressed, namely, how the low-dose combination treatment compares with full-dose nitrendipine monotherapy. The relative efficacy and tolerability of the different treatments are of interest rather than the net effect of drug therapy.
In contrast to many other studies, the present trial took advantage of a semiautomatic BP device that included printouts of measurements with the date and time of data collection. Observer bias23 and digit preference in BP studies can considerably influence the outcome. Automatic devices ensure standardization and verification of the measurements and therefore substantially increase the validity of the results.24
In this large trial, the decreases in SBP and DBP produced by the low-dose reserpine-clopamid combination were significantly greater than those induced by either reserpine or clopamid alone. With rising baseline BP values, the efficacy of the combination increases to a greater extent than is the case with the other treatments. Since the two drugs act synergistically to achieve a better therapeutic response than either drug alone, the primary requirement for any fixed combination25 is fulfilled.
More important than the average BP-lowering effect of a drug in a study population is the question of how many patients actually achieved the primary goal of antihypertensive treatment, that is, normalization of elevated BP. The percentage of responders in the reserpine-clopamid group was substantially higher than in all other monotherapy groups. The responder rates for the low-dose combination were in the range of what can be expected for combination therapy19 ; however, the rates for the monotherapy groups seemed to be somewhat low compared with rates in recent studies26 27 that compared the efficacy of different monotherapies. In an investigation restricted to men, Materson et al26 reported normalization rates from 75% to 54% for sustained-release diltiazem, clonidine, atenolol, hydrochlorothiazide, prazosin, or captopril. However, in the present study, mean baseline BP at entry was rather high, as was the percentage of patients whose BP had not been normalized during previous therapy (approximately 60%). Moreover, in the reserpine and clopamid monotherapy groups, relatively low doses even after the titration step were used. Thus, this protocol did not explore the upper limits of the reserpine dose range (nowadays 0.3 mg) and clopamid dose range (20 mg) or alternating dosing intervals, such as twice daily, which might have resulted in higher normalization rates. On the other hand, the BP reduction in these groups was comparable to the effect achieved in the nitrendipine group, which was started at a common dose (20 mg) and titrated to a high dose (40 mg). In that respect, all monotherapies were comparable in their efficacy, the low-dose combination being superior.
The rate of adverse events in all groups appeared to be rather high; this is because of the use of standardized questions asked during each visit. Between groups, the incidences of adverse experiences and early withdrawals were lowest with the reserpine-clopamid combination. Usually, certain adverse effects are attributed to reserpine therapy; one is depression, typically noted in studies of the 1950s,28 29 in which the reserpine doses used were as much as 5 to 40 times higher than those currently recognized as appropriate. In both the reserpine monotherapy and combination groups, depression was not observed, consistent with the low incidence rate reported in other studies30 31 and pharmacovigilance surveys.32 Adverse metabolic effects commonly associated with diuretic therapy, such as increases in glucose or lipid abnormalities, were not a clinical problem with the combination and all monotherapy groups. Nonetheless, the possibility of hypokalemia cannot be ruled out in a diuretic-containing preparation; in one instance in the combination group, hypokalemia led to a withdrawal from the study. Interestingly, reserpine in the present study appeared to counterbalance the lipid changes induced by the diuretic component. A similar observation has also been described in other studies.12 30 Whether this effect is due to or independent of reserpine has not been studied.
The relatively high incidence of adverse events in the nitrendipine group is consistent with typical findings of other studies: A meta-analysis of the safety data of 17 450 individuals revealed that most adverse events, such as hot flushes (12%), headache (10%), and peripheral edema (5%), are extensions of the pharmacological action of the drug.33
Reserpine in combination with a diuretic belongs to those antihypertensive therapies that have proved to reduce stroke and coronary disease in hypertensive patients. Other advantages of reserpine include once-daily dosing, an absence of rebound effect if doses are missed, and when used with a diuretic, a reduction of left ventricular hypertrophy.34 It is also the least expensive agent, costing about as much for a 1-year supply as many newer agents cost for a 1-week supply.35 The decline in the use of reserpine and thiazides in the last several years36 may be due to the fact that newer drugs, such as calcium antagonists, seem to be more attractive to investigate than the older ones, and studies using them are more likely to be funded by industry.35
The results of the present study indicate that reserpine in low-dose combination with a thiazide remains a drug of choice for the treatment of hypertension. The components act synergistically and display a strong antihypertensive effect, resulting in high normalization rates. The present direct comparison with nitrendipine as a representative of the modern group of calcium antagonists reveals an impressive advantage of reserpine-clopamid in terms of efficacy and tolerability. Low-dose combinations of reserpine with thiazides, such as that investigated, appear to be a rational treatment for patients with mild to moderate hypertension.
We thank the investigators of the Briserin in Hypertension Study Group for their clinical research support. We are also indebted to statistician H.J. Kaiser, PhD, for his excellent assistance with the statistical work.
Reprint requests to Dr D.B. Pittrow, Sandoz AG, Department of Clinical Research, Deutschherrnstraße 15, D-90429 Nürnberg, FRG. E-mail email@example.com
- Received March 26, 1996.
- Revision received April 18, 1996.
- Revision received August 27, 1996.
Fingl E, Woodbury DM. General principles. In: Goodman LS, Gilman A, eds. The Pharmacological Basis of Therapeutics. 4th ed. New York, NY: Macmillan Publishing Co; 1970:24-25.
Materson BJ, Cushman WC, Goldstein G, Reda DJ, Freis ED, Ramirez EA, Talmers FN, White TJ, Nunn S, Chapman RH, Khatri I, Schnaper H, Thomas JR, Henderson WG, Fye C. Treatment of hypertension in the elderly, I: blood pressure and clinical changes: results of a Department of Veterans Affairs Cooperative Study. Hypertension. 1990;15:348-360.
McMahon FG, Brobyn R, Kann J, Levy B, Margolis R, Reeves R. A double-blind comparative study of nitrendipine and propranolol in the treatment of hypertension. J Cardiovasc Pharmacol. 1987;9(suppl 4):228-231.
Association for the Advancement of Medical Instrumentation. American National Standard: Electronic or Automated Sphygmomanometers. Arlington, Va: Association for the Advancement of Medical Instrumentation, 1987.
Reid JL. First-line and combination treatment for hypertension. Am J Med. 1989;86(suppl 4A):2-5.
Henry D. Comparing treatments. BMJ. 1995;310:1279.
Holzgreve H, Distler A, Michaelis J, Philipp T, Wellek S. Verapamil versus hydrochlorothiazide in the treatment of hypertension: results of a long term double blind comparative trial. BMJ. 1989;229:881-886.
Materson BJ, Reda DJ, Cushman WC, Massie BM, Freis ED, Kochar MS, Hamburger RJ, Fye C, Lakshman R, Gottdiener J, Ramirez EA, Henderson WG, for the Department of Veterans Affairs Cooperative Study Group on Antihypertensive Agents. Single-drug therapy for hypertension in men: a comparison of six antihypertensive agents with placebo. N Engl J Med. 1993;328:914-921.
Freis ED. Mental depression in hypertensive patients treated for long periods with large doses of reserpine. N Engl J Med. 1954;251:1006-1008.
Muller JC, Pryor WW, Gibbons JE, Orgain ES. Depression and anxiety occurring during rauwolfia therapy. JAMA. 1955;159:836-839.
Goldstein G, Materson BJ, Cushman WC, Reda DJ, Freis ED, Ramirez EA, Talmers FN, White TJ, Nunn S, Chapman RH, Khatri I, Schnaper H, Thomas JR, Henderson WG, Fye C. Treatment of hypertension in the elderly, II: cognitive and behavioral function—results of a Department of Veterans Affairs Cooperative Study. Hypertension. 1990;15:361-369.
Annual Report on Adverse Drug Events 1993. Basel, Switzerland: Drug Monitoring Center, Sandoz Ltd; 1993.
Bayotensin product leaflet. Bayropharm GmbH Internal data 1992. Wuppertal, Germany.
Hypertension Detection and Follow-Up Program Cooperative Group. Five-year findings of the Hypertension Detection and Follow-Up Program: prevention and reversal of left ventricular hypertrophy with antihypertensive drug therapy. Hypertension. 1985;7:105-112.