Original article 1327 Main results of the losartan versus amiodipine (LOA) study on drug tolerability and psychological general well-being Björn Dahlöf, Lars H. Lindholm*, Shane Carney †, Pertti J. Pentikäinen‡ and Jan Östergren § for the LOA Study group Objective To compare two losartan regimens (with and without hydrochlorothiazide) and amiodipine in treating mild-to-moderate hypertension regarding their bloodpressure-lowering effect, drug tolerability and quality of life. Design A 12-week, randomized, double-blind, parallelgroup, multi-centre study. After 4 weeks of placebo, patients with a diastolic blood pressure (DBP) in the range 95-115 mmHg were allocated randomly to be administered 50 mg losartan (increased to 100 mg if the DBP was 90 mmHg or more after 6 weeks), 50 mg losartan (plus 12.5 mg hydrochlorothiazide under the above conditions), or 5 mg amiodipine (increased to 10 mg under the above condition). The tolerability of the treatment and the quality of life were evaluated by spontaneous reporting, active questioning and the Psychological General Well-Being (PGWB) index. Study population In total 898 hypertensives, mainly referred from primary health care (mean age 57.8 years) of whom 52% were men. Results Administration of 50 mg losartan (plus 1 2.5 hydrochlorothiazide if necessary) and of 5 mg amiodipine (or 10 mg if necessary) lowered the blood pressure as well as or better than did 50 mg losartan (or 100 mg if necessary). The incidence of any discomfort' and 'swollen ankles' increased with amiodipine but not with losartan treatment. The opposite was found for 'dizziness upon standing'. The incidence of drug-related adverse events and the number of patients withdrawn from therapy were higher with amiodipine than they were with losartan treatment. The PGWB index at week 12 indicated that improvements from baseline had occurred in some domains for the losartan groups whereas it remained unchanged for the amlodipine group. Conclusion Both losartan and amiodipine were effective in lowering the blood pressure and were tolerated well. Administration of 50 mg losartan (plus 12.5 mg hydrochlorothiazide if necessary) and of 5 mg amlodipine (or 10 mg if necessary) lowered the blood pressure equally well or better than did 50 mg Iosartan (or 100 mg if necessary). Drug-related adverse effects and withdrawal from the study were more common for the amlodipine group. The clinical significance of the improvements in the PGWB index with losartan needs to be studied further. Journal of Hypertension 1997, 15:13 27-1335 Keywords: amlodipine, general well-being, hypertension, losartan, tolerability, quality of life From the Clinical Trial Unit, Department of Medicine, Sahlgrenska University Hospital/Östra, Göteborg, Sweden, the *Department of Community Health Sciences, Lund University, Lund, Sweden, the †Division of Medicine, John Hunter Hospital, University of Newcastle, Australia, the ‡ Department of Medicine, Helsinki University. Helsinki, Finland, and the §Department of Medicine, Karolinska Hospital, Stockholm, Sweden. Sponsorship: Financial support was provided by Merck & Co. Inc. Requests for reprints to Dr Lars H. Lindholm, Department of Community Health Sciences, Lund University, Helgeandsgatan 16, S-223 54 Lund, Sweden. Received 20 February 1997 Revised 1 July 1997 Accepted 2 July 1997 © Rapid Science Publishers ISSN 0263-6352 Introduction Pharmacological treatment of arterial hypertension is one area in medicine that has shown remarkable progress and development during recent decades. This-is true not only with regard to the number of efficacious and generally well-tolerated antihypertensive drugs that have become available but also to the substantial documentation of a clinically relevant reduction in cardiovascular morbidity and mortality brought about by antihypertensive drug treatment . We now have a choice of effective blood-pressure-lowering agents with better and better drug tolerability. β-Blockers and diuretics are the drug classes so far shown to reduce cardiovascular morbidity and mortality risk [2,3] but have at the same time been linked to a number of well-known side effects [1,4]. 'Newer' treatment regimens have been shown to be tolerated well in general by hypertensive patients but have also been associated with specific side effects [e.g. calcium antagonists with ankle oedema [1,5,6], angiotensin converting enzyme (ACE) inhibitors with coughing [7,8] and -blockers with orthostatic hypotension ]. Members of one of the newest antihypertensive classes of drugs avail- 1328 Journal of Hypertension 1997, Vol 15 No 11 able, namely the selective angiotensin II antagonists, have shown promising results in terms of their improved drug tolerability, few drug-specific side effects and satisfactory blood-pressure-lowering effects [9-11]. Table 1 Needless to say, the ultimate goal for the treatment of high blood pressure is to prevent or delay cardiovascular disease. This has not yet been established to be achieved in hypertensives for most of the 'newer' blood-pressurelowering drugs (α-blockers, ACE inhibitors, calcium antagonists and selective angiotensin II antagonists), however. One prerequisite for a satisfactory effect on the blood pressure is that the patient be compliant and ingest his or her medication as prescribed. A large follow-up survey in the United Kingdom showed that less than 50% of the patients were continuing their initial antihypertensive therapy after 6 months . There are of course many reasons for a patient to discontinue drug therapy for a chronic disease. A lack of effect is one, drug-related side effects and an impaired quality of life are others. It is believed generally that the 'newer' drugs, having fewer side effects, maintain general well-being better than do blockers and diuretics . This is likely to lead to better patient compliance. Men (%) 53 53 52 Age (years) Aged < 65 years (%) Aged 65-74 years (%) Aged > 74 years (%) Caucasians (%) Severity of hypertension Mild (DBF 95-105 mmHg) (%) Moderate (DBF 106-115 mmHg) (%) Laboratory values Blood glucose (mmol/l) Serum cholesterol (mmol/l) Serum triglycerides (mmol/l) 57.8 71 24 5 98 57.9 57.8 Large prospective studies concerning cardiovascular morbidity and mortality are under way comparing calcium antagonists, ACE inhibitors and angiotensin II antagonists with β-blockers and diuretics . While awaiting the outcome of these trials, we could and probably should evaluate the effects on intermediate (surrogate) endpoints of the 'newer' antihypertensive drugs as well as their impact on tolerability and general well-being. For many years, drug tolerability was evaluated only with structured interviews on specific symptoms (e.g. tiredness, ankle oedema and coughing). Studies on the quality of life and well-being really began to evolve with the publication of a comparison of captopril, propranolol and methyldopa by Croog e.t al.  in 1986 using, among other things, a psychological general well-being (PGWB) index. Compositions of the study groups at baseline A Losartan monotherapy (n = 300) B Losartan plus HCTZ (n = 300) C Amlodipine monotherapy (n = 298) 70 22 8 100 72 23 5 100 68 72 74 29 26 24 5.5 6.1 1.7 5.4 6.1 5.6 6.1 1.6 1.7 HCTZ, hydrochlorothiazide; DBF, diastolic blood pressure. At baseline, the groups did not differ regarding age, sex distribution, heart rate (73-74 beats/min), and prevalences of other diseases such as cardiovascular disease (16-23%), diabetes mellitus (6-7%), musceloskeletal disease (39-41%), neurological and psychiatric disorders including migraines and headaches (25-26%) and respiratory diseases (20-22%). Blood pressure values at baseline are given in Table 2 and the severity of hypertension and blood lipid levels in Table 1. The vast majority of the patients was recruited from primary health care. The six key exclusion criteria (of 22 applied) were that we excluded women of childbearing age, those with significant renal impairment, those who had suffered myocardial infarction within the last 6 months, those who had angina pectoris, those who had congestive heart failure, those who were being administered β-blockers and oth antihypertensive drugs and those who had previous., been treated with angiotensin II antagonists and calcium antagonists. Study design The overall aim of the present study was to compare two losartan treatments [with and without hydrochlorothiazide (HCTZ)] and amlodipine in treating mild-to-moderate hypertension concerning their blood-pressure-lowering effect, drug tolerability and the quality of life using the PGWB index. Study population and methods In total, 898 patients with mild-to-moderate hypertension were included in the study; 637 (71%) came from Sweden, 150 (17%) from Australia and 137 (12%) from Finland. Their baseline characteristics are given in Table 1; 52% were men. The men were aged 56.3 ± 1 1 . 4 years (mean ± SD) and the women 59.5 ± 9.9 years; almost all were Caucasians. The study was a 12-week, multi-centre (63 centres), randomized, double-blind, double-dummy, parallel-group trial. After 4 weeks of placebo, patients with a diastolic blood pressure in the range 95-115 mmHg when they were seated were allocated randomly to be administered either 50 mg losartan (groups A and B) or 5 mg amlodipine (group C) in a ratio of 2 : 1. If the diastolic blood pressure remained above 90 mmHg after 6 weeks of therapy, 100 mg losartan was administered to subjects in group A, 50 mg losartan plus 12.5 mg HCTZ to subjects in group B and 10 mg amlodipine to subjects in group C for the remaining 6 weeks of the study. All of the drugs were administered once a day, in the morning. Patients who had been being administered blood-pressure-lowering drug treatment at the time of their recruitment went without antihypertensive drugs for a minimum of 7 days before starting the placebo period. Quality of life during losartan treatment Dahlöf et al. 1329 Table 2 Systolic blood pressures (SBP) and diastolic blood pressure (DBP) of sitting patients at baseline and during week 12 Treatment groups Baseline (mmHg) Week 1 2 (mmHg) Mean change from baseline (mmHg) 95% Confidence interval (mmHg) Significance (P) 161.9 ±16.4 148.3 ± 17.7 -13.6 -1 5.0 to -1 1 .9 < 0.001 160.5 ±17.0 143.4 ± 16.6 -17.1 -18.5 to -15.7 < 0.001 160.3 ± 15.8 143.5±13.8 -16.8 -18.3 to -15.2 < 0.001 102.6 ± 5.9 92.3 ± 8.2 -10.3 -11.1 to -9.4 < 0.001 101.8 ± 5.6 90.4 ± 7.2 -11.4 -12.2 to -10.6 < 0.001 101.8 ± 5.6 89.3±17.1 -12.5 -13.3 to -11.8 < 0.001 SBP Losartan monotherapy (A) (n = 298) Losartan plus HCTZ (B) (n = 300) Amlodipine monotherapy (C) (n = 298) DBP Losartan monotherapy (D) (n = 298) Losartan plus HCTZ (E) (n = 300) Amlodipine monotherapy (F) (n = 298) Values are expressed as means±SD. HCTZ, hydrochlorothiazide. Between-group comparisons: A versus B, P<0.001; A versus C, P=0.003; B versus C, NS; D versus E, P= 0.022; D versus F, P< 0.001; E versus F, NS. Patients were examined during clinical visits during weeks -4 (the start of the placebo run-in period), 0 (random allocation to treatment), 3, 6, 9 and 12. A routine clinical examination was performed during all of the visits. Blood chemistry determinations and urine analyses were performed during weeks -4, 0, 6 and 12 (data not given). The blood pressure (the mean of two recordings) and heart rate were recorded at trough (i.e. 24 h after the latest morning dose). The blood pressure was recorded after the subject had sat for at least 5 min and after the patient had stood for 2 min. Cuffs were of appropriate size and the blood pressure was measured with the arm on which the measurement took place supported at heart level. Adverse events and drug tolerability were monitored during all visits. The tolerability was evaluated by use of a structured interview consisting of one general question to be answered by 'yes' or 'no' ('Have you felt any kind of discomfort since your last visit?') and 24 questions on specific symptoms to be answered in the same way. The structured interview had previously been evaluated [16). Moreover, during each visit the number of adverse events considered drug-related ones by the investigator was recorded, as was the number of patients who had discontinued their participation in the study because of adverse events. During each visit, the global symptom score was computed by giving a score equal to zero when a symptom was absent and a score equal to unity when a symptom was present and then summing the scores obtained for each of the 24 symptoms. If a patient did not answer three or more questions, the global symptom score for that patient was set to 'missing'. The PGWB index was used to evaluate changes in quality of life [17,18]. The patients filled in this questionnaire at home on the day before visits during weeks -4, 0, 6 and 12 and brought the questionnaire to the investigator in a sealed envelope. The PGWB index comprises 22 questions (range 22-132 points) divided into six domains (anxiety, depressed mood, positive well-being, selfcontrol, general health and vitality). The PGWB index has been used in several previous studies and, in contrast to most other scales, includes both positive and negative aspects. It is not intended to discriminate between psychiatric cases and healthy subjects . All of the patients had given their informed consent to participate in the study and the study was approved by research ethics committees of the University Hospitals of Goteborg, Helsinki and Newcastle (Australia). Statistical analysis To address the tolerability objective further, the area under the curve (AUC0-12 weeks) was calculated for each patient during the 12 week treatment period using the trapezoidal method. The AUC0-12weeks measured was standardized for time of follow-up by dividing by the number of days that the patient was monitored. It was calculated as the area between the curve (change from baseline in global symptom score versus time) and the abscissa (time axis). The area below the abscissa (time axis) was considered negative and the area above was considered positive. The efficacy and tolerability analyses were performed using an intention-to-treat approach, including all patients for whom we had efficacy and tolerability data both at baseline and for during treatment; additional analyses based on a per-protocol approach were also performed but gave no further insight (data not given). The comparability of the three treatment groups with respect to patient characteristics was assessed by use of the x2 test for the dichotomous variables and the Kruskal-Wallis test for the ordered categorical variables. Baseline efficacy variables (the global symptom score and the blood pressure of seated subjects) were analysed by analysis of variance 1330 Journal of Hypertension 1 997, Vol 1 5 No 11 including the treatment and investigators as factors. Within-group tests of the change from baseline in blood pressure, symptom score and effect on the quality of life were based on a two-sided Student's t test. The betweengroup comparisons were performed in two steps. A closed testing procedure of Bauer  was used to address-the multiplicity issue.''When the overall treatment effect was significant at the level P<0.05 according to analysis of "arian.ce, three pairwise treatment comparisons were performed on the basis of a least-squares analysis of the means. The differences between treatment groups were estimated in a pair-wise fashion in terms of the difference in least squares and its associated 95% confidence interval (CI). Within-group changes in the incidence of patients suffering any kind of discomfort (from- the general questioning) and in the incidence of patients with each symptom separately were assessed by the McNemar test. Between-treatment comparisons were performed in two Steps, as described above. Pairwise treatment comparisons were carried out by means of a logarithmic-linear model. The incidence of patients having drug-related adverse experiences and patients discontinuing the study was compared • between groups in a pairwise fashion using Fisher's exact test. For the global symptom score, with a sample size of 898 patients; the study had a 90% power of detecting a between-treatment difference of 0.59 units in the area under the curve, if such a difference existed. Results Blood pressure and heart rate The blood pressure (of sitting patients) was lowered significantly in members of all of the groups. After 6 weeks (monotherapy) amlodipine treatment had reduced the blood pressure by 13.8 ± 12.2/9.9 ± 6.4 mmHg (mean ± SD) in comparison with 12.7 ± 12.1/8.5 ± 6.6 mmHg for losartan treatment; in a pairwise comparison the difference between the groups was 0.9/1.3 mmHg (95% CI 0.7 to 2.6/0.4-2.2). After 6 weeks, 33% of the patients in the losartan group and 40% of those being administered amlodipine had a diastolic blood pressure below 90 mmHg (P = 0.012). The lowering of the blood pressure attained during week 12 is shown in Table 2. The results observed for the blood pressure of standing patients were comparable to those obtained for the blood pressure of seated patients (data not given). After 6 weeks, 61% of the, patients in the 50 mg losartan group (group A) had been transferred to the higher dose of 100 mg. The corresponding perce ages in the 50 mg losartan and 12.5 mg HCTZ (group bj and amlodipine (group C) groups were 62 and 56%, respectively (NS). Administration of 50 mg losartan (plus 12.5 mg HCTZ if necessary) and of 5 mg amlodipine (or 10 mg if necessary) reduced the blood pressure equally well. Administration of 50 mg losartan (or 100 mg if necessary), however, was not so effective as were the other two treatments (Table 2). This is also shown in Figure 1, in which the percentages of patients who responded to treatment are given for the three treatment groups. The heart rate did not change significantly for any group (data not given). Subgroup analyses by country, sex, age group, and degree of hypertension gave no further insight (data not given). Symptoms After 12 weeks 'any discomfort' was reported by 33.1% of the patients in the amlodipine group in comparison with 23.3% at baseline (P = 0.002). The correspondin- Fig. 1 Categories (l-lll) of response to antihypertensive treatment during week 12 for the diastolic blood pressures of seated patients. HCTZ, hydrochlorothiazide; I, diastolic blood pressure <90 mmHg; II, diastolic blood pressure 90 mmHg and reduced by 10 mmHg; III, neither I nor II. Quality of life during losartan treatment Dahlöf et al. 1331 Fig. 2 Changes in incidence of individual symptoms and 95% confidence intervals during week 1 2 for (a) 50 or 100 mg losartan versus 5 or 10 mg amlodipine and (b) 50 mg losartan plus 12.5mg hydrochlorothiazide if necessary versus 5 or 10mg amlodipine. percentages for losartan monotherapy were 22.5 and 23.2% (NS) and those for losartan plus HCTZ therapy were 23.5 and 19.8% (NS). The symptom 'swollen ankles' was reported by 30.6% of the patients in the amlodipine group in comparison with 8.4% at baseline (P < 0.001). The corresponding percentages for losartan monotherapy were 7.0 and 7.0% (NS) and those for losartan plus HCTZ therapy were 8.4% after treatment and 10.0% at baseline (NS). The symptom 'swollen ankles' was reported by 15.2% of the patients in the amlodipine group already after 6 weeks, which was a significant increase from baseline (P = 0.003). 'Dizziness upon standing' was more prevalent during week 12 than it was at baseline among patients in both losartan groups; 10.1 versus 6.0% for the losartan-alone group. (P= 0.028) and 17.1 versus 9.0% for the losartan plus HGTZ group (P = 0.001). The symptom 'dizziness upon standing' was reported by 10.2% of patients in the losartan group(s) as early as after 6 weeks, which was a significant increase from baseline (P = 0.027). When the 53 patients in the losartan group who reported 'dizziness upon standing' during week 12 were analysed further we could not find any difference in lowering of the blood pressure or in heart rate, with the patients sitting and standing, compared with those of other patients. We also considered age, sex and the percentage increased doses, but this gave no further insight (data not given). There was no increase in prevalence of the symptom 'dizziness upon standing' among patients in the amlodipine group during the study. In Figure 2(a) the changes in incidence of individual symptoms and 95% CI during week 12 are given for 50 or 100 mg losartan versus 5 or 10 mg amlodipine; in Figure 2(b), the corresponding values are given for 50 mg losartan plus 12.5 mg hydrochlorothiazide if necessary versus 5 or 10 mg amlodipine. A 'global score' based on the 24 symptoms, integrating data from all of the visits, revealed no differences between the three treatment groups (Fig. 3). Finally, there were more patients with drug-related adverse events in the amlodipine group and more patients withdrew from that group because of adverse events than occurred for either of the losartan groups (Table 3). Quality of life The evaluation of the quality of life in terms of the PGWB score revealed improvements for patients in the losartan groups after 12 weeks in comparison with baseline for 1332 Journal of Hypertension 1997, Vol 15 No 11 Fig. 3 Tables Patients whose Psychological General Well-Being Index score improved, did not change and worsened in total score by week 12 compared with baseline Treatment groups Losartan monotherapy (A) (n = 298) Losartan plus HCTZ (B) (n = 300) Amlodipine monotherapy (C) (n = 298) Improved (%) 60 No change (%) Worsened (%) 8 6 33 6 54 39 50 44 HCTZ, hydrochlorothiazide. Between-group comparisons; A versus B, NS; A versus C, P= 0.011; B versus C, NS. Fig. 4 Evaluation of tolerability in terms of the global symptom score (0—24) and treatment group (mean changes from bassline): — losartan monotherapy; - • -, losartan plus hydrochlorothiazide if necessary; ..... , amlodipine monotherapy. Table 3 Drug-related adverse experiences and patients withdrawn from the study due to adverse experiences A B C Losartan Completed the study (%) Drug-related adverse experiences (°/o) Withdrawn due to adverse experiencea (%) monothera Losartan plus HCTZ py (n = 300) (n=300) 94 92 Amlodipine monotberapy (n = 298) (n = 298) 89 22 24 3 6* * † † † 2 5 8*† † HCTZ, hydrochlorothiazide. *P<0.06, **P = 0.01, versus B; P= 0.001, versus A. †† The total Psychological General Weil-Being (PGWB) index score plotted for the visits (weeks -4, 0, 6 and 1 2) and treatment groups: —-, losartan monotherapy (n = 267); ------ , losartan plus hydrochlorothiazide it necessary; ..... , amlodipine monotherapy (n = 259); only data for patients for whom we had a total score value for each visit are shown. P=0.01, Table 4 Evaluation of quality of life with the Psychological General Well-Being index (mean change from baseline by week 12) Domain Anxiety Depressed mood Positive well-being Self-control General health Vitality Total score A Losartan monotherapy 0.76 (P< 0.001) 0.25 (P= 0.042) 0.72 ( P < 0 .0 0 1 ) 0.14 (NS) 0.13 (NS) 0.47 (P = 0.01) 2.44 (P< 0.001) B Losartan plus HCTZ 0.38 C Amlodipine monotherapy 0.33 (P= 0.044) -0,04 (NS) 0.27 (NS) 0.04 (NS) 0.35 (P = 0.004) 0.42 (P= 0.035) 1.65 (P =0. 01 5) (NS) -0.09 (NS) 0,05 (NS) 0.14 (NS) -0.20 (NS) 0.23 (NS) 0.5 (NS) HCTZ, hydrochlorothiazide. Pair-wise between-group comparisons: for positive well-being; overall between-group, P = 0.015; A versus B, P= 0.057; A versus C, P = 0.005; B versus C, NS; and for general health, overall between-group, P= 0.009; A versus B, NS; A versus C, P= 0.097; B versus C, P= 0.002. several domains (Table 4), whereas the quality of life remained unchanged for all domains for patients in the amlodipine group. The total score after 12 weeks was 110.0 versus 107.5 at baseline for the losartan monotherapy group (P< 0.001) and 109.8 versus 108.1 at baseline (P = 0.002) for the losartan plus HCTZ group. The corresponding values for the amlodipine group were 108.7 and 108.2, respectively. The proportions of patients whose total score improved, did not change, and worsened are given in Table 5; for the three treatment groups, improvements were seen for 60, 54 and 50% of the patients, respectively. The changes in total score during the study (visits during weeks -4, 0, 6 and 12) are shown in Figure 4. Only data for patients for whom we had a total score value for each visit were included. The data correspond to 267 patients for losartan monotherapy, 261 for losartan plus HCTZ if necessary and 259 for amlodipine monotherapy; if data were missing for weeks 0 and 12, data from weeks -4 and 6, respectively were carried forwards. The slight improvement in total score from the visit during week -4 to the visit during week 0 (during placebo) was statistically significant (mean change 0.8, P= 0.037). Quality of life during losartan treatment Dahlöf et al. After 6 weeks, the change in total score from baseline was -0.14 (95% CI -1.07 to 0.79) for losartan and 0.65 (-0.73 to 2.03) for amlpdipine; further subgroup analyses have not been carried out. The Pvalue for the between-group comparison was 0.380. Discussion This study, which was intended to evaluate the differences among the effects of two losartan regimens (with and without hydrochlorothiazide) and amlodipine therapy in treating mild-to-moderate hypertension in terms of their blood-pressure-lowering effects, tolerability and effects on the quality of life, measured in terms of PGWB, could be of interest for several reasons. First, it was a -fairly large study comparing a new drug, namely an angiotensin II antagonist, with a well-tolerated, long-acting dihydropyridine, of established properties. Smaller studies have shown previously that losartan (with or without the addition of hydrochlorothiazide) is tolerated better than and is as effective as many other antihyper-tensive regimens [911,20-24]. Second, the vast majority of the patients was recruited from primary health care and hence our patients, were likely to have been representative of hypertensives in the general population. Third, the tolerability and quality of life were evaluated by using several methods, namely spontaneous reporting, active questioning and validated questionnaires. Fourth, and maybe most importantly, patients who had previously been treated with angiotensin II antagonists and calcium antagonists were excluded from the study in order not to bias any of the treatment groups. Blood pressures measured at trough were reduced equally well by administration of 5 mg amlodipine (or 10 mg if necessary) as they were by 50 mg losartan (plus HCTZ if necessary) and better than they were with losartan monotherapy. This was also evident from the difference in the number of patients for whom we attained the target diastolic blood pressure of below 90 mmHg. In a previous smaller study, a combination of 50 mg losartan and 12.5-25 mg HCTZ had been found equally effective in lowering the blood pressure as 5-10 mg amlodipine plus 12.5 mg HCTZ . On the other hand, losartan mono-therapy compared favourably with the use of other dihydropyridines (nifedipine gastrointestinal therapeutic system and felodipine extended release) regarding its blood-pressure-lowering efficacy [9,25]. However, amlodipine, possibly partly because of its long half life, is somewhat more effective than is felodipine extended release when the blood pressure is recorded 23-25 h after administration of the dose [26J. That the addition of 12.5 mg hydrochlorothiazide to 50 mg losartan improved the blood pressure control was not at the expense of reduced treatment tolerability and changes in quality of life. The usefulness of an antihypertensive medication depends on the balance between efficacy and tolerability, 1333 so the evaluation of a therapeutic regimen should consist of valid measurements of both of these variables. In the present study, among the patients administered initially amlodipine, compared with losartan-treated patients, there was a significantly higher proportion of individuals answering 'yes' to the question 'Have you felt any kind of discomfort since the last visit?'. The difference between losartan and amlodipine monotherapies was just below 10%, with a wide confidence interval but separated from zero. A similar pattern could be seen in the percentage of patients who withdrew from the study due to adverse effects; again significantly fewer patients withdrew with losartan (2%) than did with amlodipine (8%). When the patients were questioned actively regarding specific symptoms, an expected increase in the symptom of swollen ankles was reported with amlodipine but not with losartan. The increased incidence of swollen ankles with amlodipine was already evident after 6 weeks (a 6.8% increase versus baseline), that is, with the lower dose of the drug. The increase was even more pronounced after 12 weeks (a 22.2% increase versus baseline). Ankle oedema or swelling of the ankles is a well-established side effect of calcium antagonists [1,5,6] that is not seen with losartan; this has previously been established (e.g. in comparisons with nifedipine gastrointestinal system  and amlodipine ). An unexpected increase in reporting of 'dizziness upon standing' was evident among patients in both of the losartan groups, with a tendency towards a higher reported incidence among patients in the group who also needed to be administered hydrochlorothiazide. We have no valid explanation for this finding. Those patients reporting 'dizziness upon, standing' did not have any orthostatic fall in blood pressure or increase in heart rate on standing. Symptoms and side effects of antihypertensive therapy can be multi-dimensional and influence the patient's physical state, emotional well-being, sexual and social functioning and cognitive functioning, and thus affect his or her quality of life negatively. Furthermore, this could have implications for compliance and the patient's motivation to submit to long-term therapy [12,27,28]. Measurements of the quality of life and well-being are very important but have been a much neglected area, mainly due to the lack of good, validated measurement scales. Unfortunately, many studies have been open or single-blinded ones with undocumented methods, lacking control groups and concerning small unselected populations. This study fulfils most of the methodological requirements put forward for studies of the quality of life (i.e. the trial was randomized, blinded and had a parallel-group design with an adequate sample size and a 12-week follow-up ). The questionnaire used had been validated in many previous studies and was developed to provide a reproducible measure of subjective well-being 1334 Journal of Hypertension 1997, Vol 15 No 11 or distress. The PGWB instrument has been documented extensively in terms of its reliability and validity [17,18]. The improvement in PGWB index before randomization, when previous therapy was discontinued (Fig. 4), is interesting and probably reflects 'hidden' side effects of the chronic therapy being administered before the study , The general well-being, assessed in terms of the PGWB index total score, at baseline was slightly higher than that of the general population, indicating that we had an asymptomatic hypertensive population; it was thus comparable to other studies of the quality of life . During the study, there was a significant improvement from baseline in several domains with losartan (Table 4) but nor with amlodipine. In the between-group analyses, only two domains (positive well-being and general health) exhibited significant differences, both in favour of losartan treatment (Table 4). The proportion of patients whose total score had improved by week 12 was 60% for the group being administered losartan monotherapy and 50% for those being administered amlodipine monotherapy (P = 0.011, Table 5). The corresponding percentage for patients in the group in which 62% of the patients had been administered supplementary treatment with hydrochlorothiazide from week 6 was 54% (NS). Does this mean that a patient administered losartan in this study actually felt better than 'normal'? That is not our conclusion, but it is possible that the patients administered losartan felt better than did those administered amlodipme. This could have been due to the fact that a lowering of blood pressure equates to an improvement in well-being, whereas more symptoms equates to a worsening in well-being. If this were true, the result of this study would be that the balance in this regard is more favourable for losartan treatment than it is for arniodipine treatment. One could argue that the sequence of treatment could have influenced the quality of life and that the mere fact of being in a study could have changed a patient's quality of life for the better. We believe, however, that this could not have favoured any of the groups in this randomized study in which members of all groups were subjected to the same procedures. The absence of a change in PGWB index for amlodipinetreated patients is compatible with previous reports that calcium antagonists have no positive or adverse effect on the well-being and quality of life of patients [24,25,30-33]. Only high-dose nifedipine has affected well-being negatively and contributed to a high withdrawal rate due to complaints of symptoms [16,34]. Furthermore, in another study comparing amlodipine and captopril, the conclusion drawn was that both drugs maintained the patient's quality of life equally well . Previously reported maximum interdrug differences in general well-being score between therapies, assessed in terms of the PGWB, have been of the order of 3 -4 between enalapril and atenolol , captopril and methyldopa , captopril and verapamil  and captopril and propranolol . Here, we observed a numerical difference (P.= 0.058) of 1.9 between losartan and amlodipine monotherapies after 12 weeks. In summary, after 12 weeks' treatment, 50 mg losartan with 12.5 mg HCTZ added after 6 weeks (in 62% of cases) was equally effective at lowering the blood pressure as was 5 mg amlodipine increased to 10 mg (in 56% of cases). Both of these regimens were superior to 50 mg losartan increased to l00 mg (in 61% of cases). Both the losartan and the amlodipine regimens were safe and tolerated we but there were fewer adverse effects and withdrawals from the study for the losartan groups than there were for the amlodipine group. Moreover, losartan scored better in some domains of PGWB than did amlodipine. The clinical significance of this observation, however, needs to be studied further. Acknowledgements We would like to acknowledge the co-ordination efforts of the monitors and especially Dr Hans Bostrom, Merck Sharp & Dohme (Sweden). We would also like to express our sincere thanks to Mrs Cecile Dubois, Brussels, for her dedicated work with the statistical analyses. Moreover, we would like to acknowledge the following investigators who participated in the study: in Australia Eddy Bajrovic (West Perth), Michael Beckoff (Murray Bridge), Bruce Jackson (Preston), Anthony Johnson (Brisbane), Jeffrey Karrasch (Kippa Ring), John Kelly (Kogarah), Patrick Phillips (Adelaide), Craig Smee (Dickson), Andrew Tonkin (Heidelberg), Anne Tonkin (Adelaide), Judith Whitworth (Kogarah), Leigh Wilson (Adelaide) and Roger Wyndham (Sydney), in Finland Mikael Hedenborg (Helsinki), Antti Jounela (Oulu), Leena Lahdenne (Helsinki), Urpo Laisi (Helsinki), Maini Leinonen (Mantsala), Jarmo Lumme (Oulu),' Tapio Ropponen (Helsinki), Raimo Siloaho (Pieksamaki), Jukka Ulkuniemi (Kiiminki), Mauno Vanhala (Pieksamaki), and in Sweden Jan Alvang (Trollhattan), Mats Andersson (Vannas), Per Andren (Vaxjo), Jan-Axel Axelsson (Boras), Anders Berglund (Borlange), Eva-Pia Darsbo (Ostersund), Gunnar Ekblad (Lidkoping), Mark Flfast (Ytterby), Ulla Eriksson (Upplands Vasby), Lars Froberg (Hoganas), Peter Gadd (Norrahammar), Carin Hallendal (Han inge), Per Hauschildt (Skillingaryd),. Samy Hellerstedt (Kallhall), Thomas Hermansson (Haninge), Ingalill Hildebrand (Linkoping), Christer Hjortsberg (Helsingborg), Sofia Hollenberg (Upplands Vasby), Bernt Johansson (Horndal), Anders Juhlin (Tyringe), AnnChristine Knutsson (Angelholm), Magnus Karegard (Kristianstad), Inger Larsbrink (Vasterhaninge), Hakan Quality of life during losartan treatment Dahlöf et al. Larsson (Byske), Lars Larsson (Trollhättan), Åsa Larsson (Trollhattan), Mikael Lilja (Husum), Lena Linden (Tullinge), Anders Lindh (Åkersberga), Christina Lithner (Krokom), Peter Nilsson (Dalby), Niels Norgaard (Hoganas), Birgitta Strang-Olander (Malmö), Ivar LundOlseri (Snndviken), Birger Ossiansson (Vaxjö), Eva Pavek (Knivsta). Karel Pavek (Uppsala), Lars Nygaard Pedersen (Helsingborg), Mats Persson (Umea), Hillevi BlomPfeiffer (Upplands Väsby), Tom Roffey (Lidköping), Bengt Silfverhielm (Ytterby), Anita Spjuth (Trollhättan), Christina Storm (Sandviken), Dag Stromberg (Stockholm), Bo Sundqvist (Vannäs), Ulf Svensson (Lidkoping), Peter Ten brock (Tumba), Bengt-Olov Tengmark (Kista), Gunilla Vatnaland (Ängelholm), Niels Voergaard (Solleftea), Per Wendell (Haninge), Mats Wernerson (Norrköping), Lena Westerberg (Sandviken), Bo Westerdahl (Trollhattan), Kerstin Westergren (Sandviken), Evert Zetterberg (Sandviken), Ingemar Zettergren (Sandviken), Kristina Åresund (Sandviken) and Birgitta Örtelind (Hisingsb'acka). References 1 2 3 A 5 6 7 8 9 10 11 12 13 14 15 Hansson L, Svensson A, Dahlof B, Eggertsen R: Drug treatment of hypertension. In Handbook of Hypertension. Edited by Robertson JIS. Amsterdam: Elsevler; 1992, 15:655-708. Collins R, Peto R, MacMahon S, Herbert P, Fiebach NH, Eberlein KA, et al.: Blood pressure, stroke and coronary heart disease. Part 2, short-term reductions in blood pressure: overview of randomised drug trials in their epidemiological context. Lancet 1990, 335: 827-839. MacMahcn S, Rodgers A: The effects of antihypertensive treatment on vascular disease: reappraisal of the evidence in 1994. J Vase Med Biol 1993, 4:265-271. Report of Medical Research Council Working Party on mild to moderate hypertension: Adverse reactions to bendoflumethiazide .and propranolol for the treatment of mild hypertension. Lancet 1981, ii:539-543. Hansson L, Dahlof B. Treatment of hypertension with calcium antagonists. In Clinical Aspects of Calcium Entry Blockers. Edited by van Zwieten PA. Basel: Karger; 1989, 2:103-124.. Gustafsson D: Microvascular mechanisms involved in calcium antagonist edema formation. J Cardiovasc Pharmacol 1987, 10 (suppl 0:3121-5131. Karlberg BE: Cough and inhibition of the renin-angiotensin system. J Hypertens 1993, 11 (suppl):S49-S52. Fletcher AE, Palmer AJ, Bulpi't CJ: Cough with angiotensin converting enzyme inhibitors: how much of a problem? J Hypertens 1994, 7:416-424. Goldberg Al, Dunlay MC, Sweet CS: Safety and tolerability of losartan potassium, an angiotensin II receptor antagonist, compared with felodipine ER and angiotensin converting enzyme inhibitors for the treatment of systemic hypertension. Am J Cardiol 1995, 75: 793-795. Carr AA, Prisant LM: Losartan: first of a new class of angiotensin antagonists for the management of hypertension. J din Pharmacol 1996, 36:3-12. Goa KL. Wagstaff AJ: Losartan potassium: a review of its pharmacology, clinical and tolerability in the management of hypertension. Drugs 1996, 51:820-845. Jones IK, Gorkin L, Lian JF, Staffa JA, Fletcher AP: Discontinuation of and changes in treatment after start of new courses of antihypertensive drugs: a study of a United Kingdom population. BMJ 1995, 311:293-295. Hjemdal P, Wiklund IK: Quality of life on antihypertensive drug therapy: scientific endpoint or marketing exercise? J Hypertens 1992, 10:1437-1446. Wanda G, Grassi G: Considerations on current and future trials in hypertension. Blood Pressure 1996, 5:327-332; Croog SH, Levine S, Testa MA, Brown B, Bulpitt CJ, Jenkins D, et al.: The effects of antihypertensive therapy on quality of life. N Engl J Med 1986 314:1657-1664. 1335 16 Os I, Bratland B, Dahlof B, Gisholt K, Syvertsen JO, Tretli S: Lisinopril or nifedipine in essential hypertension? A Norwegian multicenter study on efficacy, tolerability and quality of life in 828 patients. J Hypertens 1991, 9:1097-1104. 1 7 Dupuy HJ: Psychological General Weil-Being (PGWB) index: assessment of quality of life in clinical trials of cardiovascular therapies. Edited by Wenger NK, Mattson ME, Furberg CF, Elinson J. New York: Le Jacq Publishing Inc.; 1984:170-183. 18 Wiklund I, Karlberg J: Evaluation of quality of life in clinical trials. Selecting QoL measures. Contr din Trials 1991, 12 (suppl):S204-S216. 19 Bauer P: Multiple testing in clinical trials. Statist Med 1991, 10:871-890. 20 Weber MA, Neutel JM, Smith DGH: Controlling blood pressure throughout the day: issue in testing a new antihypertensive agent. J Hum Hypertens 1995, 9 (suppl 5):S29-S35. 21 MacKay JH, Arcuri KE, Goldberg Al, Snapinn SM, Sweet CS: Losartan and low-dose hydrochlorothiazide in patients with essential hypertension. Arch Intern Med 1996, 156:278-285, 22 Ruilope LM, Simpson RL, Toh J, Arcuri KE, Goldberg Al, Sweet CS: Controlled trial of losartan given concomitantly with different doses of hydrochlorothiazide in hypertensive patients. Blood Pressure 1996, 5:32-40. 23 Dahlof B, Keller SE, Makris L, Goldberg Al, Sweet CS, Lim NY: Efficacy and tolerability of losartan potassium and atenolol in patients with mild to moderate essential hypertension. Am J Hypertens 1995, 8:578-583. 24 Oparil S, Barr E, Elkins M, Liss C, Vrecenak A, Edeiman J: Efficacy, tolerability and effects on quality of life of losartan, alone or with hydrochlorothiazide versus amlodipine, alone or with hydrochlorothiazide in patients with essential hypertension. Clin Therapeutics 1996, 18:608-625. 25 Weir MR, Elkins M, Liss C, Vrecenak AJ, Barr E, Edeiman JM: Efficacy, tolerability and quality of life of losartan alone or with hydrochlorothiazide versus nifedipine GITS in patients with essential hypertension. Clin Therapeutics 1996, 18:411-428. 26 Hoegholm A, Wiinberg N, Rasmussen E, Nielsen PE: Comparative effects of amlodipine and felodipine ER on office and ambulatory blood pressure in patients with mild to moderate hypertension. J Hum Hypertens 1995, 9 (suppl 1):S25-S28. 27 Luscher TF, Vetter H, Sigenthaler W, Vetter W: Compliance in hypertension: facts and concepts. J Hypertens 1935, 3 (suppl 1):S3-S9. 28 Perry HM Jr, Came! GH: Survival of treated hypertensive patients as a function of compliance and control, J Hypertens 1984, 2 (suppl 3):S197-S199. 29 Dahlof B, Andren L, Eggertsen R, Jern S, Svensson A, Hansson L: Potentiation of the antihypertensive effect of enalapril by randomised addition of different doses of hydrochlorothiazide. J Hypertens 1985, 3 (suppl 3):S483-S486. 30 Croog SH, Kong BW, Levine S, Weir MR, Baume RM, Saunders £: Hypertensive black men and women. Quality of life and effects of antihypertensive medications. Arch Intern Med 1990, 150:1733-1741. 31 Fletchei AE, Chester PC, Hawkins CM, Latham AN, Pike LA, Bulpitt CJ: The effects of verapamil and propranolol on quality of life in hypertension. J Hum Hypertens 1989, 3:125-130. 32 Zachariah PK, Brobyn R, Kann J, Levy B, Margolis R,; McMahon FG, et al.: Comparison of quality of life on nitrendipine and propranolol. J Cardiovasc Pharmacol 1988, 12 (suppl 4):S29-S35. 33 Fletcher AE, Bulpitt CJ, Chase DM, Collins WC, Furberg CD, Goggin TK, et al.: Quality of life with three antihypertensive treatments. Cilazapril, atenolol, nitrendipine. Hypertension 1992, 19:499-507, 34 Testa MA, Hollenberg NK, Anderson RB, Williams GH: Assessment of quality of life by patient and spouse during antihypertensive therapy with atenolol and nifedipine gastrointestinal therapeutic system. Am J Hypertens 1991, 4:363-373. 35 Omvik P, Thaulow E, Herland OB, Eide I, Midha R, Tuner RR: Doubleblind parallel, comparative study on quality of life during treatment with amlodipine or enalapril in mild or moderate hypertensive patients: a multicentre study. J Hypertens 1993, 11:103-113. 33 Applegate WB, Phillips HL, Schnaper H, Sheperd AMM, Schocken D, Luhr JC, et al.: A randomised controlled trial of the effects of three antihypertensive agents on blood pressure control and quality of life in older women. Arch Intern Med 1991, 151:1817-1823.