tolerability, safety, and quality of life and hypertensive therapy: the case for low-dose diuretics

Tolerability, Safety, and Quality of Life and Hypertensive Therapy: The Case for Low-Dose Diuretics Matthew R. Weir, MD, Baltimore, Maryland, John M. Flack, MD, MPH, Winston-Salem, worth Carolina, William B. Applegate, MD, MPH, Memphis, Tennessee

The safety and tolerability of antihypertensive therapies are an important clinical concern, because the demonstrated benefits of successful blood pressure-lowering depend on long-term compliance with pharmacologic treatments. Thiazide diuretics and p blockers have been specifically recommended as preferred initial drug therapy by the Fifth Joint National Committee on Detection, Evaluation, and Treatment of High Blood Pressure (JNC-V), unless their use is contraindicated by concomitant disease, there is intolerance to these agents, or there is a specific indication for another drug class. These recommendations are a result of the lengthy clinical experience with these drugs and the results of long-term clinical trials that have demonstrated significant reductions in cardiovascular morbidity and mortality. However, data from these same clinical studies have also shown that diuretics (and p blockers) can cause abnormalities in carbohydrate, electrolyte, and lipid metabolism and also may influence quality of life. The safety of diuretics was evaluated with regard to effects on carbohydrate, electrolyte, and lipid metabolism by seeking references from a MEDLINE search of documents published from 1966 to 1994 based on the search terms “hypertension, ” “human,” and “hydrochlorothiazide” (HCTZ) dosed in a range of 12.5-25 mg daily. Two long-term clinical trials using low-dose (12.5-15 mg/day) chlorthalidone-the Systolic Hypertension in the Elderly Program (SHEP) and the Treatment of Mild Hypertension Study (TOMHS)-were also included. During the course of treatment with HCTZ in these studies, serum potassium was reduced and uric acid was increased in a dose-

I I From the Division of Nephrology and Clinical Research Unit, Depart-

ment of Medicine, Unrversity of Maryland School of Medicine, Baltimore, Maryland (M.R.W.); the Divisions of Surgical Sciences, Medicine, and Public Health Sciences, Bowman Gray School of Medicine, Winston-Sa- lem. North Carolina fJ.M.F.1: and the Department of Prevenhve Medicine, University of Tennessee, Memphis, Tennessee (W.B.A.).

Requests for reprints should be addressed to Matthew R. Weir, MD, Diii sion of Nephrology, Unrversii of Maryland School of Medicine, 22 South Greene Street, Baltimore, MD 21201-1595.

I I

01996 by Excerpta Medica, Inc. All rights reserved.

dependent manner. Although low doses of HCTZ elevated serum glucose, cholesterol, and triglycerides, the magnitude of effect was rsmall in most cases and was probably of no clinical significance. Other laboratory parameters were not adversely affected, and subjective reporting of clinical adverse events was generally lower with low-dose HCTZ than with placebo or standard HCTZ dosing. The literature on thte effects of low-dose diuretic therapy on quality of life is not large, although the results from tie SHEP and TOMHS studies support the concept that diuretics either do not interfere with, or may actually improve, quality of life in hypertensive patients. Low-dose thiazide treatment is a well- tolerated, excellent first-line choice for hypertensive patients, especially older patients. However, diuretics should probably be avoided, whenever possible, in patients with preexisting diabetes, gout, and in men with erectile dysfunction. Am J Med. 1996;1Ol(suppl 3A):83S-92s.

T hiazide diuretics are indicated for the management of hypertension either as a sole therapeutic

agent for stage 1 or stage 2 (mild-to-moderate ) blood pressure elevations or in combination witlh other antihypertensive agents to control more severe forms of hypertension. Traditionally, when used alone, hydrochlorothiazide (HCTZ) has been pre- scribed in doses of 225-50 mg once daily. These tablet strengths are the only ones currently available in the United States. However, when used in combination with other antihypertensive products, lower doses of 12.5-15 mg daily have proven to be effective hypotensive agents.

The fifth Report of the Joint National Committee on Detection, Evaluation, and Treatment of High Blood Pressure (JNC-V) stated that the goal for treating patients with hypertension is to “prevent morbidity and mortality associated with high blood pressure and to control blood pressure by the least intrusive means possible.“’ After a trial of nonphar- macologic interventions, diuretics or p b1oc.ker-s are specifically recommended by JNC-V as first-line agents, unless their use is contraindicated by concomitant disease, there is intolerance to these agents, or an antihypertensive agent from a different

0002-9343/96/$15.00 3A-83s Pll SOOO2-9343(96)00271-g

therapeutic class is specifically indicated. This rec- ommendation is partly the result of the lengthy clinical experience with these two classes of agents, particularly with the diuretics, in a wide variety of patient types. Furthermore, diuretics and p blockers are the only two classes of antihypertensive drugs used in long-term controlled clinical trials showing a clear reduction in cardiovascular morbidity and mortality.

It is a well-accepted principle in medicine that the lowest effective dosage of a therapeutic agent should be chosen, to minimize the likelihood of adverse events occurring as a result of the treatment. Higher doses of HCTZ have been implicated in alterations of carbohydrate, lipid, and electrolyte metabolism, and adverse clinical effects, such as headache, impaired libido, lethargy, and gastrointestinal discomfort. Con- sequently, efforts to assess the efficacy, safety, and tolerability of lower doses of HCTZ have been ex- plored. With regard to quality of life, higher doses of diuretics (250 mg) have been associated with negative impact, whereas low doses of chlorthalidone (15 mg) may actually improve quality of life.

METHODS Literature Review

References were sought from a MEDLINE search of English-language documents from 1966 to 1994 based on the search terms “hydrochlorothiazide,” “hypertension,” “human,” and “safety” to identify per- tinent literature, including review articles and edi- torials. The citations within each retrieved article were also searched for additional references. The comprehensive list was then limited to studies that evaluated low-dose HCTZ therapy (defined in this article as 12.5 mg HCTZ once daily) alone or in combination with other antihypertensive agents. Rea- sons for exclusion of an article included study of a low-dose of HCTZ other than 12.5 mg daily, lack of a description of safety results, strictly observational or noncomparative design, or the reporting of only pharmacokinetic information. In addition, two long- term clinical trials using low-dose chlorthalidone, the Systolic Hypertension in the Elderly Program (SHEP)’ and the Treatment of Mild Hypertension Study (TOMHS) ,3 were also included. The discussion about quality of life involves all diuretics due to the limited amount of available information.

Analysis Data from identified articles were extracted inde-

pendently by the authors using standard tables eval- uating the effect of HCTZ therapy on the following safety parameters: serum potassium, glucose, cholesterol, triglycerides, uric acid, subjective side effects, and spontaneously reported clinical adverse events. These findings were then validated for agree- ment before summarizing. Discussion and consensus

were used to resolve discrepancies with1 reference to the original article.

RESULTS Studies Identified

Fifteen studies that were published during 1976- 1994 that evaluated the safety and efficacy of low- dose HCTZ were identified (Table I). A total of 3,300 patients were randomized in these studies across all treatment groups (not all patients completed the studies). Eight of the trials were placeb,o controlled, and seven compared low-dose HCTZ monotherapy to a 25-112.5 mg daily dose. Safety data were available for 651 patients receiving low-dose HCTZ. Two long-term clinical trials using low-dose chlorthalidone will be discussed separately.

Demographically, these studies consisted of approximately 60% male and 40% female patients in an age range of 20-87 years. Two of the studies specifically enrolled elderly patients.4s5 Approximately 80% of the patients were white and 15Oh were black. How- ever, one study focused on postmenopausal black women.‘j In contrast to the clinical data, the safety data were generally reported in aggregate; therefore, an evaluation of the safety by racial subpopulation was not possible.

Study Design The most commonly used study design was a dou-

ble-blind, randomized, parallel-group ~clinical trial, generally with a placebo lead-in period and dose es- calation if blood pressure was not adequately controlled at a given dose leve1.“‘5’7m13 McKenney et al per- formed a randomized, double-blind, crossover study of low-dose versus high-dose ( 112.5 mg daily) HCTZ therapy.‘j More recently, a double-blind, randomized, factorial design has been used to allow for an expan- sion in the number of comparative group~.‘~~~~

Effect on Serum Potassium HCTZ and other thiazide diuretics have been rec-

ognized to reduce serum potassium levels in a dose- dependent manner. ” This reduction has been of concern to clinicians when diuretics were used in the management of hypertension due to a fear that iat- rogenic hypokalemia could provoke cardiac arrhyth- mias. In addition, it has been demonstrated that diuretic-induced hypokalemia may leaId to insulin resistance, since repletion of potassium restores insulin sensitivity.” Despite this concern, only nine of the 15 published studies of low-dose HCTZ reported specific effects of HCTZ therapy on serum potassium.4,6,8,11,13-17 The impact of HCTZ treatment on serum potassium was also reported subjectively (e.g., “hypokalemia”) in two studies, ‘.’ whereas no impact of treatment was reported in four of the 15 studies.g,‘0~12~1” The use of potassium sup:plementation was described in only two studies.16,17

3A-84S September 30, 1996 The American Journal of Medicine’ Volume 101 (suppl 3A)

Six of the 15 studies did not find a statistically or clinically significant change of serum potassium from baseline or placebo in patients treated with low-dose HCTZ.6’“-‘3~‘“~‘s However, six of the 15 studies showed either a statistically or clinically signiflcant change in serum potassium at higher doses (25- 112.5 mg daily) ,%llJ3J5J7J9

In eight studies, mean changes in serum potassium levels ranged from 0.1 to -0.3 mEq/L relative to baseline or placebo in patients treated with low-dose HCTZ.‘.6,s,11,13,15-17 These results were in contrast to average declines of 0.2-0.7 mEq/L in serum potassium in patients receiving 25 mg HCTZ daily.6*13.15-17

In another study, 21% of patents treated with low- dose HCTZ experienced a decrease in serum potassium of 0.5 mEq/L. In comparison, 30% of patients treated with 25 mg HCTZ daily experienced a decrease of 0.5 mEq/L of serum potassium.14

Three recently completed studies showed a significant decrease in serum potassium at the 25 mg HCTZ daily dose but not at the 12.5 mg HCTZ daily dose.13-15 This dose-dependent effect on serum potassium reduction had been shown previously for daily doses of 25-100 mg of HCTZ.” A clear dose- response relationship with regard to the magnitude of effect of HCTZ therapy on serum potassium was therefore established by these comparative studies; however, a slight, but clinically insignificant, decrease in serum potassium is seen at the 12.5 mg HCTZ daily dose. In the 4Smonth TOMHS study,3 serum potassium fell an average of 0.4 mEq/L among those taking 15 mg/day of chlorthalidone versus a fall of 0.02 mEq/L in the placebo group.

Effect on Serum Glucose It is well recognized that serum glucose levels

should be monitored in patients on chronic diuretic therapy, particularly if they have preexisting diabetes mellitus, borderline glucose tolerance tests, elevated fasting blood glucose levels, or a family history of diabetes mellitus.

Nine of the 14 studies reviewed here reported information regarding the impact of HCTZ on serum glucose levels. 4,5*8~11,13-18 In general, the changes in serum glucose levels relative to placebo or baseline were not clinically significant.“,12,1s

Intravenous glucose tolerance tests in low-dose HCTZ patients revealed no changes between on- treatment and pretreatment values, I7 In those studies reporting changes in serum glucose values, levels increased by a mean of 2-9 mg/dL in patients receiving any dosage level of HCTZ.4,s,13,16 The effects were generally not dose related, with the exception of the findings of Chrysant et al, l3 who found no change in serum glucose levels relative to baseline in the low- dose HCTZ group, compared with a mean 9 mg/dL increase in the 25 mg/dL HCTZ group. In the TOMHS study,3 where chlorthalidone was given in comunc-

SYMPOSIUM ON HYPERTENSION/WEIR ET AL

TABLE I Low-Dose Hydrochlorothiazide Studies identified for this

Review

Sample Reference Size Study Design

Berglund and Andersson, 40 Rand, d-b, pla’cecon, 1 97616 cross

Beermann and Groschinsky- 9 S-b, dose-titration Grind, 1978l’

Lorimer et al, 198318 18 Rand, d-b, place-con, cross

McKenney et al, 19866 9 Rand, d-b, cross Pool et al, 19877 394 Rand, d-b, para-grp Rosenthal et al, 19908 81 Rand, d-b, para-grp Hart, 19914 329 Rand, d-b, par-a-grp Miller, 1991g 394 Rand, d-b, para-grp Stein et al, 1992lO 19 Rand, d-b, para-grp Dal Palu et al, 19935 96 Rand, d-b, par-a-grp Mersey et al, 1993ll 345 Rand, d-b, para-grp Scholze et al, 199315 534 Rand, d-b, place-con,

para-grp, 4 x 3 factorial design

Canter et al, 199414 460 Rand, d-b, place-con, para-grp, 4 x 4 factorial design

Fernandez et al, 1994l* 67 Rand, d-b, place-con, para-grp

Chrysant et al, 199413 505 Rand, d-b, place-con, para-grp

Cross = crossover; D-b = double-blind; Para-grp = parallel-group; Place con = placebo controlled; Rand = randomized; S-b = single-blind.

tion with ongoing lifestyle modifications (weight loss, salt and alcohol restriction, and increased physical activity), the average change in serum glucose levels was -0.5 mg/dL versus -0.4 mg/dL in the placebo (lifestyle modifications only) group.

Glucose intolerance in patients receiving 100 mg HCTZ has been abrogated by restoring serum polas- sium levels2’ By using glucose clamp technology, it has been demonstrated that diuretic-induced potassium deficiency interferes with glucose tolerance by antagonizing both pancreatic P-cell sensitivity to glucose and tissue sensitivity to insulin. Consequently, lower doses of HCTZ are less likely to induce glucose intolerance, largely because they cause hypo kalemia less frequently.

While thiazide diuretics rarely precipitate the de- velopment of a frank case of diabetes, clinical experience dictates that care should be exercised in the treatment of patients with underlying diseases that would predispose the patient to experiencing adverse outcomes from even slight increases in serum glucose.

Effects on Serum Lipids In the studies analyzed in this review, only three

described the results of serum cholesterol changes in response to low-dose HCTZ therapy. Two of the studies reported no changes of clinical signifi-

September 30, 1996 The American Journal of Medicine@ Volume 101 (suppl 3A) 3A-85s

cance.6,21 In the third study, mean serum cholesterol in the low-dose HCTZ group actually decreased by 0.44 mg/dL relative to baseline, whereas a mean increase of 6.51 mg/dL occurred in the placebo group. l1 Investigators found significant elevations in serum cholesterol of similar magnitude in both the low-dose and high-dose ( 112.5 mg daily) HCTZ groups.6

Two studies evaluated the effect of HCTZ treatment on triglyceride levels and reported no significant changes among patients treated with 12.5- 112.5 mg HCTZ daily.‘,i7 The effect of HCTZ therapy on serum triglyceride levels was not reported in other studies.

A review of several clinical trials concluded that studies of long-term treatment with thiazide diuretics indicated that the modest elevation of serum cholesterol levels observed were transient and subsided back to, or below, baseline values after approximately 1 year of treatment. The doses used in all of the studies reviewed were 225 mg HCTZ daily. Nevertheless, one study among individuals taking either HCTZ or chlorthalidone over 6 years who were then randomized to continue diuretic treatment or placebo= doc- umented a fall in cholesterol in the participants randomized to placebo as compared with those remaining on diuretics. A 1995 report by Elliotz3 found that, by using conventional intention-to-treat analyses, if as few as 8% of diuretic-treated patients dis- continued treatment, the rise in cholesterol was no longer significant after the first year of treatment.

More recently, a large meta-analysis evaluated the effects of all classes of antihypertensive agents on serum lipid profiles. ” The results of 236 studies in which diuretics were used (thiazide-like diuretics in 214 studies, high doses of diuretics in 47 studies) demonstrated that cholesterol and triglyceride levels were increased primarily when high doses of diuretics were used and that these increases were greater among blacks than whites. Diuretic-induced changes in triglyceride levels were seen only in short-term studies, and the effect was greater in men than women. High-density lipoprotein cholesterol levels were only noted to be increased in diabetic patients, although diuretics usually do not influence high-density-lipoprotein cholesterol levels. Finally, chlorthalidone was found to be more closely related to increases in low-density lipoprotein cholesterol levels than were other diuretics.

These larger reviews, 20~24 as well as data from the present analysis, indicate that low-dose HCTZ may be expected to have a minimal and, in most patients, a clinically insignificant effect on serum lipid profiles.

Using an intention-to-treat analysis, the changes in total and low-density-lipoprotein cholesterol over 48 months in the TOMHS study with low-dose chlorthalidone (15 mg/day) were similar to the changes seen in the placebo group.25 Changes in total and

low-density-lipoprotein cholesterol averaged -4.5 and -3.6 mg/dL and -5.1 and -3.6 mg/dL, respectively, in the chlorthalidone and placebo (lifestyle modification only) groups. However, the impact of low-dose chlorthalidone on cholesterol was less than that previously reported at higher doses.24’2G’27 Aver- age changes in total cholesterol were, however, significant through 12 months in the chlorthalidone (3.8 mg/dL) versus placebo (-7.3 mg/dL) groups, but not thereafter. Nevertheless, an on-treatment analysis between 12 and 48 months showed that total cholesterol levels averaged 7.7 mg/dL higher in the chlorthalidone compared with the placebo group. In TOMHS, the chlorthalidone lipid effect was most prominent among individuals who either did not lose or actually gained weight during follow-up.

Effect on Serum Uric Acid Levels Whether due to increased resorption or decreased

secretion, HCTZ and chlorthalidone axe known to cause elevations in serum uric acid levels. The hyper- uricemia that occurs with thiazide diuretics is usually mild and asymptomatic, although patients predisposed to acute attacks of gout may be clinically susceptible to the fluctuations caused by these agents.

Three studies subjectively stated that no signill- cant change in serum uric acid levels occurred during study treatment. ‘,i3.17 Four studies described ob- jective changes in serum uric acid levels. Berglund and Andersson” noted an apparent dose-dependent increase of 0.6-0.95 mg/dL in patients treated with 12.5-50 mg HCTZ daily [P = 0.01). Hart4 reported a significant (P = 0.0082) increase when 12.5 mg of HCTZ was added to 20 mg of lisinopril in a group of 282 elderly patients. Others observed a mean increase of 0.39 mg/dL in the parallel placebo group.” Chrysant et all3 reported an increase of 0.7 mg/dL and 1.1 mg/dL in the low-dose HCTZ and 25 mg daily dosing groups, respectively, after l2 weeks of therapy.

Therefore, increases in serum uric acid levels also appear to be dose dependent but not of clinical significance in most hypertensive patients treated with thiazides. Nevertheless, whenever possible, thiazide diuretics, particularly higher doses, should be avoided for long-term management of hypertension in individuals with gout.

Effects on Other Laboratory Parameters Various studies evaluated other laboratory param-

eters (serum sodium, calcium, magnesium, phos- phate, creatinine, blood urea nitrogen, liver function tests, platelets); clinically signiflcant changes in these parameters were not reported in any study. Because of their effects on lithium excretion, thiazide diuretics are contraindicated in patients receiving lithium; however, none of the studies reported here described an impact of study treatment on lithium levels.

3A-86S September 30, 1996 The American Journal of Medicinee Volume 101 (suppl 3A)


Effects on Subjectively Repotted Clinical Adverse Events

The reporting of clinical adverse events is inher- ently more subjective than that of laboratory adverse events. Two studies did not mention clinical adverse events, 6,17 and a third reported only that no patients experienced clinical adverse events, which resulted in study drug discontinuation. lo

In six of the 15 studies, clinical adverse events that were attributed only to the assigned treatment regimen were discussed, but safety information regarding each dosage level was not provided.5S7-g*14,” Others, however, noted that “headache was most common in the placebo group,” l5 while another study reported that the overall incidence of adverse events was higher in the placebo group (19%) than in patients treated with HCTZ at 6.25-25 mg daily ( 15%) .I4

In the remaining six studies4,“-‘3,‘6~‘s information regarding clinical adverse events was reported separately for 331 patients receiving low-dose HCTZ. Berglund and Andersson16 used a unique patient questionnaire for subjective adverse events that, the authors believed, best measured overall “well-being” rather than specific drug-induced adverse events. Their results showed that patients in the low-dose HCTZ group reported better tolerance than patients receiving placebo, 25 mg HCTZ daily, or 50 mg HCTZ daily. In these six studies, the incidence of clinical adverse events, independent of causality, was 5.3- 58.8% for placebo patients, O-35% for low-dose HCTZ patients, and 6-35.6% for patients treated with 225 mg HCTZ daily (Table II).

Headache was consistently the most frequently reported clinical adverse event. Other adverse events that appeared in 25% of patients in any patient group were abdominal distention, asthenia, cough, dizziness, palpitations, and pharyngitis (Table III).

Overall, in these published trials, low-dose HCTZ proved to be safe and well tolerated by the great majority of patients. It is interesting to note that low-

TABLE II Overall Incidence (%) of Clinical Adverse Events Reported

by Treatment Regimen

HCTZ HICTZ Reference Placebo 12.5 mg qd 25 mg qd

Berglund and Anderson, 1 97616 5.3 4.7 7.7G3.5t

Mersey et al, 1993ll 40 35 N/S Lorimer et al, 198318* 6 0 Hart, 19914 N/S 9 N”/S Fernandez et al,

199412 58.8 29.4 N/S Chrysant et al, 199413 29.5 25.6 .35.6 HCTZ = hydrochlorothiazide; N/S = not specified. * One case of skin rash occured in a HCTZ patient (doses not specified). + 25 mg and 50 mg daily doses of HCTZ represented.

TABLE Ill Frequency (%) of Individual Adverse Effects

HCTZ 12.5 HCTZ 25 Placebo mg Daily

Adverse Effect (n = 168) (n = 173)

Abdominal distention 0 11.8 Abnormal urination 3 3 Asthenia 4.9 2.3 Chest pain 3 3 Cough 1-11.8 1.1 Dizziness 1.2-11.8 l-5.9 Edema 5.9 N/R Fatigue 6 3 Headache 7-17.6 5.9-10.3 Musculoskeletal pain 1 1 Nausea/vomiting 4 3 Palpitations N/R 5.9 Pharyngitis 8.6 8.0 Rash 1 1 Stress reaction 1 3 Syncope 5.9 N/R Tinnitus 5.9 N/R

mg Daily (n := 84) --

rm rw 6.0

WR 6.0 4.8

N/R IW! 1.6.7 IN/R N/R N/R 2.4

N/R N/R N/R N/R

N/R = not reported. From references 11-13

dose HCTZ was at least as well tolerated as placebo. The consistent finding of a higher prevalence of

no effect on hepatic insulin action. Low-dose torase-

headache in the placebo group leads one to question mide has been shown to have an effect on lowering

whether essential hypertension is, in fact, a symp- blood pressure similar to that of the thiazides, but

tomless disease. with fewer changes in metabolic parameters than the

Individual clinical experience with low-dose ther- higher-dose torasemide regimen, which is used pri-

apies has generally preceded the published literature marily for treatment of congestive heart failure.

in this area. Most recently, improved study designs In conclusion, the safety profile of low-dose HCTZ

with a more critical approach toward dose ranging is notable for good tolerability and a relative lack of

and use of adequate sample sizes to distinguish be- significant adverse effects. Laboratory parameters

tween dose levels has addressed these concerns. Low- remained largely unchanged, with a trend toward de-

dose regimens have been included in studies of two creasing serum potassium and increasing serum uric

other diuretics: bendrofluazide and torasemide.2”~2g acid levels that was accentuated with increasing

The study of low-dose bendrofluazide demonstrated doses. Overall, low-dose HCTZ was comparable to

an equivalent antihypertensive effect with reduced placebo in terms of clinical adverse effects.

metabolic effects on potassium and serum lipid levels Quality-of-Life Experience in comparison with higher dose regimens. In contrast For years, neither researchers nor physicians re- to conventional doses, low-dose bendrofluazide had ally examined the impact of various antihypertensive

September 30, 1996 The American Journal of Medicine@ Volume 101 (suppl 3fY 3A-87s

regimens on a patient’s quality of life or general well- being. Although studies did look at specific adverse effects and withdrawal from treatment, as a result, general effects on quality of life were often ignored. For instance, issues related to mood, intellectual function, social activity, and energy level were usually not addressed. It was not appreciated that somewhat subtle effects on mood and energy level could have profound effects on quality of life. For instance, we now know that persons who become somewhat depressed and lose energy as a result of antihypertensive treatment may well withdraw from family and friends, decrease social activities, and function less well at work.30

A few years ago, a landmark study definitively showed that use of the ACE inhibitor, captopril, pro- duced less deleterious effects on quality of life than did use of the ,0 blocker propranolol or the adrener- gic agonist a-methyldopa. ” Although all three treatments were equally efficacious in lowering blood pressure, for the first time evidence emerged that certain medications had more detrimental effects on quality of life than did other medications.

A later article by Williams et a,l,32 based on data from the same study, examined the additive effect on quality of life of the diuretic hydrochlorothiazide (25 mg twice daily) as a second-step agent. In this secondary analysis, more patients who had a diuretic added experienced sexual dysfunction and a wors- ening of general well-being, especially in the captopril- and propranolol-treated groups. However, these analyses are very difficult to interpret, because the findings are confounded by the fact that the diuretic was used as a second-step drug and at relatively high doses (50 mg/day).

In a large randomized trial of drugs in African American patients with high blood pressure, Saun- ders and colleagues 33 and Croog and colleaguesX compared the efficacy of a /3 blocker, calcium antagonist, and ACE inhibitor in hypertensive blacks and found differences in blood pressure control, but no differences in quality of life variables. However, no diuretics were used in this study.

The Veterans Affairs collaborative group looked at diuretics as first-step agents, with patients randomized to 50 or 100 mg of hydrochlorothiazide, with other classes of drugs as second-step agents. In low doses, diuretics had few adverse effects on quality of life and relatively minor metabolic effects as well. The VA investigators then compared the efficacy and side-effect profile of various step-two antihypertensive regimens when added to the diuretic. Side effects were more frequent with hydrala- zine and methyldopa than with reserpine or meto- prolol, although none of the drugs studied actually showed any deterioration in quality of life.35 Since a diuretic was uniformly used as the first-step in this VA study, quality-of-life comparisons among the

various second-step drugs could have been confounded by interactive effects between the diuretics and second-step drugs.35,36

These two studies illustrate the problems that are sometimes hidden within averages. Although both the VA study35 and a study by Applegate and co- workers37 showed significant increases, in moderate to severe side effects in some drugs colmpared with others, the more severe side effects, od course, occurred in a minority of patients. Therefo’re, when various treatment groups were compared with regard to average quality-of-life measures, no differential effect was seen on quality-of-life measures between treatment groups even though there were some significant differences in some of the more severe side effects.

Another trial randomized 218 hypertensive patients with a DBP of 95- 114 mm Hg to amlodipine (2.5- 10 mg), bisoprolol(2.5-10 mg) combined with hydrochlorothiazide 6.25 mg, or enalapril (5-20 mg) .38 The bisoprolol/HCTZ and amlodipine groups showed somewhat greater blood-pressure lowering from baseline than did the enalapril group. There were no significant differences in adverse effects among the three treatment groups. Although not al- ways statistically significant, the bisoprolol/HCTZ and amlodipine groups showed a consistent ten- dency toward improvement of quality-of-life scores, whereas the enalapril group tended to have a decline on more quality-of-life variables. Unfortunately, this is not a true examination of the diuretic’s effect on quality-of-life variables, because the investigators examined a low-dose diuretic- p blocker combination compared with other individual drugs. Interestingly, across all three groups, patients with no reported adverse effects tended to have an improvement in general well-being versus a decline in perceived general well-being in those patients reporting adverse events. On the other hand, prior antihypertensive therapy and the magnitude of blood-pressure lowering were not correlated with quality-of-life scores.

An important study published recently regarding the effect of a low-dose diuretic regimen on quahty- of-life variables is from the SHEP trial.3g In this randomized trial, a large cohort of older persons with isolated systolic hypertension were randomized to a chlorthalidone-based regimen (12.5-25 mg) or placebo. By having no active control group, this study provides one of the most accurate evaluations of the effect of a diuretic on quality-of-life variables available in the literature today.

SHEP was a multicenter clinical trial. Men and women aged 260 years with isolated systolic hypertension (n = 4,736) were recruited into this double- blind, placebo-controlled, stepped-care treatment program and were followed for an average of 5 years. The primary endpoint of the trial was fatal and nonfatal stroke. Eligibility was determined1 at two base-

3A-88s September 30, 1996 The American Journal of Medicinem Volume 101 (suppl 3A)

line visits. The participant was eligible for randomization if the average of four blood-pressure readings at the first and second baseline visit was SBP ~160 mm Hg and DBP <90 mm Hg, and if other inclusion and exclusion criteria were met and written in- formed consent was given.

The step 1 drug was chlorthalidone 12.5-25 mg. The step 2 drug was atenolol25-50 mg or reserpine 0.05-0.1 mg if atenolol was not tolerated. In each case participants in the control group were given a matching placebo.

The SHEP Behavioral Evaluation was administered after randomization. Some components were administered semiannually and some annually. The protocol called for the entire SHEP cohort (n = 4,736) to receive part 1 of the Behavioral Evaluation. Part 1 provided periodic evaluation of global cognitive level, depressive symptoms, and physical and social functioning. Part 1 consisted of the SHORT- CARE,40 CES-D,41 Activities of Daily Living,“’ and Social Network questionnaires.43

The SHORT-CARE and CES-D were administered at baseline and 6-month intervals to screen for possible depression or cognitive impairment. If the participant reached or exceeded a clinical cut point on either the depression or dementia scales of the SHORT-CARE, the SHORT-CARE was administered at the next quarterly visit. Participants who reached or exceeded the cut points on either the depression or dementia scales of the SHORT-CARE on two con- secutive visits were referred for the appropriate di- agnostic evaluation.

In addition to part 1, six clinical centers administered part 2 of the Behavioral Evaluation to all of their participants (n = 2,034) at baseline and at an- nual visits. Part 2 included more specific tests of cognitive function, including psychomotor speed, atten- tion, visual scanning, mental calculations, expressive language function, verbal memory, and hypothesis testing. Additionally, part 2 contained questions on quality of life, expression of anger, and leisure activities.44-47

The SHEP study demonstrated that active treatment of isolated systolic hypertension in the SHEP cohort had no measured negative effects on activities of daily life, and, for some measures, a slight positive effect on cognitive, physical, and social function (Tables IV and V) . The positive findings in favor of the treatment group were small. There was no effect on emotional state. Measures of cognitive and emotional function were stable in both groups for the duration of the study. There was a modest trend toward deterioration of some measures of physical function in both treatment groups over the study period (Table V), as might be expected in a long-term study of elderly patients.

Another important study regarding the impact of low-dose diuretic therapy on quality of life was the

TOMHS study.3 TOMHS was a double-blind, lplacebo- controlled trial with an average follow-up of 4.4 years in 902 men and women aged 45-69 years with stage 1 diastolic hypertension (DBP < 100 mm Hg) . Participants were randomized to lifestyle modifications (weight loss, salt and alcohol restrictions, and increased physical activity) plus one of five (drugs or placebo; active drug treatments included chlorthalidone ( 15 mg/day), amlodipine (5 mg/day) , acebutolol (400 mg/day), doxazosin (2 mg/day), and enalapril (5 mg/day) . Quality-of-life measures represented a subset from the larger set used in the Rand Medical Outcomes Study.48 Seven scales rep- resenting four health dimensions (health perceptions, emotional well-being, bodily well being, and functioning) were assessed using a 35-item self-administered questionnaire. Only the chlorthalidone (P = 0.008) and acebutolol (P <O.OOl) groups showed an improvement in the overall test for change in the global quality-of-life score compared with the placebo group.

Sexual Dysfunction The prevalence of sexual dysfunction in hyperten-

sive patients has been greater among men than women; among men, the prevalence of erectile dysfunction increases with advancing age, higher SBP levels, and previous use of antihypertensive medication.4g For unclear reasons, diuretic ther,apy has been more often associated with sexual dysfunction in men than other drug classes, including [3 blockers 4g-‘1 which like diuretics, tend to cause more sexual dysfunction in men than placebo. Male sexual dysfunction has been seen even at low doses ( 15 mg/ day) of chlorthalidone, 4g although most ‘TOMHS men experiencing erectile problems actually continued chlorthalidone treatment. Interestingly, in chlorthalidone-treated men in the Trial of Antihyperten- sive Interventions and Management (TAIM) study, 5o weight loss ameliorated the problem of chlorthalidone-induced male sexual dysfunction.

DISCUSSION Thiazide diuretics have been noted to be highly

effective in reducing blood pressure in many populations, particularly the elderly. During 199 1- 1992, three major intervention trials that dealt with the value of antihypertensive treatment in the elderly- SHEP, the Swedish Trial in Old Patients With Hy- pertension (STOP-Hypertension), and the British Medical Research Council (MRC) Trial of Treatment of Hypertension in Older Adults-were published. These three studies showed that the treatment of hypertension reduced the risk of stroke and cardiovascular events in the elderly. In one of the trials, total mortality was also positively affected. Low-dose thiazides or p blockers, or a combination of the two, were the most commonly used treatments. Further-

September 30, 1996 The American Journal of Medicine” Volume 101 (suppl 3Al 3A-89s


TABLE IV SHEP Results

Score Active Placebo Difference*

n Baseline Last Change (SD) n Baseline Last Change (SD) (95% Cl)

All participants, using last SHORTCARE in study and ignoring events Cognitive impairment 2,317 0.37 0.26 -0.11 (0.90) 2,291 0.37 0.32 -0.06 (1.03) +0.05

(-0.0006, +O.ll) Depressive symptoms 2,264 3.13 2.95 -0.18 (2.90) 2,229 3.11 3.09 -0.02 (3.12) +0.16

i-0.02, iO.34) CES-D score 2,131 4.13 4.33 +0.20 (5.63) 2,117 4.01 4.59 +0.58 16.09) iO.38

(+O.OZ, +0.73) %116 4.6% 5.6% 4.6% 6.9% P = 0.050 last visit

Mean SHORTCARE Cognitive Impairment, Depressive Symptom, and CESD scores at baseline and last evaluation by treatment group. Cl = confidence interval; SD = standard deviation, * Difference = placebo change - active change. (Reprinted with permissron from Arch Intern Med.3q)

TABLE V SHEP Results Activities of Dailv Life (ADL)* bv Treatment GrouD at Baseline and Last Evaluation

(All Participaks, U&g Last ADL in Study and ignoring Events)

Difference

Active (n = 2365) Placebo (n= 2371) deterioration in active vs

Baseline-Needs help Any deterioration Baseline-Needs help Any deterioration placebo Activity or unable to do to last visit or unable to do to last visit grows (PI Any basic ADL 101/2263 (4.5%) 421/2260 (18.6%) 126/2205 (5.7%) 443/2200 (20.1%) 0.20 Any Moderate ADL 248/2232 (11.1%) 450/2040 (22.1%) 267/2154 (12.4%) 461/1968 (23.4%) 0.30

Baseline-At least Any deterioration Baseline-At least Any deterioration some difficultv to last visit some difficultv to last visit

Any Advanced ADL 600/2260 (26.6%) 1070/2203 (48.6%) 593/2196 (27.0%) 1050/2138 (49.1%) 0.72 Any Basic-

Advanced ADL 672/2240 (30.0%) 1215/2172 (55.9%) 690/2175 (31.7%) 1177/2106 (55.9%) 0.97

‘Basic self-care items included bathing, personal grooming, dressing, eating, ambulation, using the toilet and transferring from bed to chair. Moderate ADL included writinflandling small objects, walking up and down stairs, extending the arms above shoulder level, walking l/2 mile. Advanced ADL Included the above plus the ability to perform such tasks as carrying groceries, moving furniture, lifting and carrying weights ~10 pounds, pulling or pushing large objects, or crouchrng or kneeling. (Reprinted with permlssion from Arch Intern Med.?

more, subanalyses indicated that the cost-benefit aspects of such treatment regimens were positive.52

The decision to treat hypertension in elderly patients has been somewhat controversial due to the potential for adverse effects related to treatment to outweigh the benefit of blood pressure reduction. A recent review of trials that used thiazides in selected elderly hypertensive patients concluded that treatment leads to a reduction in both nonfatal stroke and coronary events.53

Another recent review 54 of the morbidity and mortality results from 17 large, controlled clinical trials demonstrated large reductions in the incidence of stroke among study populations that primarily used thaizides to treat more severe hypertension (DBP 2 100 mm Hg) . In patient populations with less severe hypertension, the incidence of both stroke and total mortality were also improved.

Comparative trials of antihypertensive agents have generally not been able to demonstrate clear superiority of one agent over another, in part be-

cause of the lack of comparative trials~ of sufficient duration with an adequate sample size t.o do so. Nev- ertheless, it has been suggested that what matters most is the magnitude of the reduction in blood pressure in these patients, independent of the mecha- nism by which it is achieved.j5 Therefore, although low-dose HCTZ has not been studied longitudinally to the same extent as conventional HCTZ dosing, because the magnitude of blood pressure reduction is similar, the beneficial long-term effects on morbidity and mortality could be expected to be similar.

A recent review of the long-term effects of antihypertensives notes that the literature is “extensive and conflicting” regarding the effects of thiazides and p blockers on serum lipid profiles.j6 A generali- zation can be made, however, regarding the literature on thiazides, i.e., that short-term use of thiazides results in a rise in serum lipid profiles, with a possible very slight fall to no change in high-density lipoprotein levels. If this effect were maintained over a long period of time, it could be expected to be

3A-90s September 30, 1996 The American Journal of Medicine@ Volume 101 (suppI 3A)

associated with an increased risk of cardiovascular disease. However, the literature regarding the long- term effects (>6 months) of thiazides is reasonably clear. Several long-term studies have demonstrated that serum lipid levels return toward baseline during this period of time, 57-5g although this may be the result of the effect of therapeutic dropouts on conventional intention-to-treat analyses. Nevertheless, the thiazide lipid effect is small and is probably not of major consequence in most patients. As with the other laboratory safety measurements, this effect should be minimized since low-dose HCTZ has demonstrated the least effect on serum lipid profiles and, in relative terms, the lipid effects tend to be less with HCTZ than chlorthalidone.

The two definitive studies3p3’ on the effect of a diuretic-based regimen on quality-of-life variables found a slight trend toward benefit on some variables, no consistent evidence of harm, and, in TOMHS,3 a greater improvement in the global quality-of-life score compared with placebo. Although the differences reported in both the Croog et a13r and Applegate et al37 studies (neither of which used a diuretic as the first-line choice) are not overly dra- matic, it is clear that some drugs may have less deleterious effects on overall quality of life and, in par- ticular, on subtle issues related to general well-being and mood. Therefore, physicians in clinical practice should begin to focus more on subtle issues related to quality of life.

The fact that measures of quality of life do not differentiate drug treatment groups with higher side- effect rates (the study by Weir and colleagues3’ is the only potential exception) should come as no sur- prise. Since there are only very modest differences in adverse event rates between randomized comparisons (most of which are not life threatening), the quality-of-life variables are biased against detecting such differences. Thus, small differences in adverse event rates will be hidden in such averages of quality- of-life variables.

Although the literature on the effect of low-dose diuretic on quality-of-life variables is not large, the definitive SHEP study is convincing that diuretics are unlikely to harm the overall quality of life in hypertensive patients. The TOMHS suggests that quality of life actually improves during low-dose chlorthalidone therapy. The most important issues for physicians to examine are the patients’ perceptions of their overall well-being, their overall mood, their energy level, their level of sexual activity, and their level of social interaction. To evaluate these issues, the physician can either use standardized instru- ments or simply question the patient closely in the usual clinical manner. In addition, it is very helpful to question the patient’s spouse with regard to these specific issues. Although most antihypertensive medications are effective to some degree in lowering

blood pressure, it is now incumbent upon physicians also to examine whether the medication regimen is having an adverse effect on subtle issues regarding quality of life.

In summary, the existing clinical data demonstrate that thiazide diuretics have been shown to reduce coronary and vascular events significantly. Com- pared with traditional diuretic doses, low-dose thiazides reduce blood pressure to a similar degree and, therefore, are expected to result in these same im- provements in morbidity and mortality. Finally, low- dose thiazide treatment of hypertension is well-tolerated and efficacious, especially in older patients but should be avoided unless necessary in patients with preexisting diabetes, gout, and in men with erectile dysfunction.

REFERENCES 1. Joint National Committee on Detection, Evaluation, and Treatment of High Blood Pressure. The fifth report of the Joint National Committee orI Detection, Evaluation, and Treatment of High Blood Pressure (JNC-V). Arch Intern Med. 1993;153:154-183. 2. SHEP Cooperative Research Group. Prevention of stroke by anti-hyperten sive drug treatment in older persons with isolated systolic hypertension: final results of the Systolic Hypertension in the Elderly Program (SHEP) JAMA. 1991;265:3255-3264. 3. Neaton JD, Grimm Jr RH, Prineas RJ, et al. Treatment of Mild Hypertension Study. JAMA. 1993;270:713-724. 4. Hart W. Lisinopril-hydrochlorothiazide combination compared with the monocomponents in elderly hypertensive patients. J Hum Hypertens. 1991;5(suppl 2):85-89. 5. Dal Palu C, for the Lisinopril Study Group. Lisinopril versus hyclrochlorothi- azide in mid-to-moderate systolic/diastolic or isolated systolic hypertensron in the elderly. Curr Ther Res. 1993;54:779-787. 6. McKenney JL, Goodman R, Wright JT Jr, Rifai N, Aycock DG, King ME. The effect of low-dose hydrochlorothiazide on blood pressure, serum potassium and kpoproteins. Pharmacotherapy. 1986;6:179-184. 7. Pool JL, Gennari J, Goldstein R, et al. Controlled multi-center !study of the antihypertensive effects of lisinopril, hydrochlorothiazide, and lisinopril plus hydrochlorothiazide in the treatment of 394 patients with mild to moderate essential hypertension, J Cardiovasc Pharmacol. 1987;9tsuppl 3):S36-S42. 8. Rosenthal T, Grossman E, Rathaus M, et al. Treatment of hypertension by enalapril and hydrochlorothiazide separately and together: a multicenter trial. Isr J Med Sci. 1990;26:63-66. 9. Miller E. Introduction to lisinopril-hydrochlorothiazide combina~tion. J Hum Hypertens. 1991;5(suppl 2):49-51. 10. Stein CM, Neil1 P, Kusemamuriwo T. Antihypertensive effects of low doses of hydrochlorothiazide in hypertensive black Zimbabweans. Int J Cardiol. 1992;37:231-235. 11. Mersey J, D’Hemecourt P, Blaze K. Once-daily fixed combination of captopril and hydrochlorothiazide as first line therapy for mild to moderate hypertension. Curr Jher Res. 1993;53:502-512. 12. Fernandez M, Madero R, Gonzalez D, Camacho P, Villalpando J, Arriago J. Combrned versus single effect of lisinopril and hydrochlorothiazide in hypertensive patients. Hypertension. 1994;23tsuppl l):l-207-I-210. 13. Chrysant S, for the Lisinoprrl-Hydrochlorothrazide Group. Antihypertensrve effectiveness of low-dose lisinopril-hydrochlorothiazide combination: a large multicenter study. Arch Intern Med. 1994;154:737-743. 14. Canter D, Frank GJ, Knapp LE, et al. Quinapril and hydrochlorothiazide combination for control of hypertension: assessment by factorial design. J Hum Hypertens. 1994;8:155-162. 15. Scholze J, Breitstadt A, Cairns V, et al, for the East German ICollaborative Trral Group. Short report: ramipril and hydrochlorothiazide combination therapy


September 30, 1996 The American Journal of Medicine@ Volume 101 (suppl 3A) 3A-91s


in hypertension: a clinical trial of factorial design. J’ ffypertens. 1993;11:217- 221. 16. Berglund G, Andersson 0. Low doses of hydrochlorothiazide in hypertension: antihypertensive and metabolic effect. Eur J C/in Pharmacof. 1976;10:177-182. 17. Beermann B, GroschinskyGrind M. Antihypertensive effect of various doses of hydrochlorothiazide and Its relation to the plasma level of the drug. Eur J C/in Pharmacol. 1978;13:195-201. 18. Lorimer AR, Barbour MB, Lawrre TDV. An evaluation of the effect on resting and exercrse blood pressure of some first line treatments in hypertension. Ann Clin Res. 1983;15:30-34. 19. Degnbol B, Dorph S, Marner T. The effects of different diuretics on elevated blood pressure and serum potassium. Acta Med Scand. 1973;193:407-410. 20. Helderman JH, Elahi D, Andersen DK, Raizes GS, Tobin JD, Shocken D, Andres R. Prevention of the glucose intolerance of thiazide diuretics by main- tenance of body potassium. Diabetes. 1983;32:106-111. 21. Kaslske BL, Ma JZ, Kalil RSN Louis TA. Effects of antihypertensive therapy on serum lipids, Ann intern Med. 1995;122:133-141. 22. Cutler JA. Grandits GA, Grimm RH Jr, et al. Risk factor changes after ces- sation of intervention in the Multiple Risk Factor Intervention Trial. The MRFIT Research Group. PrevMed. 1991;20:183-196, 23. Efliott WJ. Glucose and cholesterol elevations during thiazide therapy: in- tentionto-treat versus actual on-therapy experience. Am J Med. 1995;99:261- 269. 24. Weidmann P, Uehlinger DE, Gerber A. Antihypertensive agents on serum lipids. Arch Intern Med. 1988;148:1280-1287. 25. Grimm RH Jr, Flack JM, Grandits GA, et al. Long-term effects on plasma lipids of diet and drugs to treat hypertension. JAMA. 1996;275:1549-1556. 26. Grimm RG, Leon AS, Hunninghake DB, et al. Effect of thiazide diuretics on blood lipids and lipoproteins in mildly hypertensive patients: a double-blind controlled trial. Ann Intern Med. 1981;94:7-Il. 27. Ames RP, Hill P. Increase in serum lipids during treatment of hypertension with chlorthalidone. Lancet. 1976;i:721-723. 28. Achhammer 1, Metz P. Low dose loop druretics in essential hypertension: experience with torasemide. Drugs. 1991;41 tsuppl3);80-91. 29. HarperR, Ennis CN, Sheridan B, Atkinson AB, Johnston GO. Bell PM. Effects of low dose versus conventional dose thiazide diuretics on insulin action in essential hypertension. Br Med J. 1994;309:226-230. 30. Levine S, Croog SH. What constitutes quality of life? A conceptualization of the dimensions of life quality in healthy populations and patients with cardiovascular disease. In: Wenger NK, Mattson ME, Furberg CD, Elison J, eds. As- sessment of Quality of Life. New York: Le Jacq Publishing, 1984;46:66. 31. Croog SH, Levine S, Testa MT, et al. The effect of antihypertensive therapy on quality of life. N Engl J Med. 1986;314:1657-1664. 32. Williams GH, Croog SH, Levine S, Testa MA, Sudilovsky A. Impact of antihypertensive therapy on quality of life: effect of hydrochlorothiazide. J Hyper- tension. 1987;5tsuppI):S29-S35. 33. Saunders E, Weir MR, Kong W, et al. A comparison of the efficacy and safety of a beta-blocker, a calcium antagonist, and a converting enzyme inhibitor in hypertensive blacks. Arch Intern Med. 1990;150:1707-1713. 34. Croog SH, Kong BW, Levine S, et al. Hypertensive black men and women: quality of life and effects of antihypertensive medications. Arch Intern Med. 1990;150:1733-1741. 35. Goldstein G, Matterson BJ, Cushman WC, et al, for the Department of Veterans Affairs Cooperative Study. Treatment of hypertension in the elderly: II. Cognitive and behavioral function. Hypertension. 1990;15:361-369. 36. Materson BJ, Cushman WC, Goldstein G, et al, for the Department of Vet- erans Affairs Cooperative Study. Treatment of hypertension in the elderly: i. Blood pressure and clinical changes. Hypertension. 1990;15:348-360. 37. Applegate WB, Phillips HL, Schnaper H, et al. A randomized controlled trial of the effects of three antihypertensive agents on blood pressure con-

trol and quakty of fife in older females. Arch Intern Med. 1991:151;1817- 1823. 38. Weir MR, Prisant LM, Papademetriou V, et al. Antihypertensive therapy and quality of life: influence of blood pressure reduction, adverse effects, and prior antihypertensive therapy. Am J Hypertens. 1996. In press. 39. Applegate WB, Pressel S, Wittes J, et al. Impact of the treatment of isolated systolic hypertension on behavioral variables: the SHEP study. Arch Intern Med.

1994;154:2154-2160. 40. Gurland B, Golden R, Challop J. Unrdrmensronal and mulbdimensional ap proaches to the differentiahon of depression and dementia in the elderly, In: Corkin S, Davis KI, Crowden JH, Usdin E, Wurtman RJ, eds. Alzheimer’s Disease: A Report of Progress in Research. New York: Raven Press, 1981. 41. Radloff LS. The CES-0 Scale. Appi Psycho! Meas. 1977:1:385-401. 42. Jette AM, Branch LG. The Framingham Disabikty Study: physical drsability amongtheaged.AmJPubllfealth. 1981;71:1211-1216. 43. Berkman LF. The assessment of networks and social support in the elderly. J Am Geriatr Sot. 1983;12:743-749. 44. Weiler PG, Camel GH, Chiappiani M, et al. Behavioral characteristics. Hy- pertension. 1991;17:11-152-I-161. 45. Katz S, Ford AB, Moskovitz RW, et al. Studies of illness In the aged: the index of ADL: a standardized measure of biological and psychologrcal function, JAMA. 1963;815:94-98. 46. Nagi SZ. An epidemiology of disability among adults III the United States. Milbank Memorial Fund Q. 1976;54:439-468. 47. Rosow I, Breslau H. A Guttman health scale for thsa aged. J Gerontol. 1966;21:556-559, 48. Tarlov AR, Ware JE. Greenfield S, et al. The Medical outcomes study: an application of methods for monitoring the results of medical care, JAMA. 1989;262:925-930. 49. Grimm RH Jr, Grandits GA, Prineas RJ, et al. Long-term effects on sexual function of five antihypertenslve drugs and nutrrhonal-hygtssnic treatment rn hypertensive men and women: Treatment of Mild Hyperten’sion Study (TOMHS) Hypertension 1996;in press. 50. Wassertheil-Smaller, Blaufax MD, Oberman A, et al. Elfect of antihypertensives on sexual function and quality of Irfe: the TAIM Study. Ann Intern Med. 1991;114:613-620. 51. Medical Research Council Working Party. Adverse reactions to bendrofluazide and propranolol for the treatment of mtld hyisertensron. Lancer 1981;11:539-543 52. Hansson L. Future goals for the treatment of hypertension in the elderly with reference to STOP-Hypertension, SHEP, and the MRC Trial in Older Adults, Am J Hypertens. 1993;6:4OS-435. 53. Kendall MJ, Tse WY, Head A. The treatment of elderly hypertensive patients. J C/in Pharm Jher. 1993;18:9-14. 54. Cutler JA, MacMahon SW, Furberg CD. Controlled clinrcal trials of drug treatment for hypertension: a review. Hypertension. 1989;13tsuppll):136-144. 55. Hampton JR. Comparative efficacy of diuretics: benefit versus risk: results of clinical trials. Eur Heart J. 1992;13(suppl G):85-91. 56. Ballantyne D. Long-term effects of antihypertensives OII blood lipids. J Hum

Hypertens. 1990;4(suppl2):35-37. 57. Amery A, Birkenhager W, Bulpitt C. et al. influence of antihypertensrve therapy on serum cholesterol In elderly hypertensive patients: results of trial by the European Working Party on High Blood Pressure in the Elderly (EWPHE). Acta Cardiol. 1982;37:235-244. 58. Curb JD, Maxwell MH, Schneider KA, Taylor JO, Shulman NB. Adverse effects of anti-hypertensive medications in the Hypertension Detection and Fok low-up Program. Prog Cardiovasc Dis. 1986;29(suppl 1):73-88. 59. Veterans Administration Cooperative Study Group on Antrhypertenave Agents. Comparison of propranolol and hydrochlorothtazide for the inrtial treatment of hypertension: II, Results of long-term therapy. JAM4. 1982;248:2004- 2011.

3A-92s September 30, 1996 The American Journal of Medicine* Volume 101 fsuppl 3A)

tolerability, safety, and quality of life and hypertensive therapy: the case for low-dose diuretics

Documents