SYSTEMIC HYPERTENSION Age and Antihypertensive Drugs (Hydrochlorothiazide, Bendroflumethiazide, Nadolol and Captopril) EDWARD D. FREIS, MD, for the VETERANS ADMINISTRATION COOPERATIVE STUDY GROUP on ANTIHYPERTENSIVE AGENTS’ 25 to 100 mg twice daily to low- er diastolic blood pressure (BP) to <90 mm Hg. Of 121 patients 65 the decrease in BP ay- Hg, while in the 191 pa- reduction averaged ypertension in the elderly has a different hemody- namic profile than in the young.’ It characteristically has a higher systolic blood pressure (BP) in relation to the diastolic level. This is due mainly to increased stiffness of the large arteries secondary to arterioscle- tosis. Also, cardiac output, heart rate, stroke volume, blood volume, renal blood flow and plasma renin ac- tivity are lower than in yourger hypertensive patients. The prevalence of hypertension increases with age. According to the National Health Survey 34% of the From the Cooperative Studies Program, Medical Research Ser- vice, Veterans Administration, Washington, D.C. Manuscript received June 4, 1987; revised manuscript received September 11, 1987, and accepted September 12. Address for reprints: Edward D. Freis, MD, Hypertension Research, Veterans Administration Medical Center, 50 Irving Street, Northwest, Washington, D.C. 20422. *See the Appendix for principal participants. wT (Am J Cardiol 1988;61:117-121) population aged 55 to 65 years have a BP > 160 systolic, 295 mm Hg diastolic, or both? Further, the elderly constitute an increasing percentage of the total popula- tion. Therefore, hypertension in the elderly is a major public health problem. Because of the many changes that occur with aging including differences in hemo- dynamics, elderly patients may respond differently than younger patients to antihypertensive drugs.} For these reasons there has been considerable interest re- cently in the epidemiology*-5 and treatment*’ of hy- pertension in the elderly. The Veterans Administration Cooperative Study Group has carried out controlled trials of a number of antihypertensive drugs. These trials have included pa- tients of various ages permitting comparison under controlled conditions of age-related responses to a va- riety of antihypertensive drugs. Age 55 was chosen as the line dividing young from old because it permitted a larger sample in the older age group. 118 AGE AND ANTIHYPERTENSIVE ORUGS TABLE! Bicod Pressure Changes Alter Hydrochiorothiazide or Propranolol by Age Change in Blood Pressure (mm Hg) Age <55 Age 55 to 65 Systolic Diastolic Systotic Diastotic No. Gase Change Sase Change No. Base Change Sase Change Hydrochiorothiazide Blacks 112 143 —~17 101 -13 59 156 —26! 101 14 Whites 79 142 —14 102 -10 62 152 —- 18° 101 —12° Ait 194 142 —18 101 —-12 121 154 —2at 104 —~13¢ Propranoiol Blacks 119 (141 -8 101 -9 57 148 -9 101 -10 Whites 78 148 —14 103 —-12 50 181 —-12 102 ~13 All 191 143 —-10 101 —11 107. 18011 101 -" Comparing bicod pressure changes for age <55 vs 255. °9 <0.05. *p <0.001. Hydrochlorothiazide Versus Propranolol Hydrochlorothiazide: The first study presents age- related data from a trial comparing hydrochlorothia- zide with propranolol given as monotherapy. The trial included nonhospitalized male veterans aged 21 to 65 years with untreated diastolic BP between 95 and 114 mm Hg. After a single-blind placebo period of 4 weeks, patients with placebo counts indicating compli- ance and with diastolic BP in the acceptable range, were randomized to receive either propranolol or hy- drochlorothiazide given double-blind. This was fol- lowed by a 10-week dose-finding period. Doses were increased until the diastolic BP decreased to <90 mm Hg (goal BP) or the maximum dose was reached or side effects intervened. There were significant age-related effects associ- ated with hydrochlorothiazide. The drug was titrat- ed from 25 mg twice daily (50 mg/day), increasing until the diastolic BP remained below 90 mm Hg on 2 successive visits or until the maximum dose of 100 mg twice daily (200 mg/day} had been given. There were 121 patients aged 55 to 65 and 191 patients younger than 55 years (Table 1). In those aged 55 to 65 the average reduction of BP was —21.8/—12.9 mm Hg (systolic/diastolic). Both the systolic and the dia- stolic BP decreases were significantly greater than in patients younger than 55 years. In the latter group, the BP decreased by ~15.7/-11.5 mg Hg (p <0.001; p = 0.048). With respect to age and race, the age-related reduc- tions were greater in biacks only with regard to systolic BP, the systolic reduction averaging —26.2 mm Hg in those aged 55 to 65 years and -17.2 mm Hg in patients younger than 55 years (p <0.001). In whites both the systolic and diastolic reductions were significantly greater in older patients. In the group of whites aged 55 to 65, the reductions averaged —17.5/-12.1 mm Hg while those younger than 55 years the reductions were —13.6/~9.9 mm Hg (p = 0.052; p = 0.018). The racial differences were especially prominent in both age groups with significantly greater decreases, especially in systolic BP in blacks compared with whites in both age groups. With respect to the percent of patients achieving goal diastolic BP <90 mm Hg with hydrochlorothia- zide there also was a significant difference related to age (Table II). Among the patients aged 55 to 65, 72% attained goal BP compared with 59% of the patients younger than 55 years (p = 0.03). When the groups were subdivided by race there was still a strong trend indicating a greater percent reaching goal BP with hy- drochlorothiazide among the older patients in both racial groups. Unlike the change in average BP the difference in the percent reaching goal BP did not quite reach the 0.05 level of significance. This may have been due to the relatively small sample sizes resulting from both race and age subdivisions. Among whites 65% of those aged 55 to 65 years attained a diastolic BP <90 mm Hg compared with 48% of pa- tients younger than 55 years (p = 0.08}. As expected. blacks had a greater response rate than whites with 80% responders in the older age group and 67% among the younger patients (p = 0.12]. Propranolol: An additional 298 patients were ran- domly assigned in a double-blind study to propranolol. Of these 107 were aged 55 to 65 years and 191 were below this age. Doses of the drug were titrated from 40 mg twice daily to 320 mg twice daily as need- ed to achieve goal BP. Reduction of BP averaged —10.6/-11.4 mm Hg in the group aged 55 to 65 years compared with -10.3/—10.5 mm Hg in the patients younger than 55 years (Table I). The differences be- tween these 2 age groups were not significant. There also were no age-related differences when blacks and whites were analyzed separately. However, there were significant age-related differences in response to propranolol when white patients aged 260 years were compared with the younger patients. Their systolic BP reduction averaged only —5.4 mm Hg compared with —14.9 mm Hg in patients aged <60 years (p <0.001). The diastolic differences were in the same direction but were not significant. Blacks showed no age-related differences to propranolol in this age group. Dose requirements: Age-related responsiveness !0 propranolol and hydrochlorothiazide also was evalu- ated in terms of the dose of drug required to achieve January 1.1988 THE AMERICAN JOURNAL OF CARDIOLOGY TABLE Percent Attaining Diastolic Blood Pressure <90 mm Hg (Goal Blood Pressure) with Propranotel or Hydrochiorothiazide Above end Below Age 55 Years Volurne 41 118 TABLE = Change In Blood Pressure (BP) After Treatment with Hydrochiorothiazide, Captopril or the Combination According to Age Age <55 Age 55 to 65 Attaining Attaining No. Goal 8P(%) No. Goal BP (%) p Value Hydrochiorothiazide Blacks 112 87 59 80 0.12 Whites 79 48 62 65 0.08 All 191 59 121 72 0.03 Propranciol Blacks 113 54 57 53 1.00 Whites 78 56 50 70 0.18 all 191 55 107 61 0.40 goal BP. As an index of responsiveness the doses of hydrochlorothiazide were divided into low and high dose ranges; the low dose consisted of 50 or 100 mg and the high dose 200 mg/day. With propranolol the low dose range was 80 to 240 mg and the high dose 320 to 640/day. In whites only there was a significant difference with age in the percentage of responders to low doses of hydrochlorothiazide. In those aged 55 to 65 years 83% responded to the low doses compared with a 53% response rate to the same dose in those younger than 55 years (p = 0.01). There were no significant age- related differences in the dose responsiveness to propranolol. Medical terminations and age: Medical termina- tions were not significantly related to age. Termina- tions were divided into medical and administrative causes. Medical terminations included such events as major cardiovascular complications and elevated BP. During the 14-week titration period the percent of hydrochlorothiazide-treated patients with medical terminations was 1.0% in patients younger than 55 and 3.0% in those 55 to 65 years of age (p = 0.16 by chi- square test), which was an insignificant difference. Medical terminations in the propranolol-treated pa- tients were 4.2% in those younger than 55 years and 6.5% in those 55 to 65 (p = 0.37). Nadolol and Bendroflumethiazide Another Veterans Administration Cooperative Study® was carried out comparing nadolol (80 to 240 mg/day) with bendroflumethiazide (5 mg increased to 10 mg/day) or the combination of the 2 drugs. The number of randomized patients was 365 and the entry BP was in the range of 95 to 114 mm Hg. The age range was 20 to 69 years (mean 50.5). The average pretreat- ment diastolic BP in the group younger than 55 years and the group 55 to 69 years was similar. In the patients who received bendroflumethiazide alone the decrease in BP averaged —20.5/-11.7 mm Hg in the patients aged 55 to 69 and —14.8/-11.6 mm Hg in the group younger than 55. The difference in the systolic reduc- tion between the older and the younger patients was of borderline significance (p = 0.10, Student t test). Treatment Age <55 Age 55 to 69 Pp Value Hydrochiorothiazide (n= 28) (n = 43) Systolic BP 12 —10 0.56 Diastotic BP -7 -9 039 Captopril aione* (n = 35) (n= 65) Systolic BP ~10 —10 08g Diastolic BP -9 -10 073 Captopril plus (n = 85) (n = 127} hydrochiorothiazide Systolic BP —21 —28 0.0015 Dlastolic BP -16 ~—16 0.69 * Represents one-third of captopril patients who did not have hydrochioroth+- azide added. Nadolol alone resulted in no difference in the de- gree of BP reduction between the 2 age groups. The average reduction for both groups combined was ~10.6/-12.3 mm Hg. A significant age-related differ- ence in systolic BP was observed in the patients who received the combination of the diuretic and the 3 blocker. Patients younger than 55 years showed a sys- tolic reduction of —22.4 while those aged 55 to 69 years averaged ~29.2 mm Hg systolic BP reduction (p = 0.01). Despite the enhanced response, hypotensive side effects were not reported. Captopril With and Without a Diuretic The third study compared the antihypertensive ef- fects of various doses of captopril both alone and with a diuretic’® in patients with uncomplicated hyperten- sion. The patients were men age 20 to 69 years (mean 55.2) whose diastolic BP during the placebo run-in pe- riod was in the range of 92 to 109 mm Hg on 2 succes- sive clinic visits. The regimens were randomly as- signed in parallel and double-blind manner. After 7 weeks of captopril alone hydrochlorothiazide, 25 mg twice daily, was added double-blind in two-thirds of the patients. Of the 421 patients randomly assigned in the trial, 166 were younger than 55 while 255 were aged 55 to 69 years. The average baseline diastolic BP in the 2 age groups <55 and 55 to 69 years was 98.4 and 97.3 mm Hg, respectively. With captopril alone the reductions of BP averaged —10.3/—10.0 mm Hg in the group aged 55 to 69 and —9.8/—9.3 mm Hg in patients younger than 55 years (Table III]. There was no age-related difference in the percent of patients whose diastolic BP decreased to <90 mm Hg compared with those who did not. There were 16% terminations in the group younger than 55 and 11% in those 55 to 69 years. After the addition of hydrochlorothiazide there was a definite further decrease in BP. The reduction from pretreatment baseline was greater in the group 55 to 69 years of age: it averaged —27.5/-16.2 mm Hg com- pared with —20.7/—15.7 mm Hg in the younger group. The systolic difference from baseline was significant (p <0.0015). 120 AGE AND ANTIHYPERTENSIVE ORUGS A multiple regression analysis of factors predicting changes in diastolic BP with captopril alone indicated that age as a continuous variable contributed very little to the model predicting BP change from baseline to the eighth week of treatment. It accounted for less than 2% of the variability of the BP change. With captopril plus diuretic there was a trend suggesting that the combina- tion was somewhat more effective in older patients. Discussion These data demonstrate that age affects the antihy- pertensive responsiveness to some drugs and not to others. Elderly patients, both black and white, exhibit a significantly greater antihypertensive response to hydrochlorothiazide than younger subjects but this did not occur with @-adrenergic blocking drugs. The older age group in these studies was between the ages of 55 and either 65 or 69. The very old were excluded. Most patients with isolated systolic hyper- tension were not included. The group aged 55 to 65 represents an intermediate group between middle- aged and very old patients. They comprise a very large and important proportion of the hypertensive popula- tion.? Comparing those older with those younger than 55, the percent of patients receiving hydrochlorothia- zide who achieved diastolic BP <90 mm Hg was great- er in the older age group. The older white patients also required a lower average dose of hydrochlorothiazide to achieve goal BP. As noted in previous trials there was a significant racial influence: blacks exhibited a greater antihypertensive response to diuretics than did whites, ®2.11 There were no age-related differences with pro- pranolol (Table I). Most other studies agree with the present results, that there is no correlation be- tween age and the response to 8-adrenergic blocking drugs.'*-'* However, when we compared only white patients aged 60 to 65 with the patients younger than 60 years there was a significant blunting of the antihy per- tensive response to propranolol among the older pa- tients. This diminished response in patients aged 60 to 65 years confirms the observations of Buhler et ai," who found that only 20% of these patients achieved a diastolic BP <90 mm Hg compared with 50% of pa- tients 40 to 60 years of age. The diuretic bendroflumethiazide, administered alone, was also more effective in patients aged 55 to 69 years compared with those younger than 55 years, al- though the difference reached only borderline (p = 0.10) significance. Also, the combination of diuretic plus 8 blockade resulted in a significantly greater de- crease in systolic BP in the older patients compared with the younger group (p = 0.01). Nadolol alone failed to show a significant difference in BP related to age. An age-related effect of diuretics also was found in the third study involving captopril and hydrochlor- othiazide. When captopril alone was administered there was no significant age-related difference in BP. When hydrochlorothiazide was added, the response was significantly greater in the older age group. There- fore, in all 3 trials the antihypertensive effectiveness of the diuretic was greater in the older patients. The Hypertension Detection and Follow-up Pro. gram,'! which prescribed diuretics as primary treat- ment also found an enhanced antihypertensive re- sponse in the older age groups compared with younger patients. Not only was the average BP lower but there also was a greater reduction in mortality in the step care compared with the referred care patients in the older age groups. The European Working Party on High Blood Pres- sure in the Elderly’* found a significant decrease in BP in patients older than 60, which averaged —20/-8 mm Hg lower than the placebo control subjects. These pa- tients were treated with 25 mg hydrochlorothiazide plus 50 mg triamterene daily, with a methyldopa add- ed when needed. Side effects in some patients consist- ed of increases in creatinine and uric acid, but not in cholesterol, and impairment of glucose tolerance after 2 years of treatment. In general, however, the medica- tions were well tolerated. In the present studies drug-related symptoms in- cluding mental symptoms were not significantly great- er in the group aged 55 to 65 years than in the younger patients. Lindeman et al’” administered hydrochloro- thiazide-reserpine treatment to elderly patients in a placebo-controlled double-blind trial. They failed to find any changes in psychological tests or in the elec- troencephalogram before and after treatment despite considerable decreases in BP in some patients. In 14 patients the reductions in BP were considerable. aver- aging ~51 mm Hg systolic and —-21 mm Hg diastolic. Elderly patients may develop hypotensive re- actions when treated with certain antihypertensive agents. For example, among the drugs used in 1955. Freis!® found that reserpine, veratrum and hydrala- zine were better tolerated than pentolinium, a gangli- on blocking drug. The latter resulted in more hypoten- sive symptoms, particularly postural hypotension in the elderly. Jackson et al!? warned of the dangers of treating elderly patients. They reported 6 patients aged 64 to 84 who were usually treated with a thiazide pius methyldopa. The side effects consisted mostly of hypo- tensive symptoms. Many of these side effects, including hypotensive side effects in the elderly, can be avoided by initiating treatment with half doses of antihypertensive drugs. These doses are then increased gradually as needed until the desired response occurs. Antihypertensive reactions were not seen in the present trials nor were they observed in the Australian trial on the treatment of mild hypertension in the elderly.® The latter study used 300 to 1,000 mg/day of chlorothiazide as initial therapy followed by step 2 drugs as needed. Hemodynamic changes that may affect the antihy- pertensive responses with aging include reductions in cardiac output, heart rate, stroke volume, intravascular volume, renal blood flow and plasma renin activity.’ Plasma volume appears to correlate inversely with to- tal peripheral resistance, which is increased. Because plasma volume is already somewhat reduced in the January 1.1988 THE AMERICAN JOURNAL OF CARDIOLOGY elderly, the further reduction induced by thiazides may contribute to the enhanced antihypertensive re- sponse found in the elderly. For any given level of BP elderly hypertensive patients exhibit more hypertro- phy of the left ventricle than do younger patients. Re- nal blood flow is reduced. One of the most important changes with aging with respect to cardiovascular con- trol is the alteration in the baroreceptors.2° With in- creasing age and associated increased stiffness of the carotid arteries the baroreceptors become less effec- tive in modifying BP changes including the reductions induced by antihypertensive drugs including ortho- static hypotension. It can be concluded that antihypertensive drug therapy is effective and well tolerated in elderly pa- tients with uncomplicated hypertension. However, reasonable precautions should be taken when pre- scribing thiazide diuretics in the elderly because of the increased antihypertensive responsiveness of such pa- tients to these agents. In.elderly patients diuretic treat- ment should probably begin with half doses and then increased as needed. Additional studies are needed to determine if elderly patients with cardiovascular com- plications, such as old myocardial infarction, stroke or congestive heart failure, can tolerate drug treatment as well as the present patients who had no complications. Also needed are studies of responses to antihyperten- sive drugs in patients older than 65 or 69 years and especially in those with isolated systolic hypertension. References 1. Messerli FH, Sundgaard-Ruse K, Ventura HO. Dunn FG, Glade LB, Froh- lich ED. Essential hypertension in the elderly: haemodynamics, intravascular volume, plasma renin activity and circulating catecholamine leveis. Lancet 1983;2:983- 986. 2. U.S. Department of Health and Human Services. Blood pressure levels in persons 18-74 years of age in 1976-1980, and trends in blood pressure from 1960 to 1980 in the United States. Vital and Health Statistics, series [[, No. 234, fuly 1986:37. 3. Roine JW. Systolic hypertension in the elderly. N Engl | Med 1983:309: 1246-1247. 4. Kannel WB, Dawber TR, McGee DL. Perspectives in systolic hypertension. The Framingham study. Circulation 1980:61:1179-1182. 5. Garland C. Barrett-Conner E, Suarez L. Criqui MH. Isolated systolic hyper- tension and mortality after age 60 years. Am | Epidemioi 1983;118:365-376 8. Management Committee of the Australian Therapeutic Trial in Mild Hy- pertension. Treatment af mild hypertension in the elderly. Med | Aust 1981:2:398-402. 7. Curb JD, Borhani NO. Schnaper H, Kass E. Entwistle G. Williams W, Bernan R. Detection and treatment of hypertension in older individuais: the HDFP. Am J Epidemiol 1985;121:371-378. 8. Veterans Administration Cooperative Study Group on Antihypertensive Agents. Comparison of propranojol and hydrochiorothiazide for the initial treatment of hypertension. I. Results of short-term titration with emphasis an racial differences is response. [AMA 1982:248:1996-2003. $. Veterans Administration Cooperative Study Group on Antihypertensive Agents. Efficacy of nadolol alone and combined with bendroflumethiazide and hydralazine for systemic hypertension. Am | Cardiol 1983:52:1230-1237. 10, Veterans Administration Cooperative Study Group on Antihypertensive Votume 61 121 Agents. Low dose captopril for the treatment of mild to modercte hy perten- sion. Hypertension 1983:5.[11-139-(I-144. 11. Hypertension Detection and Foilaw-up Program Cooperative Group Five-year findings of the hypertension detection and follow-up program [1 Mortality by race, sex and age. [AMA 1979;242:2572-2577 12. Hansson L. Beta-adrenergic blockade in essential hypertension. Effects of propranolol on hemodynamic parameters and plasma renin activity Acta Med Scand 1973:194{supp! 550|:5-38. 13. Menard |, Bertagne X, N’'Guyen PT, Degaulet P. Carval P Rapid identifi- cation of patients with essential hypertension sensitivity to acebutoloi {a new cardio-selective beta-blocker). Am | Med 1976,;60:886-890. 14. Woods |W. Pettman AW, Pulliam CC. Werk EE Jr. Waider W. Allen CA. Renin profiling in hypertension and its use in treatment with propranolol and chiorthalidone. N Engi | Med 1976:294:1137- 1143. 15. Buhler FR. Burkert K. Lutoid BE. Kung M. Morbet G. Pfisteren M. Antihv- pertensive pathogenetic mechanisms in essential hypertension. Am { Cardiol 1975:36:653-669. 16. Amery A. Birkenhager W. Bogaert M. Brixko P, Bulpitt C. Clement D, Le- Leeuw P, De-Plaen |F, Deruyttere M. De-Schaepdryuer A. Fagard R. Forette F, Forte j, Hamdy R, Hellemans j, Henry |F. Koistinen A. Laaser U. Laher M. Leonetti G. Lewis P, Lund-fohansen P. MacFarlane J, Meurer K. Miguel P, Morris |, Mutsers A, Nissinen A. O’Brien E. Ohm Of, O'Malley K. Pelemans W, Perera N, Tuomilehto |, Verschueren L], Willemse P. Williams 8. Zan- chettl A. Antihypertensive therapy in patients above age 60 with svstolic hypertension. A progress report of the European Working Party on High Blood Pressure in the Elderly (EWPHE). Clin Exp Hypertens 1982:4.1151- 1176. 17. Lindeman RD. Buthilet GN. Ashley WR. Morris {R Jr. Effect of hydrochio- rothiazide-reserpine therapy on cerebral function in elderiy hypertensive patients. { Am Geriatr Soc 1963:11:597-605. 18. Frets ED. Treatment of the elderly hypertensive patient. | Am Geriatr Soc 1955;3:7 10-713. 19. Jackson G. Mahon E. Pierscianowski TA, Condon J. [Inappropriate antihy- pertensive therapy in the elderiy. Lancet 1976:2:1317-1318. 20. Gribbin 8, Pickering TG, Sleight P, Peto R. Effect of age and high blood pressure on baroreflex sensitivity in man. Cire Res 1971,29:424-431 Appendix—Principal Participants Cochairmen: Edward D. Freis, MD (Washington. DC) and Barry J. Materson, MD (Miami, Florida). Principal Investigators: William C. Cushman. \{D (Jackson, Mississippi}; Orlando Fernandez, MD, and Khin Mae Hla, MD (Miami, Florida): Ibrahim Khatri, MD (Washington, DC): Eli A. Ramirez, MD (San Juan. Puerto Rico); Harold W. Schnaper. MD (Birmingham. Alabama): Fredrick N. Talmer. MD (Allen Park, \fich- igan); Thomas |. White, MD (Memphis. Tennessee!: Richard Borreson, MID (Houston. Texas}; Walter Fla- menbaum, MD, and Robert Hamburger, \{D (Boston. Massachusetts); Kalman C. Mezey, MD {East Orange. New Jersey}; Bangshe Mukharji. MD (Topeka, Kan- sas}: William W. Neal. MD (Dallas, Texas): H. Mitchel Perry, MD (St. Louis, Missouri}; James T. Taguchi. MD (Dayton, Ohio}. Biostatisticians: Sharon Anderson. PhD, Thomas Tosch, PhD, Jane Foregger (Hines, Illinois}; Bor Ming Ou, MS, Samuel G. Lindle. PhD, Domenic Reda, MS. Susan Fisher, MS.