Journal of Human Hypertension (2006) 20, 177–185 & 2006 Nature Publishing Group All rights reserved 0950-9240/06 $30.00 www.nature.com/jhh ORIGINAL ARTICLE Effect of telmisartan/hydrochlorothiazide vs lisinopril/hydrochlorothiazide combination on ambulatory blood pressure and cognitive function in elderly hypertensive patients

R Fogari, A Mugellini, A Zoppi, P Lazzari, M Destro, A Rinaldi and P Preti Department of Internal Medicine and Therapeutics, Clinica Medica II, IRCCS Policlinico S. Matteo, University of Pavia, Pavia, Italy

The aim of this study was to compare the effects cantly reduced ambulatory BP. However, the telmisar- of telmisartan/hydrochlorothiazide (HCTZ) vs lisinopril/ tan/HCTZ combination produced a greater reduction in HCTZ combination on ambulatory blood pressure and 24-h, day-time and night time ABPM values. Lisinopril/ cognitive function in elderly hypertensive patients. A HCTZ did not induce significant changes in any of the total of 160 patients, 76 men and 84 women, aged 61–75 cognitive function test scores at any time of the study, years, with sitting diastolic blood pressure (DBP) whereas at both 12 and 24 weeks telmisartan/HCTZ 490 mmHg and o110 mmHg and systolic blood pres- significantly improved the word-list memory score sure (SBP) 4140 mmHg were randomized to receive ( þ 17.1 and þ 15.7%, respectively, Po0.05 vs baseline), temisartan 80 mg/HCTZ 12.5 mg o.d. or lisinopril the word-list recall score ( þ 13.5 and þ 16.9%, Po0.05) 20 mg/HCTZ 12.5 mg o.d. for 24 weeks, according to and the Trails B score (À33 and À30.5%, Po0.05). These a prospective, open-label, blinded end point, parallel- results suggest that in elderly hypertensive patients group design. At the end of a 2-week wash-out period treatment with telmisartan/HCTZ produces a slightly and after 12 and 24 weeks of active treatment, 24-h greater reduction in ambulatory BP than lisinopril/HCTZ noninvasive ambulatory BP monitoring (ABPM) was combination and, unlike this latter, improves some of performed and cognitive function was evaluated the components of cognitive function, particularly through six different tests (verbal fluency, Boston episodic memory and visuospatial abilities. naming test, word-list memory, word-list recall, word- Journal of Human Hypertension (2006) 20, 177–185. list recognition and Trails B). Both treatments signifi- doi:10.1038/sj.jhh.1001964; published online 24 November 2005

Keywords: telmisartan/HCTZ; lisinopril/HCTZ; hypertensive elderly; cognitive functions

Introduction Growing evidence has suggested that the renin– angiotensin system (RAS) in the central nervous Longitudinal and epidemiological studies have system participates in the processing of sensory shown a relationship between high blood pressure 9,10 1–3 information, learning and memory. Brain angio- (BP) and cognitive impairment later in life. This tensin II (Ang II) and its fragments Ang IV and Ang suggests that, in addition to being a powerful risk III have been shown to influence performance in factor for cardiovascular disease, hypertension in learning and memory paradigms directly or by elderly patients increases the risk of decline in modulating the activity of other neurotransmitters, cognitive function. Active lowering of BP should like acetylcholine, catecholamines, serotonin and prevent or reduce cognitive impairment, although other peptides.11–13 Beyond being implicated in not all antihypertensive agents seem to be equally 4–8 neuronal functions regulation, these neurotransmit- effective in this regard. ters also regulate local vasomotility and endothe- lium-dependent relaxation, which is of relevance for maintaining the physiological self-regulation Correspondence: Dr R Fogari, Department of Internal Medicine of brain flow, that is presumed to be necessary to and Therapeutics, Clinica Medica II, IRCCS Policlinico S. Matteo, correct performance of cerebral functions.14 Owing University of Pavia, Piazzale Golgi 19, 27100 Pavia, Italy. to the role of the RAS in modulating cognitive E-mail: [email protected] Received 21 July 2005; revised 27 September 2005; accepted 1 processes, drugs that interfere with the RAS like October 2005; published online 24 November 2005 angiotensin-converting enzyme inhibitors (ACE-I) Antihypertensive treatment and cognitive functions in the elderly R Fogari et al 178 and Ang II AT1 receptor blockers (ARB) possess Patients were checked at the end of the wash-out the potential of preventing cognitive decline and period and after 12 and 24 weeks of active treatment. even improving cognitive function in hypertensive At each visit, 24-h ambulatory BP monitoring patients. Indeed experimental and clinical studies (ABPM) was performed, using a portable, noninva- have shown that these drugs may influence cogni- sive, fully automatic BP recorder (ICR 90207; Space- tive performance positively, improving learning labs Inc., Bellevue, CA, USA), validated against and memory.15–22 intra-arterial BP measurements.31 The recorder was Both ACE-I and ARB are used in the treatment set to take readings at 15-min intervals throughout of hypertension in the elderly, due to their good the 24-h period. Recordings were always started at efficacy/tolerability profile. However, in order to the same hour in the morning and were performed achieve the newly recommended BP goals, the use throughout the 24-h period, during which patients of combination therapy has been advocated on the were allowed to follow their normal daily routine grounds that its individual components may exert after they left the laboratory. Recordings were additive if not synergistic effect on BP.23,24 In this excluded from the analysis when more than 10% regard, since the pharmacological effects of hydro- of all readings or more than one reading per hour chlorothiazide (HCTZ) include the activation of the were missing or incorrect. RAS, its association with an ACE-I or an ARB is a For each patient, the following data relating to rational and useful approach, affording effective BP SBP, DBP and heart rate (HR) were provided by the control while minimizing possible side effects.25,26 analysis of the recordings: 24-h mean values, day- To date, at our knowledge, no data have been time (0700–2300) mean values, night time (2300– reported about the influence of this type of combi- 0700) mean values and hourly mean values. nation therapy on cognitive performance. Furthermore, for each patient the percentage of With this background, the present study was 24-h, day time and night time SBP and DBP readings undertaken to compare the effects of the fixed 4140 and 490 mmHg, respectively, was provided combination of the ARB telmisartan27 and HCTZ by the system. with those of the fixed combination of the ACE-I Following the recommendations of US Food and lisinopril28 and HCTZ on ambulatory BP and Drug Administration (FDA),32 the trough-to-peak cognitive function of elderly hypertensive patients, (T/P) ratio, defined as the ratio between the effect who are likely to be the most susceptible to the of an antihypertensive agent at the end of the effects of antihypertensive drugs on cognitive func- interval between doses (trough) and at the time of tion.29 its maximum effect (peak), was evaluated in each treatment group. For each ABPM recording, trough SBP and DBP levels were calculated as the average Patients and methods of the BP readings taken 22–24 h after the dose, whereas the peak values were calculated as the This was a prospective, randomized, open-label, average of the hour that showed the maximal blinded end point (PROBE),30 parallel group study average decrease in BP between 2 and 8 h after the with two treatment arms. A total of 160 outpatients dose plus the preceding or the following hour aged 61–75 years entered a 2-week wash-out period, depending on which showed the greater decreae in after which they were considered eligible for BP.33 Data were averaged for all patients. To quantify enrolment if they had a sitting diastolic BP the homogeneity of the antihypertensive effect over (DBP)X95 and o110 mmHg and a sitting systolic the 24 h, the smoothness index (SI) was also BP (SBP)4140 mmHg. Exclusion criteria included: computed by dividing the average of the 24 hourly secondary hypertension, myocardial infarction or change after treatment by the corresponding (s.d.).34 cerebrovascular accident within the preceding 6 This measurement has been shown to reflect months, clinically significant valvular heart disease, whether treatment smoothly reduces BP throughout heart failure, renal or hepatic insufficiency and a 24-h period and seems to provide a better known hypersensitivity to the drugs used in the assessment of homogeneity of antihypertensive study. treatment than T/P ratio.34 The study was performed in accordance with the Cognitive function was assessed at the end of the Declaration of Helsinki; the protocol was approved wash-out period and after 12 and 24 weeks of active by the local Ethical Committee and all of the treatment by means of the following six tests. patients gave their informed consent before enrol- ment. At the end of an initial 2-week wash-out period, during which any eventual antihypertensive Verbal fluency drug was discontinued, patients fulfilling the inclu- The subjects were asked to name as many animals as sion criteria were randomized to the telmisartan possible in 1 min and the score was the total number (80 mg)/HCTZ (12.5 mg) combination or to the of different animals named. Higher scores indicated lisinopril (20 mg)/HCTZ (12.5 mg) combination, less cognitive impairment. This verbal fluency test35 both given once daily at the same hour in the measures impairments in verbal production, seman- morning (approximately at 0800) for 24 weeks. tic memory and language and is particularly sensi-

Journal of Human Hypertension Antihypertensive treatment and cognitive functions in the elderly R Fogari et al 179 tive to early mental decline in older people. It is also Adverse events spontaneously reported or elicited a sensitive indicator of damage to the lateral frontal by indirect questioning by investigators blinded lobe.36 to patients’ treatment were recorded. Treatment compliance was assessed by counting the tablets remaining at each visit. Modified Boston naming test The subjects were asked to name 15 objects presented as line drawings, with a maximum of Statistical analysis 10 s being allowed for each. The modified Boston All analyses were conducted using the SAS system, naming test37 was scored on the basis of the number version 6.12 (SAS Institute Inc., Cary, NC, USA). of correct answers (maximum score: 15). Analysis of covariance (ANCOVA) was used to compare the changes in clinical variables. In addi- tion 95% confidence intervals for pairwise differ- Word-list memory test ences between treatment means were calculated. A total of 10 printed words were presented at the Differences in T/P ratios between treatments were rate of one every 2 s, and the subjects were then evaluated by nonparametric test (signed rank test asked to recall as many as possible within a period Proc Univariate), whereas the paired Student’s t-test of 90 s (maximum score: 10). This word-list memory was used to assess the differences in SI. Correlation test38 assesses the ability to remember newly learned analysis was performed using Pearson’s correlation information. test. A P-value of o0.05 was set as the level of clinical significance. All of the results are given as mean values7s.d. Word-list recall test This delayed memory test38 assesses the ability to remember the same 10 printed words after a delay of Results 10 min. A maximum of 90 s is allowed (maximum score: 10). A total of 160 patients, 76 men and 84 women, with a mean age of 67.674.5 years were randomized to receive telmisartan 80 mg/HCTZ 12.5 mg (n ¼ 80) or Word-list recognition lisinopril 20 mg/HCTZ 12.5 mg (n ¼ 80). Their The 10 words from the word-list memory test are demographic and clinic characteristics at baseline presented with 10 distractors. The subject’s score are shown in Table 1. No statistically significant in this recognition test is the number of correctly differences between groups were found in any of the recognized words.39 baseline characteristics. In all, 13 patients withdrew For the word-list tests, two different word lists after randomization, six for adverse events (two in were used at baseline and after 12 and 24 weeks of the telmisartan/HCTZ group for headache, dizziness treatment in order to avoid that the patients have and nausea, four in the lisinopril/HCTZ group: three already learned the list once. for cough and one for hypotension), four for inadequate therapeutic response (DBP reduction o10 mmHg) and three did not return for clinic visit. Trails B test A total pf 147 patients completed the study, 75 in This is a test of speeded mental operations, atten- the telmisartan/HCTZ group and 72 in the lisino- tion, visual scanning, visual sequential abilities and pril/HCTZ group. executive function.40 Scores are measured in sec- Results of 24-h ABPM, daytime and night time onds, with higher scores indicating slower or poorer SBP and DBP obtained during treatment with performance; an upper cutoff score of 181 s was used telmisartan/HCTZ and lisinopril/HCTZ combina- when time to complete the text exceeded 180 s. tion are shown in Table 2. Both treatments signifi- The cognitive function tests were performed on each visit by the same investigator blinded to clinical and active treatment findings. Persons who Table 1 Main baseline demographic and clinic characteristics of administered the tests attended a 2-day training the patients of the two treatment groups session conducted by neuropsychologists who were Telmisartan/ Lisinopril/ experts in cognitive function measurements. HCTZ HCTZ The tests used in the study were chosen because they have been shown to be reproducible and are Number of patients 80 80 38 valid measures of cognitive function. Also, in our Age (years) 67.8975.4 68.375.6 experience the test–retest reliability has been found Gender (male/female) 35/45 39/41 to be high (r ¼ 0.78 for verbal fluency test; r ¼ 0.71 Body weight (kg) 71.8712.3 72.1712.2 Duration of hypertension (years) 8.276.5 8.576.6 for Boston naming test; r ¼ 0.70 for word-list Systolic blood pressure (mmHg) 163.877.9 164.277.7 memory test; r ¼ 0.81 for word-list recall test; Diastolic blood pressure (mmHg) 101.774.0 101.374.1 r ¼ 0.77 for word-list recognition test; r ¼ 0.79 for Heart rate (beats/min) 75.976.1 76.275.8 Trails B test).

Journal of Human Hypertension Antihypertensive treatment and cognitive functions in the elderly R Fogari et al 180 cantly decreased ambulatory SBP and DBP values as altering the normal circadian BP pattern. Both compared to baseline (Po0.01 vs baseline for both combinations significantly decreased the percen- SBP and DBP with both drugs). However, analysis tages of SBP readings 4140 mmHg as well as the for adjusted means showed that the mean decreases percentages of DBP readings 490 mmHg in all in 24-h, daytime and night time SBP and DBP ABPM periods. However, telmisartan/HCTZ combi- obtained with the telmisartan/HCTZ combination nation was significantly superior to lisinopril/HCTZ were significantly greater than those obtained with in reducing the percentages of SBP readings the lisinopril/HCTZ combination, particularly at 24 4140 mmHg in the 24-h and daytime ABPM periods weeks (Po0.05 for 24-h and daytime SBP and DBP at both 12 and 24 weeks as well as the percentages of and for night time SBP; Po0.01 for night time DBP) DBP readings 490 mmHg in the 24-h and daytime (Figure 1). ABPM periods at 12 weeks (Table 3). Calculation of hourly average SBP and DBP For both treatments, the T/P ratios for SBP and DBP (Figure 2) showed that both telmisartan/HCTZ and were greater than 0.5, with no significant differences lisinopril/HCTZ combination maintained their anti- between telmisartan/HCTZ and lisinopril/HCTZ com- hypertensive efficacy throughout the 24 h, including bination. Also the average SI of SBP and DBP was the hours farthest from the last drug intake, without similar for the two treatments (Table 4).

Table 2 Mean values (7s.d.) of ambulatory BP during treatment with telmisartan/HCTZ and lisinopril/HCTZ

Variable 24 h Daytime Night time

Baseline 12 weeks 24 weeks Baseline 12 weeks 24 weeks Baseline 12 weeks 24 weeks

SBP (mmHg) Telm/HCTZ 151.579.9 132.175.0 129.375.2* 155.879.9 136.075.0* 133.175.3* 138.7711.5 120.476.8 117.476.8* Lisin/HCTZ 151.3710.2 133.675.5 131.775.4 155.5710.2 137.575.5 135.475.4 138.7711.6 121.877.4 119.877.2

DBP (mmHg) Telm/HCTZ 90.273.8 77.075.6* 75.076.6* 93.573.9 79.775.7 78.176.8* 79.975.7 68.476.5 66.277.1** Lisin/HCTZ 90.174.2 78.776.4 76.975.5 93.474.1 81.476.6 79.877.0 80.375.9 69.877.0 68.676.3

* ¼ Po0.05 vs lisinopril/HCTZ. ** ¼ Po0.01 vs lisinopril/HCTZ.

Figure 1 Mean7s.d. decreases from baseline in 24-h ambulatory blood pressure monitoring and daytime and nigth time systolic blood pressure (SBP) and diastolic blood pressure (DBP) obtained after 12 and 24 weeks of treatment with telmisartan/HCTZ and lisinopril/ HCTZ. Differences (and 95% confidence intervals) between treatments are shown at the bottom of each pair of histograms.

Journal of Human Hypertension Antihypertensive treatment and cognitive functions in the elderly R Fogari et al 181 Ambulatory HR was not affected by treatment The average cognitive test scores at baseline and with both telmisartan/HCTZ and lisinopril/HCTZ after 12 and 24 h of active treatment are shown in combination. Table 5. Lisinopril/HCTZ combination did not lead to any significant changes in any test score at any time of the study, whereas telmisartan/HCTZ com- bination significantly increased the scores of the word-list memory ( þ 17.1% at 12 weeks, Po0.05 vs baseline and vs lisinopril/HCTZ; þ 15.7% at 24 weeks, Po0.05 vs baseline), the word-list recall ( þ 13.5% at 12 weeks, Po0.05 vs baseline; þ 16.9% at 24 weeks, Po0.05 vs baseline and vs lisinopril/ HCTZ) and the Trails B (À33% at 12 weeks, Po0.05 vs baseline; À30.5% at 24 weeks, Po0.05 vs base- line). The scores of the Boston naming, word- list recognition and verbal fluency tests were not significantly changed by telmisartan/HCTZ (Figure 3). There was no significant correlation between the changes in cognitive test scores and the decrease in BP after 12 and 24 weeks of treatment with either drug. Based on counting the remaining tablets at each visit, 94% of the prescribed tablets were taken during telmisartan/HCTZ therapy and 92% during lisinopril/HCTZ therapy, indicating good treatment Figure 2 Systolic blood pressure (SBP) and diastolic blood compliance. pressure (DBP) (24 h) at baseline and after receiving telmisartan/ Both treatments were generally well tolerated. Six HCTZ for 12 and 24 weeks. patients reported one or more adverse events: two

Table 3 Mean percentages of ambulatory BP readings exceeding 140 mmHg for systolic BP (SBP) and 90 mmHg for diastolic BP (DBP) during treatment with telmisartan/HCTZ and lisinopril/HCTZ

Variable 24 h Daytime Night time

Baseline 12 weeks 24 weeks Baseline 12 weeks 24 weeks Baseline 12 weeks 24 weeks

SBP4140 (mmHg) Telm/HCTZ 77.0 24.0*** 15.4*** 89.1 31.4*** 20.2*** 40.8 1.7* 1.1 Lisin/HCTZ 75.9 29.7 20.8 87.9 38.7 26.9 39.8 5.7 2.7

DBP490 (mmHg) Telm/HCTZ 52.7 6.2** 5.0 67.7 8.0*** 6.6 9.7 0.8 0.2 Lisin/HCTZ 53.9 10.2 6.5 68.0 13.2 8.4 11.8 1.1 0.6

* ¼ Po0.05 vs lisinopril/HCTZ. ** ¼ Po0.01 vs lisinopril/HCTZ. *** ¼ Po0.001 vs lisinopril/HCTZ.

Table 4 Trough/peak ratio and smoothness index (mean values7s.d.) for systolic BP (SBP) and diastolic BP (DBP) during treatment with telmisartan/HCTZ and lisinopril/HCTZ

Variable Trough/peak ratio Smoothness index

12 weeks 24 weeks 12 weeks 24 weeks

SBP (mmHg) Telmisartan/HCTZ 0.6370.35 0.6370.30 2.3271.55 2.6971.81 Lisinopril/HCTZ 0.5770.44 0.5870.37 2.0671.35 2.4471.53

DPB (mmHg) Telmisartan/HCTZ 0.6470.36 0.6270.30 1.9670.91 2.3271.13 Lisinopril/HCTZ 0.5670.37 0.5670.26 1.8770.91 2.0870.94

Journal of Human Hypertension Antihypertensive treatment and cognitive functions in the elderly R Fogari et al 182 Table 5 Mean scores of verbal fluency, word-list memory, word-list recall, word-list recognition, Boston naming test and Trails B at the end of the wash-out period and after 12 and 24 weeks of treatment with telmisartan/HCTZ and lisinopril/HCTZ

Baseline Telmisartan/HCTZ 3 months Telmisartan/HCTZ 6 months

Verbal fluency (no. words) 14.676.4 14.876.5 14.475.9 Word-list memory (no. words) 7.072.7 8.272.1*,** 8.172.2* Word-list recall (no. words) 5.972.6 6.772.2* 6.972.1*,** Word-list recognition (n.c.a.) 7.171.5 7.971.4 7.871.3 Boston naming test (n.c.a.) 13.971.2 14.271.3 14.171.4 Trails B (s) 158.7774.4 106.3739.2* 110.2744.1*

Baseline Lisinopril/HCTZ 3 months Lisinopril/HCTZ 6 months

Verbal fluency (no. words) 14.176.8 14.576.7 14.676.9 Word-list memory (no. words) 6.972.6 6.772.7 6.872.8 Word-list recall (no. words) 5.872.3 5.772.4 5.972.5 Word-list recognition (n.c.a.) 7.271.4 7.171.3 7.371.4 Boston naming test (n.c.a.) 14.171.3 14.271.4 13.971.2 Trails B (s) 176.5770.7 143.4771.2 145.7771.4

n.c.a.: number of correct answer. * ¼ Po0.05 vs baseline. ** ¼ Po0.05 vs lisinopril/HCTZ.

administered once daily, produced a clinically relevant reduction in ambulatory BP values, without affecting the normal BP circadian profile. However, the telmisartan/HCTZ combination produced a slightly, but significantly greater decrease in 24-h mean values of SBP and DBP as well as in daytime and night time BP levels. Also the BP load, defined as the percentage of BP readings recorded by the ABPM tecnique that are abnormal in a 24-h period, was more decreased by treatment with the telmisar- tan/HCTZ combination than with lisinopril/HCTZ. This observation is important since BP load is considered a better determinant of cardiac and vascular damages than either clinic or mean 24-h ABPM values.41 When evaluating the duration and the homogene- ity of the hypotensive action over the 24 h, the T/P ratios for SBP and DBP obtained with the telmisar- tan/HCTZ and the lisinopril/HCTZ combination fulfilled the FDA guidelines (T/P ratio 450%), with no significant difference between the two treat- Figure 3 Systolic blood pressure (SBP) and diastolic blood pressure (DBP) (24 h) at baseline and after receiving lisinopril/ ments. This indicates that the antihypertensive HCTZ for 12 and 24 weeks. action of both treatments was sustained throughout the 24 h, which renders them suitable for once daily administration. Whereas the T/P ratio is exclusively (2.6%) treated with telmisartan/HCTZ (one head- an index of the duration of the antihypertensive ache, one dizziness) and four (5.5%) treated with effect, the SI is a measure of homogeneity of the lisinopril/HCTZ (two patients cough, one dizziness hypotensive action.34 In the present study, the SI of and one gastric discomfort). There were no serious SBP and DBP was similar with the two treatments, adverse experiences considered to be possibly or which indicates that both telmisartan/HCTZ and probably related to the trial medications. lisinopril/HCTZ combination smoothly reduced BP values throughout the 24 h, without difference between the two regimens. Discussion There was no significant alteration in ambulatory HR values. The results of this study indicate that in elderly The most original findings of this study, however, hypertensive patients treatment with both telmisar- regard the effects of the trial medications on tan/HCTZ and lisinopril/HCTZ combination, both cognitive function. Telmisartan/HCTZ combination

Journal of Human Hypertension Antihypertensive treatment and cognitive functions in the elderly R Fogari et al 183 significantly improved the scores of the word-list performance of elderly hypertensive or even im- memory, the word list recall and the Trails B tests. proved it, whereas contrasting data have been These findings are in agreement with the results reported in the literature about the cognitive effects obtained by ourselves in previous studies of of thiazide diuretic monotherapy.5,51,52 ARBs21,22 and suggest that the cognitive effects of these drugs may be selective. In particular, there were favourable effects on episodic memory and Conclusions visuospatial skills, but not on semantic memory and In conclusion, the results of this study indicate verbal skills. Since episodic memory and visuo- that 24 weeks of treatment with both telmisartan spatial skills reflect the operation of networks that 80 mg/HCTZ 12.5 mg and lisinopril 20 mg/HCTZ are widely distributed in the brain, their biological 12.5 combination were effective in reducing BP in complexity makes them more sensitive to numerous 42–44 elderly hypertensive patients. Telmisartan/HCTZ conditions affecting brain function. Unlike combination, however, produced a greater reduction episodic memory, memory span is relatively well in ambulatory BP levels and, unlike lisinopril/ preserved in many conditions affecting cognitive 45 HCTZ, improved some cognitive functions, particu- function, so that it might be difficult to detect any larly episodic memory and visuospatial abilities. treatment effect on memory span in otherwise Conflict of interests: None healthy hypertensive patients, at least over a relatively short period. The positive effects of telmisartan/HCTZ on memory might be explained by the interactions What is known about this topic: between Ang II and acethylcoline.46 Ang II inhibits K There is a relationship between high BP and cognitive impairment in later life3 the in vitro release of acethylcholine from the K The brain renin–angiotensin system participates in the entorhinal cortex associated with cognitive perfor- processing of attention, learning and memory10 mance and this action is reversed by ARBs.15,47 K Angiotensin II AT1 receptors blocker monotherapy seems to improve learning and memory in hypertensive elderly Besides, experimental studies have shown that 22 endogenous Ang II, which is released in response patients to AT1 receptor blockade, interacts with AT2 What this study adds: receptors, which may lead to memory-enhancing K An angiotensin II AT1 receptors blockers also in combination effects, possibly through the modulation of dopami- with hydrochlorothiazide improves cognitive functions, nergic and noradrenergic transmission.48 Blockade particularly episodic memory and visual-spatial abilities K An ACE inhibitor, given in combination with of Ang II at AT1 receptor level may also result in hydrochlorothiazide, has no effects on cognitive functions increased synthesis of Ang IV, that selectively binds K The changes in cognitive functions are unrelated with the to AT4 receptors; activation of AT4 receptors is changes in BP involved in memory acquisition and recall, perhaps by modulating neurotransmitter release and/or remodelling the cholinergic and glutamatergic path- ways in the hyppocampus.49,50 Unlike telmisartan/HCTZ combination, lisinopril/ References HCTZ combination did not significantly affect any of the cognitive function test scores at any time of 1 Skoog I, Lernfeldt B, Landahl S, Palmertz B, Andersson the study. These findings confirm that the use of this LA, Nilsson L et al. Fifteen-year longitudinal study ACE-I/HCTZ combination did not deteriorate cog- of blood pressure and dementia. Lancet 1996; 34: nitive performance over a period of 24 weeks, which 1141–1145. is reassuring when treating elderly hypertensive 2 Launer LJ, Ross GW, Petrovitch H, Masaki K, Foley D, patients, but did not afford any improving effect. White LR et al. Midlife blood pressure and dementia: the Honolulu-Asia aging Study. Neurobiol Aging 2000; Mechanisms other than the simple ACE-induced 21: 49–55. inhibition of Ang II synthesis, including the in- 3 Alves de Moraes S, Szklo M, Knopman D, Sato R. The volvement of other fragments of brain Ang II and of relationship between temporal changes in blood brain AT2 and AT4 receptors, might perhaps explain pressure and changes in cognitive function. Athero- the different effects of telmisartan and lisinopril sclerosis Risk in Communities (ARIC) Study. Prev Med on cognitive function tests. Correlation analysis 2002; 35: 258–263. showed no relationship between the changes in 4 Farmer ME, Kittner SJ, Abbott RD, Wolz NM, White the cognitive test scores and the decrease in BP after LR. Blood pressure and cognitive performance: 12 and 24 weeks of treatment with either drug, the Framingham Study. Am J Epidemiol 1987; 126: which suggests that the better outcome of telmisar- 1103–1114. 5 Starr JM, Whalley LJH, Deary IJ. The effects of tan/HCTZ combination cannot be ascribed to its antihypertensive treatment on cognitive function: greater lowering effect on ambulatory BP levels. results from the HOPE Study. J Am Geriatr Soc 1996; It is interesting to note that in the present study 44: 411–415. using HCTZ in combination with an ARB or an ACE- 6 Prince MJ, Bird AS, Blizard RA, Mann AH. Is cognitive I did not yeld to a deterioration in cognitive function of older patients affected by antihypertensive

Journal of Human Hypertension Antihypertensive treatment and cognitive functions in the elderly R Fogari et al 184 treatment? Results from the Medical Research Coun- 24 2003 European Society of Hypertension-European cil’s trial of hypertension in older adults. BMJ 1996; Society of Cardiology Guidelines for the management 312: 801–805. of arterial hypertension. J Hypertens 2003; 21: 7 Forette F, Seux ML, Staessen JA, Thijs L, Birkenhager 1011–1053. W. Prevention of dementia in randomized double- 25 Puschett JB. Diuretics. In: Oparil S, Weber MA (eds). blind placebo-controlled Systolic Hypertension in Hypertension: A Companion to Brenner and Rector’s Europe (Syst-Eur) trial. Lancet 1998; 352: 1347–1351. The Kidney. W.B. Saunders Company: Philadelphia, 8 Guo Z, Fratiglioni L, Zhu L, Fastbom J, Winblad B, 2000, pp 584–590. Viitanen M. Occurrence and progression of dementia 26 Weber MA. Rationalizing the treatment of hyperten- in a community population aged 75 years and older: sion. Am J Hypertens 2001; 14: 3S–7S. relationship of antihypertensive medication use. Arch 27 McClellan KJ, Markham A. Telmisartan. Drugs 1998; Neurol 1999; 56: 991–996. 56: 1039–1044. 9 Gaard PR. The role of angiotensin II in cognition and 28 Lancaster SG, Todd PA. Lisinopril. A preliminary behaviour. Eur J Pharmacol 2002; 438: 1–14. review of its pharmacodynamic and pharmacokinetic 10 Culman J, Blume A, Gohlke P, Unger T. The renin– properties and therapeutic use in hypertension and angiotensin system in the brain: possible therapeutic congestive heart failure. Drugs 1988; 35: 646–669. implications for AT1-receptor blockers. J Hum Hyper- 29 Starr JM, Whalley LJ. Senile hypertension and cogni- tens 2002; 16: S64–S70. tive impairment. An overview. J Hypertens 1992; 11 Baranowska D, Braszko JJ, Wisniewski K. Effect of 10(Suppl): S31–S42. angiotensin II on acquisition and extinction of condi- 30 Hansson L, Hedner T, Dahlof B. Prospective, rando- tioned avoidance in rats. Psychopharmacology 1983; mized, open label, blinded-end point (PROBE) study: a 81: 247–251. novel design for intervention trials. Blood Press 1992; 12 Braszko JJ, Wisniewski K, Kupryszewski G, Witczuk B. 1: 113–114. Psychotropic effects of angiotensin II and angiotensin 31 Groppelli A, Omboni S, Ravogli A, Villani A, Parati G, III in rat: locomotor and exploratory vs cognitive Mancia G. Validation of the Spacelabs 90202 and behaviour. Behav Brain Res 1987; 25: 195–203. 90297 devices for ambulatory blood pressure monitor- 13 Barnes JM, Barnes NM, Costall B, Horovitz ZP, Naylor ing by comparison with intra-arterial resting and RJ. Angiotensin II inhibits the release of (3H) acety- ambulatory measurements. J Hypertens 1991; 9(Suppl lcholine from the rat entorhinal cortex in vitro. Brain 3): S334–S335. Res 1989; 492: 136–143. 32 US Food and Drugs Administration. Proposed guide- 14 Vanhoutte PM. How to assess endothelial function lines for the Clinical Evaluation of Antihypertensive in human blood vessels. J Hypertens 1999; 17: Drugs. US Department of Health and Human Services, 1047–1058. Division od Cardio-Renal Drug Products: Rockville, 15 Barnes JM, Barnes NM, Costall B. Ace-inhibition and MD, 1988. cognition. In Mc Gregor GA, Sever PS (eds). Current 33 Omboni S, Parati G, Zanchetti A, Mancia G. Calcula- Advances in ACE-inhibition. Churchill-Livingstone: tion of trough: peak ratio of antihypertensive treatment London, 1989, pp 159–171. from ambulatory blood pressure. Methodological as- 16 Mondadori C, Etienne P. Nootropic effects of ACE pects. J Hypertens 1995; 13: 1105–1112. inhibitors in mice. Psychopharmacol 1990; 100: 34 Parati G, Omboni S, Rizzoni D, Agabiti-Rosei E, 301–307. Mancia G. The smoothness index: a new, reproducible 17 Starr JM, Whalley LJ. ACE-inhibitors: Central Actions. and clinically relevant measure of the homogeneity of Raven Press: New York, 1994. the blood pressure reduction with treatment for 18 Raghavendra V, Chopra K, Kulkarni SK. Comparative hypertension. J Hypertens 1998; 16: 1685–1691. studies on the memory enhancing actions of captopril 35 Isaacs B, Kennie AT. The SET test as an aid to the and losartan in mice using inhibitory shock avoidance detection of dementia in old people. Br J psychiatr paradigm. Neuropeptides 2001; 35: 65–69. 1973; 123: 467–470. 19 Tedesco MA, Ratti G, Mennella S, Manzo G, Grieco M, 36 Tranel D. Neuropsychological assessment. Psychiatr Rainone AC et al. Comparison of losartan and hydro- Clin North Am 1992; 15: 283–299. chlorothiazide on cognitive function and quality of life 37 Kaplan EF, Goodglass H, Weinstraub S. The Boston in hypertensive patients. Am J Hypertens 1999; 12: naming test. Veterans Administration Medical Center: 1130–1134. Boston, 1978. 20 Ebert U, Kirch W. Effects of captopril and enalapril on 38 Morris JC, Mohs RC, Rogers H, Fillenbaum G, Heyman electroencephalogram and cognitive performance in A. Consortium to establish a registry for Alzheimer’s healthy volunteers. Eur J Clin Pharmacol 1999; 55: disease (CERAD). Part I. Clinical and neuropsycholo- 255–257. gical assessment of Alzheimer’s disease. Neurology 21 Fogari R, Mugellini A, Zoppi A, Derosa G, Pasotti C, 1989; 39: 1159–1165. Fogari E et al. Influence of losartan and atenolol on 39 Mohs RC, Kim Y, Johus CA. Assessing change in cognitive function in very elderly hypertensive pa- Alzheimer’s disease: memory and language tests. In: tients. J Hum Hypertens 2003; 17: 781–785. Poon LW (ed). Handbook for Clinical Memory Assess- 22 Fogari R, Mugellini A, Zoppi A, Derosa G, Pasotti C, ment of Older Adults. American Psychological Asso- Fogari E et al. Effects of valsartan compared with ciation: Washington, 1986, pp 149–155. enalapril on blood pressure and cognitive function in 40 Reitan R. validity of the Trail Making Test as an elderly patients with essential hypertension. Eur J Clin indicator of organic brain damage. Percept Mot Skills Pharmacol 2004; 59: 863–868. 1958; 8: 271–276. 23 The Seventh Report of the Joint National Committee on 41 White WB. Blood pressure load and target organ effects Prevention, Detection, Evaluation and Treatment of in patients with essential hypertension. J Hypertens High Blood Pressure. JAMA 2003; 289: 2534–2573. 1991; 9(Suppl 8): S39–S41.

Journal of Human Hypertension Antihypertensive treatment and cognitive functions in the elderly R Fogari et al 185 42 Squire LR. Memory and Brain. Oxford University detected in a mouse habituation paradigm. Neuro- Press: Oxford, 1987. report 1990; 1: 239–242. 43 Cabeza R, Nyberg L. Imaging cognition: an empirical 48 Okuyama S, Sakagawa T, Inagami T. Role of the ang II review of PET studies with normal subjects. J Cogn type 2 receptors in the mouse central nervous system. Neurosci 1997; 9: 1–26. Jpn J Pharmacol 1999; 81: 259–263. 44 Fahlander K, Wahlin A, Fastbom J, Grut M, Forsell Y, 49 Wright JW, Stubley LA, Pederson ES, Kramar EA, Hill R et al. The relationship between signs of Hanesworth J, Harding JW. Contributions of the brain cardiovascular deficiency and cognitive performance angiotensin IV-AT4 receptor subtype system to spatial in old age: a population based study. J Gerontol B learning. J Neurosci 1999; 19: 3952–3961. Psychol Sci Soc Sci 200; 55: P259–P265. 50 Pederson ES, Krishnan R, Harding JW, Wright JW. A 45 Backman L, Small B, Wahlin A, Larsson M. Cognitive role for the angiotensin AT4 receptor subtype in functioning in very old age. In: Cruik FIM, Salthouse overcoming scopolamine-induced spatial memory TA (eds). The Handbook of Aging and Cognition, 2nd deficits. Regul Pept 2001; 102: 147–156. edn. Erlbaum: Muhwah, NJ, pp 499–558. 51 Goldstein G, Materson BJ, Cushman WC, Reda DJ, Freis 46 Raghawendra V, Chopra K, Kulkarni SK. Comparative ED, Ramirez EA et al. Treatment of hypertension in the studies of the memory-enhancing actions of captopril elderly(II): cognitive and behavioural function. Hyper- and losartan in mice using inhibitory shock avoidance tension 1990; 15: 361–369. in mice using inhibitory shock avoidance paradigm. 52 Vanmolkot FHM, DeHoon JNJN, Van de Ven LLM, Van Neuropeptides 2001; 35: 65–69. Bortel LMAB. Impact of antihypertensive treatment on 47 Barnes NM, Champaneria S, Costall B, Kelly ME, quality of life: comparison between bisoprolol and Murphy DA. Cognitive enhancing actions of DuP 753 bendrofluazide. J Hum Hypertens 1999; 13: 559–563.

Journal of Human Hypertension