renin-angiotensin system blockade and cognitive function in patients at high risk of cardiovascular...

www.thelancet.com/neurology Vol 10 January 2011 43

Articles

Renin-angiotensin system blockade and cognitive function in patients at high risk of cardiovascular disease: analysis of data from the ONTARGET and TRANSCEND studiesCraig Anderson, Koon Teo, Peggy Gao, Hisatomi Arima, Antonio Dans, Thomas Unger, Patrick Commerford, Leanne Dyal, Helmut Schumacher, Janice Pogue, Ernesto Paolasso, Nicolaas Holwerda, Irina Chazova, Azan Binbrek, James Young, Salim Yusuf, for the ONTARGET and TRANSCEND Investigators.

Summary Background Cardiovascular risk factors are associated with dementia and cognitive decline. We investigated the eff ects of renin-angiotensin system blockade on cognitive function in patients aged 55 years and older with established atherosclerotic cardiovascular disease or diabetes with end-organ damage in two clinical trials.

Methods In the main study, ONTARGET, a double-blind, double-dummy, randomised controlled trial, the eff ects on cardiovascular outcomes of standard doses of an angiotensin-converting enzyme (ACE) inhibitor (ramipril), an angiotensin-receptor blocker (telmisartan), and a combination of the drugs were evaluated in 25 620 participants. In the parallel TRANSCEND trial, the eff ects of telmisartan were compared with those of placebo in 5926 participants intolerant to ACE inhibitors. Secondary outcomes included cognitive impairment (defi ned by investigator-reported diagnosis of dementia or signifi cant cognitive dysfunction, or a score of ≤23 on the Mini-Mental State Examination [MMSE]) and cognitive decline (a decrease of ≥3 points on the MMSE from baseline during follow-up). Analyses were by intention to treat. We pooled data from these studies to identify baseline predictors of cognitive impairment and its frequency according to mean systolic blood pressure during follow-up. These studies were registered with ClinicalTrials.gov, number NCT00153101.

Findings During a median duration of 56 months (IQR 51–64) of follow-up in ONTARGET, cognitive impairment occurred in 652 (8%) of 7865 patients allocated ramipril, 584 (7%) of 7797 allocated telmisartan, and 618 (8%) of 7807 allocated combination treatment (combination vs ramipril, odds ratio [OR] 0·95, 95% CI 0·85–1·07, p=0·39; telmisartan vs ramipril, OR 0·90, 0·80–1·01, p=0·06). Corresponding fi gures for cognitive decline were 1314 (17%), 1279 (17%), and 1240 (17%) in each of the groups, respectively (telmisartan vs ramipril, OR 0·97, 0·89–1·06, p=0·53; combination vs ramipril, OR 0·95, 0·88–1·04, p=0·28). In TRANSCEND, cognitive impairment occurred in 239 (9%) of 2694 participants allocated telmisartan compared with 245 (9%) of 2689 allocated placebo (OR 0·97, 0·81–1·17, p=0·76). The corresponding fi gures for cognitive decline were 454 (17%) and 412 (16%; OR 1·10, 0·95–1·27, p=0·22).

Interpretation In patients with cardiovascular disease or diabetes, diff erent approaches to blocking of the renin-angiotensin system had no clear eff ects on cognitive outcomes. Although patients with the lowest systolic blood pressure had the greatest preservation of cognitive function, meta-regression analyses did not show any benefi ts of blood-pressure lowering on cognition over several years of treatment.

Funding Boehringer-Ingelheim.

IntroductionThe burden of dementia is increasing on ageing populations around the world.1 Evidence is accumulating that stroke and various other cardiovascular risk factors, especially raised blood pressure, are important in the development of both vascular and degenerative forms of dementia (ie, Alzheimer’s disease) and cognitive impairment.2–10 Therefore, eff ective management of cardiovascular risk factors could translate into major health benefi ts through the preservation of cognitive function.11

Blockade of the renin-angiotensin system reduces rates of major cardiovascular events,12 including stroke, mainly through lowering of blood pressure.13 More controversial, though, is the degree to which the benefi ts of such treatment extend to the prevention of dementia,14 or are

mediated by eff ects independent of blood pressure13 such as through anti-infl ammatory, antiglycaemic, anti-proliferative, and antioxidative eff ects.15 A potential neuroprotective eff ect of renin-angiotensin system blockade in focal cerebral ischaemia has been suggested that is diff erentially modulated by treatment specifi cally targeting the angiotensin II receptor.16 Moreover, the Fournier hypothesis proposes that angiotensin-receptor blocker (ARB) treatment has potential advantages over angiotensin-converting enzyme (ACE) inhibitors in the prevention of stroke and cognitive impairment because of more specifi c concomitant blocking of harmful eff ects (ie, vasoconstriction and proatherothrombogenesis), with neutral or possibly enhanced benefi cial eff ects (ie, vasodilatation and endothelial modulation).17,18 Although

Lancet Neurol 2011; 10: 43–53

Published OnlineOctober 27, 2010DOI:10.1016/S1474-4422(10)70250-7

See Comment page 22

See In Context page 29

The George Institute for Global Health, University of Sydney and Royal Prince Alfred Hospital, Sydney, NSW, Australia (C Anderson MD, H Arima MD); Population Health Research Institute, Hamilton Health Science Centre, Hamilton, ON, Canada (K Teo MD, P Gao MSc, L Dyal MSc, J Pogue MSc, S Yusuf DPhil); University of the Philippines, College of Medicine, Manila, Philippines (A Dans MD); Center for Cardiovascular Research and Institute of Pharmacology, Charite–Universitätsmedizin Berlin, Berlin, Germany (T Unger MD); Cardiac Clinic, Department of Medicine, University of Cape Town, Cape Town, South Africa, (P Commerford MD); Boehringer Ingelheim, Ingelheim, Germany (H Schumacher PhD); Estudios Cardiologicos Latinoamerica, Boulevard, Rosario, Argentina (E Paolasso MD); St Elisabeth Hospital, Tilburg, The Netherlands (N Holwerda MD); Arterial Hypertension Department, Cardiology Research Complex Ministry of Healthcare RF, Moscow, Russia (I Chazova MD); Rashid Hospital, Dubai, United Arab Emirates (A Binbrek MD); and Division of Medicine, Cleveland Clinic Foundation, Cleveland, OH, USA (J Young MD)

Correspondence to: Prof Craig Anderson, The George Institute for Global Health, University of Sydney and Royal Prince Alfred Hospital, Sydney 2050, NSW, [email protected]

Articles

44 www.thelancet.com/neurology Vol 10 January 2011

this hypothesis has not been substantiated in individual clinical trials,19 systematic reviews of blood-pressure lowering, mainly involving treatment with ACE inhibitors or ARBs in patients with and without established cardiovascular disease, suggest potential benefi cial eff ects on prevention of dementia and cognitive decline.20–22

The Ongoing Telmisartan Alone and in Combination with Ramipril Global Endpoint Trial (ONTARGET)23 and the parallel Telmisartan Randomized Assessment Study in ACE Intolerant Subjects with Cardiovascular Disease (TRANSCEND) trial24 have reported the eff ects of treatment with an ARB and an ACE inhibitor on major cardiovascular outcomes in patients with cardiovascular disease or high-risk diabetes without heart failure. We aimed to assess the eff ects of the study treatments on predefi ned secondary outcomes related to cognitive function.

MethodsStudy designAs described previously,23–25 the ONTARGET and TRANSCEND studies were modelled on the design of the Heart Outcomes Prevention Evaluation (HOPE) trial26 and took place at 733 centres in 40 countries, with central coordination at the Population Health Research Institute, McMaster University and Hamilton Health Sciences (Hamilton, ON, Canada) and regional coordination at Oxford University (Oxford, UK) and University of Auckland (Auckland, New Zealand). The studies included men and women who were aged 55 years and older at the time of randomisation and who had had evidence of coronary artery, peripheral vascular, or cerebrovascular disease or diabetes with end-organ damage, without any defi nite indication (eg, symptomatic congestive heart failure, haemodynamically signifi cant primary valvular or cardiac outfl ow tract obstruction) or contraindication to the study treatments (eg, inability to discontinue or known hypersensitivity to ACE inhibitors or ARBs). Patients with specifi c medical conditions including constrictive pericarditis, complex congenital heart disease, syncope of unknown cause, cardiac or revascularisation surgery within the previous 3 months, uncontrolled hypertension irrespective of treatment (>160 mm Hg systolic or >100 mm Hg diastolic blood pressure), heart transplantation, stroke due to subarachnoid haemorrhage, major renal artery stenosis, creatinine concentration higher than 265 mmol/L, or signifi cant hepatic dysfunction, which gave an indication of medical instability, diff erent natural history of cardiovascular disease, or a poor prognosis, were also excluded, as were individuals with signifi cant medical comorbidity or a social situation that prevented safe or practical long-term follow-up. In TRANSCEND, similar eligibility criteria were followed except that patients also had a history of intolerance to ACE inhibitors and an absence of overt proteinuria. Approval for the studies was obtained from the institutional ethics committee of each centre and all patients provided written informed consent.

Randomisation and maskingEligible patients entered a 3-week single-blind run-in period during which the doses of the study drugs were titrated upwards. In ONTARGET, 25 620 (88%) of 29 019 patients who entered the run-in phase were randomly allocated to treatment groups through a central automated system, with 8542 assigned to telmisartan 80 mg daily, 8576 to ramipril 5 mg daily, and 8502 to a combination of telmisartan 80 mg daily and ramipril 5 mg daily in a double-dummy fashion. After 2 weeks, patients assigned to ramipril had the dose increased to 10 mg daily. In TRANSCEND, 5926 (89%) of 6666 patients who entered the run-in were randomly assigned to treatment groups, with 2954 allocated to telmisartan 80 mg daily and 2972 to placebo. Follow-up visits were at 6 weeks, 6 months, and then every 6 months until study end.

ProceduresSite investigators and coordinators underwent training sessions before commencement, and monitoring visits were undertaken during the course of the study to ensure uniform adherence to procedures. A manual of operations provided a detailed outline of the protocol, and the case record forms included instructions for completion. At each assessment, physical measurements and demographic and medical information were obtained for all patients. More extensive assessments were undertaken at the baseline, 2-year, and penultimate (immediately before the fi nal) visits, which included a 12-lead electrocardiogram, measurement of resting blood pressure with an automated device, collection of a fasting blood sample and a urine sample, and an assessment of cognitive function. All serious adverse events and study outcomes were reported centrally and all events that formed the primary composite outcome (cardiovascular death, myocardial infarction, stroke, and admission to hospital for heart failure) and all-cause mortality were adjudicated using standardised criteria by a committee whose members were unaware of study group allocations. Details of prespecifi ed secondary outcomes including newly diagnosed heart failure, revascularisation procedures, newly diagnosed diabetes mellitus, nephropathy, new onset of atrial fi brillation, and the development of dementia or cognitive decline were provided with supporting documents where possible.

All patients were checked for evidence of dementia or clinically signifi cant cognitive impairment and had an assessment of cognitive function at baseline and at the 2-year and penultimate visits using the Mini-Mental State Examination (MMSE).27 For each successfully completed item on the MMSE, a score of 1 point was awarded to a maximum of 30, with an adjusted weighted score derived in situations in which there were missing item responses. Investigators in North America (excluding Mexico), Australia, South Africa, New Zealand, and the UK and Ireland, who contributed a total of 10 369 (33%) of the

Articles


patients randomly allocated to treatment groups in both ONTARGET and TRANSCEND, used a standardised English language version of the MMSE. Investigators in the other countries had the option of using a contextually appropriate translation of the MMSE in local clinical use or the English MMSE version administered in the local language if necessary. All non-English versions of the MMSE were checked for appropriateness before use.

For these analyses, the main outcomes were: (i) cognitive impairment, defi ned by either an investigator-reported and specialist-confi rmed diagnosis of dementia or cognitive impairment, or by a MMSE score of 23 or less during follow-up in patients without reported dementia or cognitive impairment at baseline; and (ii) cognitive decline, defi ned by a drop of 3 or more points between baseline and a follow-up recorded MMSE score in patients with a baseline and at least one follow-up measure, as used elsewhere.14 Quantitative changes in MMSE score between baseline and fi nal assessment were also compared between treatment groups.

Statistical analysisWe undertook analyses of the treatment eff ects using two-sided hypothesis tests, according to the principle of intention to treat, using logistic regression models to estimate odds ratios (ORs) and 95% CIs on the dichotomous outcomes of cognitive impairment and cognitive decline. Participants with missing baseline MMSE values were excluded and those with only a baseline MMSE were assumed not to have met the criteria for cognitive decline. Data were pooled from the ONTARGET and TRANSCEND studies to identify the predictors of cognitive impairment during follow-up, independent of treatment. Variables included were age, sex, educational level (≥9 years and <9 years), major ethnic group (white vs other), presence or absence of a history of hypertension, coronary artery disease, stroke (according to standard clinical criteria supported by neuroimaging), or transient ischaemic attack (TIA; according to standard clinical criteria within the past 12 months), and smoking, alcohol consumption (nil, less than four, and four or more

Figure 1: Trial profi le for ONTARGET MMSE=Mini-Mental State Examination.

8576 assigned ramipril 8542 assigned telmisartan

25 620 patients were randomised

Cognitive impairment 711 (8%) excluded from analysis because of cognitive impairment or dementia at baseline. These included: 148 deaths 11 missing baseline MMSE 3 lost to follow-up

Cognitive decline 974 (11%) excluded from analysis because they did not have ≥2 MMSE assessments to derive a change in scores. These included: 471 deaths 9 lost to follow-up 813 baseline MMSE score alone 161 no baseline MMSE score



29 019 patients entered run-in phase

3399 (12%) excluded 1123 (4%) poor compliance 597 (2%) consent withdrawn 492 (2%) symptomatic hypotension 287 (1%) raised creatinine or potassium 27 (<1%) deaths 873 (3%) other reasons

8502 assigned combination



Cognitive impairment 7865 included for assessment (includes 150 with missing baseline MMSE who were assumed free of the outcome)

Cognitive decline 7602 included for assessment





Articles


standard drinks per day), use of any antiplatelet agent, statin, or β blocker, low baseline MMSE score (≤25), and method of administration (English or translated version) of the MMSE. The association between 10 mm Hg increases and tertiles of mean achieved systolic blood pressure on the rate of cognitive impairment during follow-up were examined in logistic regression models. We undertook additional analyses for the outcomes of cognitive impairment using a limit for MMSE scores of fewer than 18 points and for the combination outcomes of cognitive impairment and recurrent stroke, and cognitive impairment and cognitive decline.

We also undertook a systematic review of clinical trials that assessed cognitive function (or dementia, if available) or cognitive decline using MMSE with a minimum of 1000 patient-years of follow-up in each randomised group. For each trial and each outcome, we calculated the relative risk and its variance according to the principle of intention to treat, with overall estimates of eff ect and 95% CIs calculated using a random eff ects model and

inverse variance weighting. We also did meta-regression analyses to explore the association between the diff erence in systolic blood pressure during follow-up between randomised groups and the log relative risk for outcomes. This association was investigated across trials by random eff ects meta-regression models with inverse variance weighting. STATA (release 10) and SAS (version 8.2 for Unix) were used for all analyses.

Role of the funding sourceThe studies were coordinated and analysed independently of the sponsor and the steering committee had overall responsibility for interpretation of results. The corresponding author had full access to all the data in the study and had fi nal responsibility for the decision to submit for publication.

ResultsThe trial profi les are shown in fi gures 1 and 2. The baseline characteristics of patients in the TRANSCEND

Figure 2: Trial profi le for TRANSCENDMMSE=Mini-Mental State Examination.

2954 assigned telmisartan 2972 assigned placebo

5926 patients were randomised


Cognitive decline 312 (11%) excluded from analysis because they did not have at least two MMSE assessments to derive a change in scores. These included: 264 baseline MMSE score alone 157 deaths 48 no baseline MMSE score 7 lost to follow-up


Cognitive decline 383 (13%) excluded from analysis because they did not have at least two MMSE assessments to derive a change in scores. These included: 328 baseline MMSE score alone 170 deaths 55 no baseline MMSE score 7 lost to follow-up

6666 patients entered run-in phase

740 (11%) excluded 311 (5%) poor compliance 135 (2%) consent withdrawn 37 (1%) raised creatinine or potassium 53 (1%) symptomatic hypotension 3 (<1%) deaths 201 (3%) other reasons





Articles


and ONTARGET studies, as reported previously,23–25 were generally similar between the studies with an average age of roughly 66 years and a history of stroke or TIA in about a fi fth (table 1). However, there was a higher proportion of women (43% vs 27%) and more patients with a history of hypertension (76% vs 69%) in TRANSCEND than in ONTARGET. Randomised groups were well balanced for patient characteristics that might aff ect risk of cognitive impairment, such as age, sex, level of education, eth-nic origin, baseline MMSE score, and method of administration (English vs translated version) of the MMSE. At randomisation, the median MMSE score was 29 (IQR 27–30) and 17% of patients had a score of 25 or less in both studies. At entry (before the run-in phase) in ONTARGET, mean blood pressure was 142/82 mm Hg across all three randomised groups, but fell on average slightly more in the combination (–2·4/1·4 mm Hg) and telmisartan (–0·9/0·6 mm Hg) groups than in the ramipril group during follow-up. In TRANSCEND, mean blood pressure was 4·0/2·2 mm Hg lower in the telmisartan group than in the placebo group during follow-up.

Of the 25 620 patients allocated to treatment groups, a total of 2151 were excluded from the analysis of cognitive impairment because they had cognitive impairment or dementia at baseline. Tables 1 and 2 show that cognition was assessed with the MMSE at baseline and in one or more follow-up visits for 22 629 (88%) of the 25 620 patients randomly allocated to treatment groups

in ONTARGET, allowing these patients to be assessed for the eff ects of treatment on cognitive decline. Figure 1 outlines reasons for the absence of an assessment in 2991 patients (ramipril, telmisartan, and combination groups): 2507 (84%) patients had only a baseline MMSE score and 484 (16%) had no baseline MMSE score, of whom 391 patients had documented dementia or cognitive impairment and were excluded from the analysis. The remaining 2600 patients were included in the analysis of cognitive impairment because there was a report of dementia or cognitive impairment (86 at the 2-year visit and 33 at the penultimate [or fi nal] visit), the investigator reported an absence of dementia (1388 at 2 years, 1099 at fi nal visit), patient refusal (65 at 2 years, 49 at penultimate or fi nal visit), or another reason (284 at 2 years, 117 at penultimate or fi nal visit), or because data was missing for 1217 patients; these reasons did not vary across the randomised groups.

Tables 1 and 3 show that a similar proportion (5231, 88%) of patients were assessed for cognitive decline during follow-up in TRANSCEND. The 695 patients without such an assessment included 592 (85%) with only a baseline score and 103 (15%) with no baseline score on the MMSE. However, of the 592 patients with only a baseline score, 39 (27 at 2 years and 12 at penultimate or fi nal visit) had dementia or cognitive impairment, 262 had investigator-reported absence of dementia, ten were due to patient refusal, 19 had other reasons, and 262 were excluded because of missing data,

ONTARGET TRANSCEND

Overall (n=25 620)* MMSE at baseline and ≥1 follow up visit (n=22 629)

Overall (n=5926)† MMSE at baseline and ≥1 follow up visit (n=5231)

Age (years) 66·4 (7·2) 66·2 (7·1) 66·9 (7·3) 66·6 (7·2)

Women 6831 (27%) 5917 (26%) 2547 (43%) 2229 (43%)

Education ≥9 years 17 223 (67%) 15 349 (68%) 3649 (62%) 3257 (62%)

European ethnic origin 18 708 (73%) 16 435 (73%) 3621 (61%) 3146 (60%)

Past history

Hypertension 17 607 (69%) 15 511 (69%) 4528 (76%) 3989 (76%)

Coronary artery disease 19 102 (75%) 16 932 (75%) 4418 (75%) 3930 (75%)

Stroke/transient ischaemic attack 5342 (21%) 4612 (20%) 1302 (22%) 1125 (22%)

Diabetes mellitus 9612 (38%) 8366 (37%) 2118 (36%) 1849 (35%)

Current smoker 3225 (13%) 2769 (12%) 582 (10%) 500 (10%)

1–4 alcoholic drinks per day 8565 (33%) 7678 (34%) 1559 (26%) 1401 (27%)

≥4 alcoholic drinks per day 1670 (7%) 1508 (7%) 265 (4%) 233 (4%)

Use of aspirin 20 727 (81%) 18 397 (81%) 4705 (79%) 4163 (80%)

Use of statins 15 783 (62%) 14 184 (63%) 3272 (55%) 2931 (56%)

Use of β blocker 14 583 (57%) 13 000 (57%) 3453 (58%) 3091 (59%)

MMSE assessment

Median score at randomisation 29 (27–30) 29 (27–30) 29 (27–30) 29 (27–30)

Score ≤25 4215 (16%) 3603 (16%) 1063 (18%) 902 (17%)

English version used 19 260 (75%) 17 105 (76%) 4521 (76%) 4001 (76%)

Data are n (%), mean (SD), or median (IQR). MMSE=Mini-Mental State Examination. *n=25 136 for median MMSE score. †n=5823 for median MMSE score.

Table 1: Demographics and baseline characteristics

Articles


but again the frequency of these reasons did not vary across groups.

Characteristics were broadly similar in patients with and without baseline MMSE and between those with multiple MMSE measures and the total study populations (table 1). Of 30 959 patients in both studies who had baseline MMSE scores, only 724 (2%) needed an adjusted weighted score. In both studies, there were no diff erences in the MMSE scores (median 29) between groups at the end of follow-up.

Tables 2 and 3 show results for the eff ect of the treatments on cognitive impairment and cognitive decline in the two studies. In ONTARGET, we noted no signifi cant diff erence between treatment groups in occurrence of cognitive impairment (combination vs ramipril, OR 0·95, 95% CI 0·85–1·07, p=0·39; telmisartan vs ramipril, OR

0·90, 0·80–1·01, p=0·06) or cognitive decline (telmisartan vs ramipril, OR 0·97, 0·89–1·06, p=0·53; combination vs ramipril, OR 0·95, 0·88–1·04, p=0·28). Similarly, in TRANSCEND, neither cognitive impairment (telmisartan vs placebo OR 0·97, 95% CI 0·81–1·17, p=0·76) nor cognitive decline (telmisartan vs placebo OR 1·10, 0·95–1·27, p=0·22) diff ered signifi cantly between treatment groups. Results for the other cognitive endpoints were also non-signifi cant.

Table 4 shows the variables that were associated with cognitive impairment during follow-up, on the basis of pooled data from ONTARGET and TRANSCEND. The strongest predictor variable was a baseline MMSE score of 25 or less (OR 5·51, 4·98–6·09). Other predictor variables were age, female sex, history of hypertension, previous stroke or TIA, use of the English version of the MMSE, and high mean systolic blood pressure during follow-up. Protective factors were a higher level of education, white ethnic origin, history of coronary artery disease, current smoking, use of any of antiplatelet drugs, statins, and β blockers, and alcohol consumption. None of the treatment eff ects in the studies were changed appreciably when adjustments were made sequentially, fi rst with these variables and then with average achieved systolic blood pressure (table 4).

There were no signifi cant diff erential eff ects of the treatments in patient subgroups defi ned by age, history of hypertension, or previous stroke or TIA, in either study. For the ONTARGET patients, there was no diff erence in the proportion of cases of cognitive impairment associated with a preceding stroke event during follow-up in the ramipril (44, 7%), telmisartan (40, 7%), and combined treatment (41, 7%) groups. Although a lower risk was found for the combined endpoint of stroke and cognitive impairment from treatment with telmisartan compared with ramipril in ONTARGET (OR 0·90, 0·81–0·99, p=0·03; table 2), this

Ramipril Telmisartan Combination Telmisartan vs ramipril Combination vs ramipril

OR (95% CI) p value OR (95% CI) p value

Number randomised 8576 8542 8502 ·· ·· ·· ··

Number with MMSE 7602 7566 7461 ·· ·· ·· ··

Primary outcomes

Cognitive impairment* 652/7865 (8%) 584/7797 (7%) 618/7807 (8%) 0·90 (0·80–1·01) 0·06 0·95 (0·85–1·07) 0·39

Cognitive decline† 1314/7602 (17%) 1279/7566 (17%) 1240/7461 (17%) 0·97 (0·89–1·06) 0·53 0·95 (0·88–1·04) 0·28

Secondary outcomes

Cognitive decline or impairment‡ 1488/8448 (18%) 1437/8404 (17%) 1423/8377 (17%) 0·96 (0·89–1·04) 0·38 0·96 (0·88–1·04) 0·28

Cognitive impairment or stroke 939/7865 (12%) 847/7797 (11%) 888/7807 (11%) 0·90 (0·81–0·99) 0·03 0·95 (0·86–1·04) 0·27

Cognitive impairment, MMSE score <18§

305/8473 (4%) 256/8433 (3%) 307/8396 (4%) 0·84 (0·71–0·99) 0·04 1·02 (0·86–1·19) 0·84

Data are n/N (%), OR (95% CI), and p value. OR=odds ratio. MMSE=Mini-Mental State Examination. *Defi ned by investigator-reported diagnosis of dementia or signifi cant cognitive impairment, or a score of 23 or less on the MMSE during follow-up in patients without dementia or cognitive impairment at baseline. †Defi ned by a drop of 3 or more points between baseline and follow-up MMSE score. ‡Defi ned by either cognitive decline or cognitive impairment during follow-up. §Defi ned by investigator-reported dementia or cognitive impairment, or a score of less than 18 on the MMSE.

Table 2: Cognitive outcomes in the ONTARGET study

Telmisartan Placebo Telmisartan vs placebo

OR (95% CI) p value

Number randomised 2954 2972 ·· ··

Number with MMSE 2642 2589 ·· ··

Primary outcomes

Cognitive impairment* 239/2694 (9%) 245/2689 (9%) 0·97 (0·81–1·17) 0·76

Cognitive decline† 454/2642 (17%) 412/2589 (16%) 1·10 (0·95–1·27) 0·22

Secondary outcomes

Cognitive decline or impairment‡ 520/2914 (18%) 487/2909 (17%) 1·08 (0·94–1·24) 0·27

Cognitive impairment or stroke 318/2694 (12%) 330/2689 (12%) 0·96 (0·81–1·13) 0·60

Cognitive impairment, MMSE score <18§

127/2917 (4%) 111/2932 (4%) 1·16 (0·89–1·50) 0·27

Data are n/N (%), OR (95% CI), and p value. OR=odds ratio. MMSE=Mini-Mental State Examination. *Defi ned by investigator-reported diagnosis of dementia or signifi cant cognitive impairment, or a score of 23 or less on the MMSE during follow-up in patients without reported dementia or cognitive impairment at baseline. †Defi ned by a drop of 3 or more points between baseline and follow-up recorded MMSE score. ‡Defi ned by either cognitive decline or cognitive impairment during follow-up. §Defi ned by investigator-reported dementia/cognitive impairment, or a score of less than 18 on the MMSE.

Table 3: Cognitive outcomes in the TRANSCEND study

Articles


result was not signifi cant after adjustment for average systolic blood pressure during follow-up, age, sex, ethnic origin, history of stroke, use of antiplatelet drugs, education, alcohol consumption, and baseline MMSE of 25 or less (p=0·09). Use of the more stringent 18-point threshold for cognitive impairment on the MMSE did show a signifi cant eff ect of treatment for telmisartan (OR 0·84, 0·71–0·99, p=0·04) but not for combination treatment (OR 1·02, 0·86–1·19, p=0·84) compared with ramipril (table 2), but these analyses were not prespecifi ed and are prone to random error.

Figure 3 shows that, on the basis of pooled data from the two trials, the frequency of cognitive impairment was lowest in patients with the lowest mean tertiles of systolic blood pressure up to the endpoint of dementia (overall unadjusted trend p<0·0001, adjusted trend p=0·04; tertile 2 vs 1, OR 1·12, 1·00–1·26, p=0·06; and tertile 3 vs 1, OR 1·14, 1·01–1·27, p=0·03) in multivariate logistic regression backward selection models that included the variables age, sex, ethnic origin, history of stroke, use of antiplatelet drugs, education, alcohol consumption, and baseline MMSE score of 25 or less, as well as study and treatment allocation. There was no signifi cant interaction between these variables.

We added our data to those from other randomised controlled trials of blood-pressure-lowering therapy on cognitive function (fi gures 4 and 5).14,22,28–31 Our updated meta-analysis showed that blood-pressure lowering is not signifi cantly more eff ective than placebo (or control) for reducing the risk of cognitive impairment (relative risk 0·97, 95% CI 0·92–1·01) and cognitive decline (0·97, 0·93–1·01; fi gure 4). However, although the trends were not signifi cant, meta-regression suggests reductions in the risks of cognitive impairment and cognitive decline (3·4%, 95% CI –4·6 to 10·9, p=0·40, and 1·0%, –3·4 to 5·2, p=0·66, respectively) associated with each 5 mm Hg reduction in systolic blood pressure that were consistent with the size of overall observed treatment eff ects (fi gure 5).

DiscussionIn these two clinical trials of patients at high vascular risk with established cardiovascular disease or complicated diabetes without heart failure, there were no signifi cant diff erences in the eff ects of standard fi xed high doses of ramipril, telmisartan, or their combination on diff erent indices of cognitive dysfunction over several years. The baseline characteristics of patients in these trials were similar to those of participants in the HOPE study,26 with many patients having baseline and on-treatment blood pressures within the normal range and receiving proven treatments for the prevention of cardiovascular disease during the course of the study. These fi ndings are therefore not only comparable to those seen on the overall and separate components of primary vascular endpoint cluster used in ONTARGET and TRANSCEND,23,24 but they are consistent with the cognitive outcomes seen with

Unadjusted OR (95% CI) Adjustment 1* Adjustment 2†

Age, per 1 year increase 1·07 (1·06–1·08) 1·06 (1·06–1·07) 1·06 (1·06–1·07)

Sex, female vs male 1·63 (1·49–1·77) 1·19 (1·08–1·31) 1·21 (1·10–1·33)

Education, ≥9 vs <9 years 0·39 (0·36–0·43) 0·52 (0·47–0·57) 0·52 (0·47–0·57)

Ethnic origin, white vs other 0·68 (0·62–0·74) 0·83 (0·75–0·92) 0·82 (0·74–0·91)

History of hypertension vs no history 1·36 (1·24–1·50) 1·07 (0·96–1·2) ··

History of CAD vs no history 0·71 (0·64–0·77) 0·95 (0·84–1·07) ··

History of stroke or TIA vs no history 1·62 (1·47–1·78) 1·34 (1·20–1·5) 1·37 (1·24–1·52)

Current smoker vs never or ex-smoker 0·77 (0·67–0·88) 1·00 (0·86–1·16) ··

Use of antiplatelet drugs vs no use 0·74 (0·67–0·82) 0·86 (0·77–0·96) 0·85 (0·76–0·95)

Use of statins vs no use 0·78 (0·72–0·85) 1·07 (0·97–1·18) ··

Use of β blockers vs no use 0·80 (0·73–0·87) 0·96 (0·87–1·06) ··

MMSE score, ≤25 vs >25 score 5·51 (4·98–6·09) 4·28 (3·85–4·76) 4·28 (3·85–4·75)

MMSE, English vs translated version used 1·12 (1·02–1·24) 1·02 (0·91–1·14) ··

<4 alcoholic drinks per day vs no alcohol 0·63 (0·57–0·69) 0·83 (0·74–0·92) 0·84 (0·75–0·93)

≥4 alcoholic drinks per day vs no alcohol 0·61 (0·50–0·74) 0·88 (0·71–1·09) ··

Mean achieved SBP during follow-up, each 10 mm Hg increase

1·14 (1·11–1·18) 1·03 (1·00–1·07) 1·04 (1·00–1·07)

OR=odds ratio. CAD=coronary artery disease. TIA=transient ischaemic attack. MMSE=Mini-Mental State Examination. SBP=systolic blood pressure. *Adjusted for all factors. †Adjusted for signifi cant factors with p<0·1 from backward selection.

Table 4: Analysis of variables associated with cognitive impairment, defi ned by a score of 23 or less on the MMSE, in the pooled data from the ONTARGET and TRANSCEND studies

Figure 3: Rates of cognitive impairment, defi ned by scores of 23 or less on the MMSE, according to tertiles of mean systolic blood pressure achieved during follow-up of patients in the combined ONTARGET and TRANSCEND studies Overall p<0·0001 for trend. Rates are adjusted for variables shown in table 4, including age, sex, level of education, ethnic origin, cardiovascular risk factors, stroke or transient ischaemic attack, antiplatelet use, baseline MMSE score and technique, alcohol consumption, and study treatment allocation. Error bars show 95% CIs. MMSE=Mini-Mental State Examination. SBP=systolic blood pressure.

Freq

uenc

y of c

ogni

tive

impa

irmen

t (%

) 7

5

4

3

1

0<130 >140130–140

Tertiles of mean SBP (mm Hg)

10

9

8

6

2

Unadjusted, trend p<0·0001Adjusted, trend p=0·04

Articles


antihypertensive agents in other cardiovascular prevention trials (see panel).22,28–30

The results appear robust, in view of the large number of events and the consistent lack of eff ect of the diff erent treatments on cognitive outcomes, even after adjustment for potential confounding variables and when key patient subgroups were evaluated. In ONTARGET, standard drug doses (10 mg ramipril on the basis of the HOPE trial26 and 80 mg telmisartan for maximum blood-pressure lowering) were shown to be non-inferior with respect to eff ects on major cardiovascular event outcomes.23 Even the

combination of these agents, which was able to produce greater blockade of the renin-angiotensin system compared with single agents, evident by the increased rates of hypotension and syncope, renal dysfunction and failure, and hyperkalaemia,23 failed to reduce rates of either cognitive impairment or cognitive decline.

We cannot exclude the possibility that the lack of eff ect of the study treatments was attributable to methodological limitations. In particular, that there was insuffi cient power in these trials to detect small treatment eff ects requiring much longer periods of exposure to manifest on cognition

Figure 4: Eff ects of blood-pressure lowering on cognitive impairment and cognitive declineBoxes and horizontal lines represent relative risk and 95% CIs for each trial; size of boxes is proportional to the inverse of variance of that trial result. Diamonds show 95% CIs for pooled estimates of eff ect and are centred on pooled relative risk. Half of the patients in the ACEI group were included in each comparison of ONTARGET. ARB=angiotensin-receptor blocker. ACEI=angiotensin-converting enzyme inhibitor.

Cognitive impairmentActive vs placebo ADVANCE 39 5569 37 5571 1·05 (0·67–1·65) HYVET-COG 126 1687 137 1649 0·90 (0·71–1·13) PROGRESS 193 3051 217 3054 0·89 (0·74–1·07) PRoFESS 1178 7739 1162 7726 1·01 (0·94–1·09) SCOPE 62 2477 57 2460 1·08 (0·76–1·54) SHEP 37 2365 44 2371 0·84 (0·55–1·30) Syst-Eur 11 1238 21 1180 0·50 (0·24–1·03) TRANSCEND 239 2694 245 2689 0·97 (0·82–1·15)Subtotal 1885 26 820 1920 26 700 0·98 (0·92–1·04)Test for homogeneity: χ2=6·44; df=7 (p=0·49) Test for overall effect: Z=0·63 (p=0·53)

Head to head ONTARGET (ARB vs ACEI)* 584 7897 326 3933 0·90 (0·79–1·03) ONTARGET (ARB+ACEI vs ACEI)* 618 7807 326 3933 0·96 (0·84–1·09)

Total 3087 42 424 2572 34 566 0·97 (0·92–1·01)Test for homogeneity: χ2=7·73; df=9 (p=0·56) Test for overall effect: Z=1·39 (p=0·17) Cognitive declineActive vs placebo ADVANCE 633 5569 640 5571 0·99 (0·89–1·10) HYVET-COG 485 1687 486 1649 0·98 (0·88–1·08) PROGRESS 276 3051 334 3054 0·83 (0·71–0·96) PRoFESS 795 7531 832 7518 0·95 (0·87–1·05) SCOPE 113 2416 125 2409 0·90 (0·70–1·16) TRANSCEND 454 2642 412 2589 1·08 (0·96–1·22)Subtotal 2756 22 896 2829 22 790 0·97 (0·91–1·03)Test for homogeneity: χ2=7·85; df=5 (p=0·17) Test for overall effect: Z=1·10 (p=0·27)

Head to head ONTARGET (ARB vs ACEI)* 1279 7566 657 3801 0·98 (0·90–1·07) ONTARGET (ARB+ACEI vs ACEI)* 1240 7461 657 3801 0·96 (0·88–1·05)

Total 5275 37 923 4143 30 392 0·97 (0·93–1·01)Test for homogeneity: χ2=7·92; df=7 (p=0·34) Test for overall effect: Z=1·53 (p=0·13)

Event n Event n

Active Control Relative risk, random (95% CI)

0·25 2·001·000·50

Favours active Favours control

Articles


than was expected for the primary vascular events. The patients included in these studies were at increased risk of cognitive impairment by virtue of their high cardiovascular risk status, including having high frequencies of hypertension, coronary heart disease, and stroke; the rates of the cognitive outcomes were similar to those seen in the Perindopril Protection Against Recurrent Stroke Study (PROGRESS) trial14 of patients with established cerebrovascular disease, in which an ACE-inhibitor-based blood-pressure-lowering regimen was shown to reduce risks of these outcomes, albeit more clearly when associated with recurrent stroke. However, since the average age of the patients in ONTARGET and TRANSCEND was roughly 66 years, which is about a decade less than the point at which rates of dementia increase rapidly, and because ONTARGET was an active-comparison trial with a background of high levels of treatment for cardiovascular disease, our studies might still have lacked suffi cient numbers of cognitive endpoints.

Various meta-analyses of placebo-controlled studies in similar patient groups have shown non-signifi cant eff ects of antihypertensive treatments of roughly 5–10% relative reduction on the risk of dementia or signifi cant cognitive impairment.20–22 We note with interest, therefore, that the point estimates of eff ects in both ONTARGET and TRANSCEND were within this range, and more clearly for the endpoint of cognitive impairment than cognitive decline. Any benefi cial eff ects of such treatment are, however, likely to vary according to several factors, including previous stroke and cerebral white matter disease, the type of agent, and the duration and intensity of blood-pressure-lowering treatment.

Another factor to consider is measurement error, which complicated the assessment of cognition. Although several sophisticated neuropsychological assessment and diagnostic tests of cognition are available, they are generally not practical for use in large-scale, multinational clinical trials, whereas the MMSE provides a tolerably useful, albeit simple, measure of cognition for such purposes. Even so, we recognise that the MMSE is limited by being insensitive to early (or mild degrees of) cognitive impairment, in particular to frontal-lobe executive dysfunction, which is often an early feature of vascular-related cognitive impairment. The MMSE is also aff ected by various confounding variables, such as level of education, physical impairment, and method of administration, as re-emphasised in the predictive modelling of cognitive impairment during follow-up (table 4), which although apparently well balanced between randomised groups, could have led to bias.

With respect to the Fourier hypothesis and activation (or inhibition) of neuroprotection in cerebral ischaemia,17,18 our results could have also been aff ected by a potential negative interaction between the study treatments and concomitant background therapies that decrease angiotensin II formation, such as has been proposed with ACE inhibitors and β blockers. Alternatively, the

combination of telmisartan and ramipril treatment might have had blunting adverse eff ects because of a greater (or broader) suppression of angiotensin receptors in the brain, despite greater blood-pressure lowering than that achieved with either telmisartan or ramipril alone.32

Thus, although results of observational studies suggest that control of blood pressure, particularly during middle age, is an important factor aff ecting the risk of dementia and cognitive decline in later life,4,33 trials of blood-pressure-lowering treatment to date have been inconsistent. For example, positive eff ects on cognition were seen in placebo-controlled studies that achieved fairly small reductions in blood pressure from the ACE inhibitor perindopril with or without the diuretic indapamide in PROGRESS14 and the calcium antagonist nitrendipine in the Systolic Hypertension in

Figure 5: Association of reduction in systolic blood pressure with risk reduction for (A) cognitive impairment and (B) cognitive declineThe area of each circle is proportional to inverse variance of log relative risk. Fitted lines represent summary meta-regressions for outcomes. Half of the patients in the ACEI group were included in each comparison of ONTARGET. ARB=angiotensin-receptor blocker. ACEI=angiotensin-converting enzyme inhibitor.

Rela

tive

risk

(95%

CI)

2·0

1·0

0·50

0·25–20 0–5–10–15

Difference in reduction in systolic blood pressure (mm Hg)

Rela

tive

risk

(95%

CI)

2·0

1·0

0·50

0·25–20 0–5–10–15

B

AReduction in risk for each 5 mm Hg reduction3·4% (95% CI –4·6 to 10·9), p=0·40

Reduction in risk for each 5 mm Hg reduction1·0% (95% CI –3·4 to 5·2), p=0·66

HYVET-COG SHEP

Syst-Eur

PROGRESS

ADVANCE

ONTARGET(ARB vs ACEI)

PRoFESS SCOPE

ONTARGET(ARB+ACEI vs ACEI)

TRANSCEND

HYVET-COG

PROGRESS

ADVANCEONTARGET

(ARB vs ACEI)

PRoFESS

SCOPE

ONTARGET(ARB+ACEI vs ACEI)

TRANSCEND

Articles


the Europe (Syst-Eur) trial.34 However, no clear eff ects were seen in the Study of Cognition and Prognosis in the Elderly (SCOPE) trial,28 which evaluated the ARB candesartan in older hypertensive patients, and the Prevention Regime for Eff ectively Avoiding Second Strokes (PRoFESS) study,29 which evaluated telmisartan in patients with recent ischaemic stroke. Similarly, negative results were also seen in the Hypertension in the Very Elderly Trial (HYVET)22 despite achievement of a fairly large (15 mm Hg) reduction in systolic blood pressure from the use of indapamide with or without the addition of an ACE inhibitor, although the baseline blood pressures of patients in this trial were much higher than they were in the other studies.

In our meta-analysis, blood-pressure lowering did not signifi cantly reduce risk of cognitive impairment and decline, although non-signifi cant trends suggested reductions in the risks of these conditions associated with each 5 mm Hg decrease in systolic blood pressure that were consistent with the overall treatment eff ects. In view of these pooled data, which contrast with the accumulating epidemiological data and the results of clinical observational databases35 suggesting benefi ts of blood-pressure lowering on cognition, the possibility remains that the eff ects of antihypertensive treatment on the cerebral microcirculation (or neuroprotection, or

both) require much longer periods of exposure to manifest small yet still clinically worthwhile benefi ts.

In summary, in a population of patients with established cardiovascular disease or complicated diabetes, diff erent approaches (and degrees) of blockade of the renin-angiotensin system using telmisartan and ramipril alone or in combination, in addition to background treatment, did not signifi cantly aff ect cognitive function. However, in view of the strong association between cardiovascular risk factors such as stroke and future dementia, and the potential benefi cial much smaller eff ects of blood-pressure lowering on this outcome, eff orts should continue to include cognitive function, especially with more sensitive measures, as well as having longer periods of follow-up, in the design of future trials. A proven eff ective treatment to delay the onset of dementia by only a few years could have major public health implications in ageing populations.

ContributorsAll authors were members of the trial’s steering committee and

contributed to discussion and interpretation of the data, and to the

writing of the report. The analysis was planned by CA, PG, LD, JP, and

SY. Data were analysed by PG, JP, HS, HA, and CA. All authors had full

access to data. A full list of all investigators has been published

elsewhere.23,24

Confl icts of interestCA, KT, JY, AD, TU, and SY have received consulting and lecture fees

and research grants from Boehringer Ingelheim. HS is an employee of

Boehringer Ingelheim.

References1 Murray CJL, Lopez AD, eds. The global burden of disease:

a comprehensive assessment of mortality and disability from diseases, injuries and risk factors in 1990 and projected to 2020. Cambridge, MA, USA: Harvard University Press, 1996.

2 Longstreth WT Jr, Manolio TA, Arnold A, et al. Clinical correlates of white matter fi ndings on cranial magnetic resonance imaging of 3301 elderly people. The Cardiovascular Health Study. Stroke 1996; 27: 1274–82.

3 Dufouil C, de Kersaint-Gilly A, Besancon V, et al. Longitudinal study of blood pressure and white matter hyperintensities: the EVA MRI cohort. Neurology 2001; 56: 921–26.

4 Launer LJ, Ross GW, Petrovitch H, et al. Midlife blood pressure and dementia: the Honolulu-Asia aging study. Neurobiol Aging 2000; 21: 49–55.

5 Gorelick PB. Risk factors for vascular dementia and Alzheimer disease. Stroke 2004; 35 (suppl I): 2620.

6 Solfrizzi V, Panza F, Colacicco AM, et al. Vascular risk factors, incident of MCI, and rates of progression to dementia. Neurology 2004; 63: 1852–91.

7 Vermeer SE, Prins ND, den Heijer T, et al. Silent brain infarcts and the risk of dementia and cognitive decline. N Engl J Med 2003; 348: 1215–22.

8 Hofman A, Ott A, Breteler MMB, et al. Atherosclerosis, apolipoprotein E, and prevalence of dementia and Alzheimer’s disease in the Rotterdam Study. Lancet 1997; 349: 151–54.

9 Snowdon DA, Greiner LH, Mortimer JA, et al. Brain infarction and the clinical expression of Alzheimer disease: the Nun Study. JAMA 1997; 277: 813–17.

10 Kivipelto M, Helkala EL, Laakso MP, et al. Midlife vascular risk factors and Alzheimer’s disease in later life: longitudinal, population based study. BMJ 2001; 322: 1447–51.

11 Erkinjuntti T, Román G, Gauthier S, Feldman H, Rockwood K. Emerging therapies for vascular dementia and vascular cognitive impairment. Stroke 2004; 35: 1010–17.

12 Dagenais GR, Pogue J, Fox K, Simmons ML, Yusuf S. Angiotensin-converting-enzyme inhibitors in stable vascular disease without left ventricular systolic dysfunction or heart failure: a combined analysis of three trials. Lancet 2006; 368: 581–88.

Panel: Research in context

Systematic reviewWe searched Medline (1950–2010), Embase (1986–2010), and the Cochrane Central Register of Controlled Trials (The Cochrane Library issue 3, 2010) with relevant text words and medical subject headings that included all spellings of antihypertensive agents, dementia, and cognitive impairment. We also obtained information from the Blood Pressure Lowering Trialists’ Collaboration. Trials were eligible for inclusion in our analyses if they assessed cognitive function with a minimum of 1000 patient-years of follow-up in each randomised group. We assessed reports with outcomes of cognitive impairment (or dementia, if cognitive impairment was not available) or cognitive decline (assessed with Mini-Mental State Examination), or both.

InterpretationThe ONTARGET and TRANSCEND trials showed that diff erent approaches to blocking the renin-angiotensin system had no signifi cant eff ects on cognitive function in patients with established cardiovascular disease or diabetes mellitus. Estimated small eff ects of renin-angiotensin blockade were almost similar to those expected from meta-regression analyses associated with observed reductions in systolic blood pressure. These results are consistent with those of other trials showing no clear benefi ts of blood-pressure lowering on the prevention of cognitive impairment over several years of treatment.

Articles


13 Turnbull F, Neal B, Pfeff er M, et al. Blood pressure-dependent and independent eff ects of agents that inhibit the renin-angiotensin system. J Hypertens 2007; 25: 951–58.

14 Tzourio C, Anderson C, Chapman N, et al, for the PROGRESS Collaborative Group. Eff ects of blood pressure lowering with perindopril and indapamide therapy on dementia and cognitive decline in patients with cerebrovascular disease. Arch Intern Med 2003; 163: 1069–75.

15 Schmieder RE, Hilgers KF, Schlaich MP, Schmidt BMW. Renin-angiotensin system and cardiovascular risk. Lancet 2007; 369: 1208–19.

16 Krikov M, Thoene-Reineke C, Müller S, Villringer A, Unger T. Candesartan but not ramipril pre-treatment improves outcome after stroke and stimulates neurotrophin BNF/TrkB system in rats. J Hypertens 2008; 26: 544–52.

17 Fournier A, Messerli F, Achard J, Fernandez L. Cerebroprotection mediated by angiotensin II: a hypothesis supported by recent randomised clinical trials. J Am Coll Cardiol 2004; 43: 1343–47.

18 Boutitie F, Oprisin R, Achard JM, et al. Does a change in angiotensin II formation caused by antihypertensive drugs aff ect the risk of stroke. J Hypertens 2007; 25: 1543–53.

19 Staessen JA, Richart T, Wang Z, Thijs L. Implications of recently published trials of blood pressure lowering drugs in hypertension or high risk patients. Hypertension 2010; 55: 819–31.

20 McGuinness B, Todd S, Passmore P, Bullock R. Blood pressure lowering in patients without prior cerebrovascular disease for prevention of cognitive impairment and dementia. Cochrane Database Syst Rev 2009; 4: CD004034.

21 Feigin V, Ratnasabapathy Y, Anderson C. Does blood pressure lowering treatment prevents dementia or cognitive decline in patients with cardiovascular and cerebrovascular disease? J Neurol Sci 2005; 229–30: 151–55.

22 Peters R, Beckett N, Forette F, et al, for the HYVET investigators. Incident dementia and blood pressure lowering in the Hypertension in the Very Elderly Trial cognition function assessment (HYVET-COG): a double-blind, placebo controlled trial. Lancet Neurol 2008; 7: 683–89.

23 The ONTARGET Investigators. Telmisartan, ramipril, or both in patients at high risk of vascular events. N Engl J Med 2008; 358: 1547–59.

24 The Telmisartan Randomised AssessmeNt Study in ACE iNtolerant subjects with cardiovascular Disease (TRANSCEND) Investigators. Eff ects of the angiotensin-receptor blocker telmisartan on cardiovascular events in high-risk patients intolerant to angiotensin-converting enzyme inhibitors: a randomised controlled trial. Lancet 2008; 372: 1174–83.

25 Teo K, Yusuf S, Sleight P, et al. Rationale, design, and baseline characteristics of 2 large, simple, randomized trials evaluating telmisartan, ramipril, and their combination in high-risk patients: The Ongoing Telmisartan Alone and in Combination with Ramipril Global Endpoint Trial/Telmisartan Randomized Assessment Study in ACE Intolerant Subjects with Cardiovascular Disease (ONTARGET/TRANSCEND) trials. Am Heart J 2004; 148: 52–61.

26 Yusuf S, Sleight P, Pogue J, Bosch J, Davies R, Dagenais G. Eff ects of an angiotensin-converting-enzyme inhibitors, ramipril, on cardiovascular events in high-risk patients. The Heart Outcomes Prevention Evaluation study investigators. N Engl J Med 2000; 342: 145–53.

27 Folstein MF, Folstein SE, McHugh PR. “Mini-Mental State”: a practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 1975; 12: 189–98.

28 Lithell H, Hansson L, Skoog I, et al. The study on cognition and prognosis in the elderly (SCOPE): principal results of a randomized double-blind intervention trial. J Hypertens 2003; 21: 875–86.

29 Diener HC, Sacco RL, Yusuf S, et al, for the Prevention Regimen for Eff ectively Avoiding Second Strokes (PRoFESS) study group. Eff ects of aspirin plus extended-release dipyridamole versus clopidogrel and telmisartan on disability and cognitive function after recurrent stroke in patients with ischaemic stroke in the Prevention Regimen for Eff ectively Avoiding Second Strokes (PRoFESS) trial: a double-blind, active and placebo-controlled study. Lancet Neurol 2008; 7: 875–84.

30 ADVANCE Collaborative Group. Eff ects of a fi xed combination of perindopril and indapamide on macrovascular and microvascular outcomes in patients with type 2 diabetes mellitus (the ADVANCE trial): a randomised controlled trial. Lancet 2007; 370: 829–40.

31 Di Bari M, Pahor M, Franse LV, et al. Dementia and disability outcomes in large hypertension trials: lessons learned from the systolic hypertension in the elderly program (SHEP) trial. Am J Epidemiol 2001; 153: 72–78.

32 Kehoe PG, Wilcock GK. Is inhibition of the renin–angiotensin system a new treatment option for Alzheimer’s disease? Lancet Neurol 2007; 6: 373–78.

33 Peila R, White LR, Masaki K, Petrovitch H, Launer LJ. Reducing the risk of dementia: effi cacy of long-term treatment of hypertension. Stroke 2006; 37: 1165–70.

34 Forette F, Seux ML, Staessen JA, et al, on behalf of the Syst-Eur Investigators. Prevention of dementia in randomised double-blind placebo-controlled Systolic Hypertension in Europe (Syst-Eur) trial. Lancet 1998; 352: 1347–51.

35 Li NC, Lee A, Whitmer RA, et al. Use of angiotensin receptor blockers and risk of dementia in a predominantly male population: prospective cohort analysis. BMJ 2010; 340: b5465.

renin-angiotensin system blockade and cognitive function in patients at high risk of cardiovascular...

Documents