Table 3 Summary of characteristics of the included studies

Carlsson 2014 [43]

Cohort study

To study mortality rates in men and woman with hypertension and AF prescribes different cardiovascular pharmacotherapies

(time to death between registration of AF diagnosis and 31.12.2007)

Antithrombotic drugs alone:

- antiplatelet agents vs. no treatment

- anticoagulants vs. no treatment

- anticoagulants vs. antiplatelets

Beta blocker:

- Selective vs. no treatment

- non-elective vs. no treatment

- non-selective vs. selective

RAS-blocking agents:

- vs. no treatment

- ARBs vs. ACE inhibitors

- RAS blocking agents + other vs. no treatment

- ARB + thiazide vs. ACE inhibitor + thiazide

Calcium receptor- blocking agents:

- Vessel selective vs. no treatment

- Heart active vs. no treatment

- Vessel selective vs. heart active

- - Statins

n = 5602

4748 aged ≥65 y

854 aged <65 y

Mean follow-up 3.4 years

Mortality

Collier 2011 [42]

Multicentre, international randomized trial

To evaluate the efficacy and safety of an amlodipine versus an atenolol-based antihypertensive regimen among older (≥ 65 years) and younger (<65 years) patients.

Atenolol-based (atenolol ± thiazide diuretic)

Amlodipine-based (amlodipine ± perindopril)

n = 19,257

8137 aged ≥65 and 11,020 aged <65

Median follow up 5.5 years

Nonfatal myocardial infarction and fatal coronary heart disease and cardiovascular events

Coope 1986 [34]

RCT

To explore, if the treatment of hypertension in patients between the ages of 60 and 79 years old reduces the incidence of stroke or coronary disease or affects cardiovascular or overall mortality

Intervention group:

Step 1: atenolol 100 mg

If blood pressure control was insufficient the further treatment steps were applied:

Step 2: bendrofluazide 5 mg

Step 3: alphamethyldopa 500 mg (and eventually nifedipine retard 20 mg or other antihypertensive medication)

Control group: usual care

n = 884,

intervention group n = 419

control group n = 465

Mean follow up 4.4 years

(range 1–10, SD not reported)

Primary outcomes:

myocardial infarction, major strokes, minor strokes, transient ischaemic attacks,

death

Secondary outcomes:

Congestive heart failure

Heart failure

Atrial fibrillation

Clinical gout

Diabetes

Non- fatal cancer

Vertigo

Dizzy spells

Subgroup analysis in Coope 1986 [34]

See above

To analyse whether the treatment of hypertension in patients between the ages of 70 and 79 years old reduces the incidence of all strokes

See above

Not stated

Not stated

Incidence of all strokes including major strokes, minor strokes, transient ischaemic attacks

Gelber 2013 [45]

Prospective, community- based cohort study

To determine the associations between classes of antihypertensive medication use and the risk of cognitive impairment among elderly hypertensive men.

No drug, BB alone, ACE alone, Diuretic alone, CCB alone, vasodilators alone, BB & 1 other, Other drug combinations

n = 2197

Median follow up 5.8 years (SD 5.1)

Development of cognitive impairment

CASI Score, a combination of the

Hasegawa Dementia Screening Scale, the Folstein Mini-Mental

State Examination, and the Modified Mini-Mental State Examination

Pepine 2003 [32]

RCT

To test the hypothesis, that the risk for adverse outcomes in older people with hypertension and coronary artery disease treated with a calcium antagonist based strategy or a non-calcium antagonist (atenolol) based strategy is equivalent.

Step 1: either

calcium antagonist group:

240 mg/d of verapamil sustained release

or non-calcium antagonist: 50 mg/d of atenolol

If target blood pressure was not achieved, further steps:

Step 2:

calcium antagonist group additional trandolapril

non-calcium antagonist group additional hydrochlorothiazide

Step 3:

dosage increase

Step 4:

calcium antagonist group additional hydrochlorothiazide

non-calcium antagonist group additional trandolapril.

n = 22,576

Calcium antagonist group, n = 11,267 Non-calcium antagonist group, n = 11,309

Mean follow up 2.7 years (range 1 day to 5.4 years, SD not reported)

Primary outcome:

first occurrence of

-death (all cause)

-non-fatal myocardial infarction

-non-fatal stroke

Secondary outcomes:

death (all cause), non-fatal myocardial infarction, non-fatal stroke, time to most serious event (ranked from death as most serious, to myocardial infarction, to stroke as last serious), cardiovascular death, angina, cardiovascular hospitalizations, cancer, Alzheimer disease, Parkinson, gastrointestinal bleeding, blood pressure control, new diagnosis of diabetes, adverse experiences.

Subgroup analysis in Pepine 2003 [32]

See above

See above

See above

Subgroup >70 years old: calcium antagonist group: n = 3694 non- calcium antagonist group: n = 3829

See above

See above

Ruwald 2012 [41]

Post hoc analysis of a double-blind, double-dummy randomized trial

To investigate the influence of age on the effects of losartan versus atenolol-based antihypertensive treatment

Intervention:

losartan 50 mg

(step 2: + HCT 12.5 mg HCT

step 3: losartan 100 mg + HCT 12.5 mg

step 4: losartan 100 mg + HCT 12.5-25 mg HCT + another antihypertensive treatment)

Control:

atenolol 50 mg

(step 2: + HCT 12.5 mg

step 3: atenolol 100 mg + HCT 12.5 mg

step 4: atenolol 100 mg + HCT 12.5-25 mg + another antihypertensive treatment)

n = 9193

n = 4304 < 67 years (46.8%)

n = 4889 ≥ 67 years (53.2%)

Mean follow up 4.8 years

Primary outcome:

Composite of cardiovascular death, nonfatal stroke, nonfatal MI

Testa 2014 [44]

Cross sectional study

To evaluate long-term mortality in hypertensive older adults taking atenolol

No atenolol (Intervention)/atenolol (control)

972 patients aged ≥65 with isolated hypertension

mean age 74.4 ± 6.4

Follow up of 12 years from 1992 to 2003 (not stated if median or mean follow up)

Taking atenolol showed a greater mortality (73.9% vs 55.0%; p = 0.047) and higher pulse arterial pressure values (74.7 vs 63.02 mmHg, p < 0.001) than those not taking atenolol. Atenolol use (HR 1.91; 95% CI: 1.04–4.31; p = 0.04) was predictive of long-term mortality. Pulse

arterial pressure (HR 1.02; 95% CI 1.01–1.03; p = 0.032) was weakly predictive of long-term mortality.

Matsuzaki 2011 [35] Main trial

multicentre, prospective, randomized, open-label, blinded-endpoint trial

To determine the optimal CCB benidipine-based combination therapy (angiotensin-receptor blocker (ARB), beta-blocker or a thiazide) in terms of lowering BP and preventing

cardiovascular events

3 intervention arms:

1. ARB (benidipine 4 mg + angiotensin receptor blocker usual dose)

2. BB (benidipine 4 mg + beta-blocker usual dose)

3. TD (benidipine 4 mg + half daily dose of thiazid diuretic

step 2: benidipine 8 mg (plus ARB/BB/TD)

step 3:benidipine 8 mg plus dose increase of ARB/BB/TD

n = 3293

1. ARB n = 1110 (analysed)

2. BB n = 1089 (analysed)

3. TD n = 1094 (analysed)

n = 1533 ≥ 65 years (46.6%)

n = 1760 < 65 years (53.4%)

Median follow up 3.61 years

Primary outcome:

composite of cardiovascular morbidity

and mortality (sudden death, fatal or non-fatal stroke, fatal or non-fatal

myocardial infarction, hospitalization due to unstable angina, new onset of

heart failure new onset or worsening

of peripheral arterial disease, new onset or worsening of renal failure)

Secondary outcomes:

(1) All-cause mortality

(2) Death from cardiovascular events

(3) Fatal and non-fatal cardiovascular events

(4) New onset of diabetes mellitus

(5) Safety (adverse events and adverse drug reaction)

(6) Nonfatal stroke

Hospitalization due to heart failure (results on outcome not reported)

Ogihara 2012 [40]

COPE trial

post hoc analysis of the COPE trial

To evaluate the efficacy and safety in older (≥65 years) and younger (<65 years) hypertensive patients

MRC 1992 [33]

Main trial

RCT

To establish whether treatment with diuretic or beta blocker in hypertensive older adults reduces risk of stroke, coronary heart disease, and death.

Patients were randomized to atenolol 50 mg daily; hydrochlorothiazide 25 mg or 50 mg plus amiloride 2.5 mg or 5 mg (Moduretic©) daily; or placebo.

n = 4396

Three groups

(a) potassium sparing diuretic regimen (hydrochlorothiazide +amiloride) n = 1081

(b) beta blocker atenolol n = 1102

(c, d) matching placebo tablets

n = 2213

Mean follow up 5.8 years

25,355 patients years of observation (SD not reported)

Strokes, coronary events, and deaths from all causes.

Bird 1990 [36]

RCT, secondary analysis

To explore if cognitive dysfunction, manifested as an increased incidence of confusional states or impaired concentration and attention are associated either with a particular antihypertensive regime or with the reduction in blood pressure level that it induces. To explore if a continuous reduction in mildly elevated blood pressure levels affect the rate of change in cognitive function.

Patients were randomly assigned to hydrochlorothiazide 25 mg and 2.5 mg amiloride, (Moduretic©) daily, atenolol 50 mg daily, or placebo.

n = 2401

hydrochlorothiazide amiloride 24%

atenolol 24%

Placebo 52%

Results reported at month 1 and 9 (median or mean follow up and SD not reported)

Cognitive tests:

Verbal intelligence measured with the Nelson Adult Reading Test (NART),

Non verbal intelligence (performance ability) measured with the Ravens Matrice, part a & b,

Capacity to learn and remember new material measured with the Paired Associate Learning Test (PALT),

Alertness and speed of reaction measured with the Trail-Making Test (TMT),

Depression with the self-rating Depression Questionnaire (DQ) Blood Pressure

Carr 2012

[37]

RCT, secondary analysis

To assess the impact of the blood pressure profile on cardiovascular risk in the Medical Research

Council (UK) elderly trial and to investigate whether the effects of

hypertensive drugs in reducing event rates are solely a

product of systolic pressure reduction.

Intervention:1. atenolol 50 mg daily or

2. hydrochlorothiazide 25 mg (50 mg) + amiloride 2,5 mg (5 mg)

Control:

Placebo

n = 4396

atenolol: n = 1102

hydrochlorothiazide + amiloride: n = 1081 hydrochlorothiazide + Placebo n = 2213

Mean follow up 5.8 years

Primary outcomes:

stroke

CHD

secondary outcomes:

association between BP and outcomes

Lever 1992 [48]

RCT, secondary analysis

To study the effect of two treatments for hypertension on all-cause mortality and morbidity from cardiovascular disease

see above

see above

see above

Stroke

coronary events

all cardiovascular events

all-cause mortality

withdrawal/loss to FU from treatment

Lever 1993

[38]

RCT, secondary analysis

To determine whether hypotensive drug treatment in men and women aged 65–74 reduces stroke, CHD and mortality

see above

see above

see above

Stroke

coronary events

all cardiovascular events

all-cause mortality

change in blood pressure