Angiotensin Receptor Blockers Versus Calcium Channel Blockers for First‐Line Antihypertensive Therapy and Survival in Adults Aged 75 Years or Older

ABSTRACT

Background

Evidence guiding first‐line antihypertensive drug choice in adults aged 75 years or older is limited, despite widespread use of angiotensin receptor blockers (ARBs) and calcium channel blockers (CCBs) in late older age.

Methods

We conducted a target trial emulation using a new‐user design in a nationwide linked healthcare claims database in Japan. Adults aged ≥ 75 years who had no prescription of either drug class in the preceding 12 months and initiated an ARB (n = 10,037) or a CCB (n = 19,785) were followed from treatment initiation. The primary outcome was all‐cause mortality; secondary outcomes included hospitalization for heart failure, myocardial infarction, stroke, and major adverse cardiovascular events (MACE). Intention‐to‐treat effects were estimated using inverse probability of treatment and censoring weighting with pooled logistic regression models.

Results

Among 29,822 patients, median follow‐up was 4.0 years. During follow‐up, 3487 deaths occurred. ARB therapy was associated with lower all‐cause mortality than CCB therapy (hazard ratio [HR], 0.885; 95% CI, 0.823–0.951). The estimated 5‐year risk of death was 12.7% for ARB users and 14.8% for CCB users (absolute risk difference, −2.1 percentage points; 95% CI, −3.1 to −1.0). ARB therapy was also associated with lower risks of heart failure hospitalization (HR, 0.843; 95% CI, 0.774–0.918), myocardial infarction (HR, 0.867; 95% CI, 0.795–0.945), stroke (HR, 0.931; 95% CI, 0.869–0.998), and MACE (HR, 0.889; 95% CI, 0.848–0.931). Associations were consistent across age subgroups, including adults aged ≥ 85 years.

Conclusion

In adults aged 75 years or older, ARB‐based antihypertensive therapy was associated with lower risks of mortality and cardiovascular events compared with CCB‐based therapy. These findings suggest that first‐line antihypertensive drug choice may have prognostic implications in late older age.

Summary

• Key points

  • ○
    In this nationwide target trial emulation of 29,822 adults aged 75 years or older, angiotensin receptor blocker (ARB)‐based therapy was associated with lower all‐cause mortality compared with calcium channel blocker (CCB)‐based therapy (HR 0.885; 5‐year absolute risk difference −2.1 percentage points).
  • ○
    ARB therapy was also associated with lower risks of heart failure hospitalization, myocardial infarction, and major adverse cardiovascular events, with consistent associations among adults aged 85 years or older.
  • ○
    Achieved systolic and diastolic blood pressure levels were similar between treatment strategies, suggesting that differences in outcomes may not be explained solely by blood pressure control.
• Why does this paper matter?

  • ○
    Adults aged 75 years or older are frequently underrepresented in randomized trials of antihypertensive therapy, and ARBs and CCBs are often considered interchangeable first‐line options in routine practice. In this large nationwide study, ARB‐based therapy was associated with modest but clinically meaningful reductions in mortality and heart failure compared with CCB‐based therapy. These findings suggest that antihypertensive drug choice in late older age may have prognostic implications beyond blood pressure lowering and may inform individualized prescribing decisions in geriatric care.

1. Introduction

Hypertension affects the majority of adults aged 75 years or older and remains a leading contributor to cardiovascular morbidity and mortality in this rapidly growing population. Although randomized trials have consistently shown that blood pressure lowering reduces the risks of stroke and other major cardiovascular events across a broad age spectrum, uncertainty persists regarding the optimal first‐line antihypertensive drug class in very old adults [1, 2]. Importantly, even among individuals aged 75 years or older, targeting a systolic blood pressure of < 130 mmHg has been associated with improved cardiovascular outcomes, underscoring the benefit of intensive blood pressure control [3]. However, existing evidence has largely focused on achieved blood pressure levels rather than the comparative effectiveness of specific antihypertensive drug classes, and evidence to guide drug selection in this age group remains limited [4, 5]. Clinical decision making is further complicated by multimorbidity and age‐related alterations in cardiovascular and renal physiology, which may modify both the efficacy and safety profiles of different antihypertensive strategies [6, 7].

Angiotensin receptor blockers (ARBs) and calcium channel blockers (CCBs) are among the most frequently prescribed antihypertensive agents in older adults. Both drug classes effectively lower blood pressure and are recommended as first‐line options in European Society of Cardiology and European Society of Hypertension (ESC/ESH) guidelines [8]. As a result, ARBs and CCBs are often regarded as clinically interchangeable in routine practice once blood pressure targets are achieved. However, these agents differ substantially in their mechanisms of action and downstream biological effects. ARBs modulate the renin–angiotensin system and may influence cardiac remodeling, neurohormonal activation, and volume regulation, whereas CCBs primarily exert vasodilatory effects through calcium channel inhibition [9, 10, 11]. In older adults, these mechanistic differences may translate into clinically meaningful differences in outcomes such as heart failure and mortality, beyond blood pressure reduction alone.

Despite the widespread use of both drug classes, evidence directly comparing ARBs and CCBs in adults aged 75 years or older is limited. Randomized clinical trials of antihypertensive therapy have historically underrepresented very old adults, and subgroup analyses by advanced age are often underpowered or inconclusive [12, 13]. Observational studies have provided valuable insights into antihypertensive effectiveness in broader populations, but many have focused on prevalent users or treatment continuation strategies, which address different clinical questions than those faced at treatment initiation [14, 15]. For older adults newly starting antihypertensive therapy, the comparative effectiveness of ARB‐ versus CCB‐based strategies on survival and major cardiovascular outcomes remains insufficiently characterized.

The framework of target trial emulation offers a principled approach to address such evidence gaps using routinely collected healthcare data when randomized trials are infeasible or unavailable [16, 17]. By explicitly specifying eligibility criteria, treatment strategies, start of follow‐up, outcomes, and causal estimands, this approach allows observational analyses to emulate key features of a hypothetical randomized trial and to reduce common sources of bias, including immortal time bias and inappropriate adjustment for post‐baseline variables. Importantly, a new‐user design within this framework enables evaluation of treatment initiation decisions, which are highly relevant to clinical practice but rarely examined in older populations [14, 15].

Using nationwide healthcare claims data from Japan, we applied a target trial emulation with a new‐user design to compare the ARB‐ versus CCB‐based antihypertensive therapy in adults aged 75 years or older. We focused on all‐cause mortality as the primary outcome and examined major cardiovascular events, including heart failure hospitalization, myocardial infarction, and stroke, as secondary outcomes. To explore potential heterogeneity across late older age, we conducted prespecified subgroup analyses among adults aged 75–79, 80–84, and 85 years or older. By directly addressing the clinical question of antihypertensive drug choice at treatment initiation in late older age, this study aims to provide real‐world comparative effectiveness evidence to inform individualized antihypertensive therapy in this rapidly growing population.

2. Methods

2.1. Study Design and Target Trial Specification

We conducted a retrospective observational study designed to emulate a hypothetical randomized target trial comparing ARB‐based versus CCB‐based antihypertensive therapy in adults aged 75 years or older. The study followed the target trial emulation framework, with explicit specification of eligibility criteria, treatment strategies, start of follow‐up (time zero), outcomes, causal estimands, and analytic methods before analysis [16, 17]. A detailed comparison between the hypothetical target trial and its emulation following the TARGET statement [18] is provided in Table 1.

TABLE 1.

Specification of the target trial and its emulation in the LIFE database.

Component	Target trial (hypothetical randomized trial)	Emulation using observational data
Objective	To compare the effects of initiating ARB‐ versus CCB‐based antihypertensive therapy on mortality and cardiovascular outcomes in adults aged ≥ 75 years	To emulate the same comparison using nationwide healthcare claims data
Eligibility criteria	Adults aged ≥ 75 years with an indication for initiation of antihypertensive therapy; no prior use of ARBs or CCBs; no end‐stage renal disease, dialysis, active malignancy, or palliative care	Adults aged ≥ 75 years who newly initiated an ARB or a CCB after ≥ 12 months of observable baseline data; same exclusion criteria applied using claims‐based definitions
Treatment strategies	Initiation of an ARB‐based antihypertensive strategy Initiation of a CCB‐based antihypertensive strategy	Assignment to ARB or CCB initiation based on first observed dispensing; subsequent treatment modifications allowed
Treatment assignment	Random assignment at treatment initiation	Nonrandomized assignment; adjusted using inverse probability of treatment weighting
Start of follow‐up (time zero)	Date of randomization and treatment initiation	Date of first observed dispensing of an ARB or CCB meeting initiation criteria
Baseline period	Assessment of baseline characteristics immediately before randomization	Assessment of covariates during the 12 months preceding treatment initiation
Outcomes	Primary: All‐cause mortality Secondary: Hospitalization for heart failure, myocardial infarction, stroke, MACE	Outcomes identified using validated ICD‐10 codes from hospitalization and death records
Follow‐up	From treatment initiation until outcome occurrence, death, loss to follow‐up, or end of study	From cohort entry until outcome occurrence, death, loss of data availability, or administrative censoring
Causal estimand	Intention‐to‐treat effect of initiating ARB versus CCB therapy	Intention‐to‐treat effect of ARB versus CCB initiation
Analysis approach	Survival analysis comparing randomized groups, accounting for competing events	Weighted pooled logistic regression with a complementary log–log link to approximate cause‐specific proportional hazard regression model
Confounding control	Achieved by randomization	Achieved by inverse probability of treatment weighting based on baseline covariates
Subgroup analyses	Prespecified analyses by age group	Prespecified analyses by age group (75–79, 80–84, and ≥ 85 years)
Sensitivity to bias	Minimal, assuming successful randomization	Addressed through new‐user design, explicit time zero definition, and covariate balance assessment

Abbreviations: ARB, angiotensin receptor blocker; CCB, calcium channel blocker; IPTW, inverse probability of treatment weighting.

The estimand of interest was the effect of initiating an ARB‐based versus a CCB‐based antihypertensive strategy on mortality and cardiovascular outcomes. Because treatment initiation decisions in late older age represent a clinically distinct question from treatment continuation, we adopted a new‐user design [14, 15] to align treatment assignment, eligibility assessment, and start of follow‐up at the time of antihypertensive initiation and to avoid biases associated with prevalent user designs. This study was approved by the Clinical Research Ethics Committee of Kyushu University (approval number: 226530).

2.2. Data Source

We used data from the Longevity Improvement and Fair Evidence (LIFE) study, a nationwide real‐world database in Japan constructed through systematic linkage of administrative healthcare claims, health check‐up records, and healthcare utilization data provided by participating municipalities [19]. The database includes demographic characteristics, diagnoses coded using the International Classification of Diseases, 10th Revision (ICD‐10) [20], prescription records, hospital admissions, discharge status, and selected laboratory measurements. The structure and format of data are nationally standardized and have been widely used for epidemiologic and health services research [21]. For the present analysis, we used data on 5,054,400 individuals provided by 19 municipalities during the period from April 2014 to September 2024.

2.3. Study Population

Eligible participants were adults aged 75 years or older who newly initiated treatment with either an ARB or a CCB during the study period and had at least 12 months of continuous observable data before treatment initiation to allow assessment of baseline covariates.

We excluded individuals with a history of end‐stage renal disease, dialysis, kidney transplantation, active malignancy, or receipt of palliative or end‐of‐life care before cohort entry. Individuals who initiated ARB and CCB therapy simultaneously were excluded. To avoid mechanistic overlap between treatment strategies, participants in the CCB group who were receiving angiotensin‐converting enzyme (ACE) inhibitors at the time of treatment initiation were also excluded.

Cohort entry (time zero) was defined as the date of first observed dispensing of an ARB or a CCB that met criteria for treatment initiation after the baseline period. Baseline covariates were ascertained using information from the 12 months preceding cohort entry. Information on prior intolerance to ARBs or CCBs was not available in the database and therefore could not be used as an exclusion criterion.

2.4. Treatment Strategies

Two antihypertensive treatment strategies were compared:

1. ARB treatment strategy: antihypertensive therapy with an ARB.

2. CCB treatment strategy: antihypertensive therapy with a CCB.

Concomitant use of other antihypertensive medications, including diuretics and β‐blockers, was permitted at baseline to reflect routine clinical practice in older adults. The treatment contrast of interest was ARB‐ versus CCB‐based therapy rather than specific drug doses or subsequent treatment modifications. ARB therapy was selected as the comparator to CCB therapy because ARBs are more commonly prescribed than ACE inhibitors in routine clinical practice in Japan, particularly among older adults, partly due to better tolerability. The CCB strategy predominantly comprised dihydropyridine CCBs; a detailed breakdown of individual agents is provided in Table S1.

2.5. Outcomes and Follow‐Up

The primary outcome was all‐cause mortality. Secondary outcomes included hospitalization for heart failure, myocardial infarction, and stroke, defined using ICD‐10 diagnosis codes recorded during hospital admissions.

Follow‐up began at cohort entry and continued until occurrence of the outcome of interest, death, loss of data availability, or the end of the study period, whichever occurred first.

2.6. Causal Estimand

The primary estimand was the intention‐to‐treat (ITT) effect of initiating an ARB‐based versus a CCB‐based antihypertensive strategy. This estimand represents the effect of treatment assignment at initiation, regardless of subsequent treatment changes, and reflects real‐world clinical decision making in older adults, in whom treatment modifications over time are common [22]. As a secondary estimand, we evaluated the per‐protocol (PP) effect, which represents the effect of sustained adherence to the initially assigned treatment strategy, to assess the robustness of the primary findings under alternative assumptions about treatment adherence [23]. In the PP analysis, treatment deviation was defined as initiation of the alternative antihypertensive drug class, resulting in combination therapy with both an ARB and a CCB. In addition, initiation of an ACE inhibitor in the CCB group was considered a protocol deviation.

2.7. Statistical Analysis

Follow‐up time was divided into one‐month intervals and analyzed using pooled logistic regression models with a complementary log–log link to approximate discrete‐time proportional hazards models [24]. For outcomes other than all‐cause mortality and major adverse cardiovascular events (MACE), death was treated as a competing event, and cause‐specific hazard models were fitted [25]. Temporal variation in the baseline hazard was modeled using calendar time as a continuous variable with spline terms.

To adjust for baseline confounding, inverse probability of treatment weighting (IPTW) was applied using propensity scores estimated from baseline covariates measured at cohort entry [26]. Covariates were selected a priori based on clinical relevance and included age, sex, comorbid conditions, prior cardiovascular disease, markers of healthcare utilization, and concomitant medications. Covariate balance between treatment groups was assessed using standardized mean differences, with values below 0.10 indicating adequate balance [27].

To account for informative censoring, inverse probability of censoring weights (IPCW) [26, 28] were applied. In the PP analysis, IPCW was additionally used to adjust for artificial censoring due to protocol deviation using a cloning‐censoring‐weighting approach [23, 29]. Time‐varying covariates related to heart failure risk and treatment intensity were incorporated into the censoring models as appropriate. A complete list of covariates and their definitions is provided in Table S2. Stabilized weights were consistently used and truncated at the 1st and 99th percentiles to reduce the influence of extreme values [28].

Hazard ratios (HRs) and 95% confidence intervals (CIs) were estimated using weighted pooled logistic regression models with robust variance estimators. Weighted cumulative incidence functions accounting for the competing risk of death were estimated to visualize absolute risks over time, and 5‐year cumulative event probabilities were derived [22, 25]. To explore potential mechanisms underlying the observed treatment effects, we evaluated longitudinal blood pressure trajectories in the weighted analytic population. Using combined IPTW and IPCW, we calculated year‐specific weighted mean systolic and diastolic blood pressure values and compared patterns over follow‐up between ARB‐ and CCB‐based strategies. Subgroup analyses were conducted according to age group (75–79, 80–84, and ≥ 85 years), baseline history of heart failure, coronary heart disease, and kidney function (estimated glomerular filtration rate [eGFR] ≥ 60 vs. < 60 mL/min/1.73 m²).

Missing covariate data were handled using multiple imputation by chained equations, generating 100 imputed datasets that were analyzed separately and combined using Rubin's rules [30]. All analyses were performed using R ver. 4.5.1 (R Foundation for Statistical Computing, Vienna, Austria).

3. Results

3.1. Study Population

Using the target trial emulation with a new‐user design, we identified 29,822 adults aged 75 years or older who newly initiated antihypertensive therapy with either an ARB (n = 10,037) or a CCB (n = 19,785). The mean age was 81.7 years in the ARB group and 81.8 years in the CCB group, and 57% of participants were women (Table 2). Life expectancy in Japan is approximately 81 years for men and 87 years for women, indicating that the study population represents individuals in late life. Information on race and ethnicity is not available in the LIFE database; however, the Japanese population is relatively homogeneous, with the vast majority identifying as Japanese. The distribution of the study population across calendar years reflects changes in the number of participating municipalities over time. The median duration of follow‐up was 4.00 years (interquartile range, 1.92–6.25).

TABLE 2.

Patient characteristics of the ARB and CCB groups.

Characteristics	Original data			IPTW‐adjusted data
	ARB (N = 10,037)	CCB (N = 19,785)	SMD	ARB (N = 10,037)	CCB (N = 19,785)	SMD
Age, years	81.7 (5.0)	81.8 (5.0)	0.03	81.8 (5.0)	81.8 (5.0)	0.00
Gender, female	5742.0 (57.2%)	11304.7 (57.1%)	0.00	5708.6 (56.9%)	11288.5 (57.1%)	0.00
BMI, kg/m2	22.7 (3.4)	22.5 (3.3)	0.05	22.5 (3.3)	22.5 (3.3)	0.00
SBP, mmHg	140.3 (18.6)	140.6 (19.0)	0.02	140.5 (18.6)	140.5 (19.0)	0.00
DBP, mmHg	76.7 (11.4)	76.5 (11.3)	0.01	76.6 (11.4)	76.6 (11.3)	0.00
LDL cholesterol, mg/dL	117.8 (29.7)	119.4 (30.8)	0.05	118.9 (29.7)	118.9 (30.9)	0.00
eGFR, mL/min/1.73 m2	63.8 (35.2)	64.9 (36.6)	0.04	64.3 (37.3)	64.5 (34.6)	0.01
Proteinuria	2463.4 (24.5%)	4707.3 (23.8%)	0.02	2442.1 (24.3%)	4775.8 (24.1%)	0.00
Smoking	600.3 (6.0%)	1379.5 (7.0%)	0.04	666.6 (6.6%)	1314.5 (6.6%)	0.00
Alcohol consumption	3228.8 (32.2%)	6496.9 (32.8%)	0.02	3275.3 (32.6%)	6443.3 (32.6%)	0.00
Diabetes	4879.2 (48.6%)	8870.5 (44.8%)	0.08	4631.2 (46.1%)	9110.9 (46.0%)	0.00
Chronic liver disease	2573.2 (25.6%)	4931.4 (24.9%)	0.02	2544.3 (25.3%)	4993.8 (25.2%)	0.00
Dyslipidemia	6379.5 (63.6%)	11419.9 (57.7%)	0.12	5996.8 (59.7%)	11805.3 (59.7%)	0.00
Diuretics use	1256.1 (12.5%)	973.6 (4.9%)	0.27	761.6 (7.6%)	1450.4 (7.3%)	0.01
β‐blocker use	748.3 (7.5%)	1360.1 (6.9%)	0.02	714.3 (7.1%)	1405.8 (7.1%)	0.00
α‐blocker use	129.8 (1.3%)	308.4 (1.6%)	0.02	151.3 (1.5%)	293.4 (1.5%)	0.00
Statin use	3814.0 (38.0%)	7012.5 (35.4%)	0.05	3650.3 (36.4%)	7189.2 (36.3%)	0.00
Antidiabetic use	1221.7 (12.2%)	1735.5 (8.8%)	0.11	1014.1 (10.1%)	1963.7 (9.9%)	0.01
Antiplatelet use	2031.1 (20.2%)	3864.8 (19.5%)	0.02	2010.3 (20.0%)	3941.5 (19.9%)	0.00
Anticoagulant use	702.5 (7.0%)	1118.6 (5.7%)	0.06	630.3 (6.3%)	1211.8 (6.1%)	0.01
Prior CHD	3470.4 (34.6%)	6321.2 (31.9%)	0.06	3313.2 (33.0%)	6492.1 (32.8%)	0.00
Prior stroke	3726.2 (37.1%)	7726 (39.0%)	0.04	3868.6 (38.5%)	7624.8 (38.5%)	0.00
Prior heart failure	3313.9 (33.0%)	5722.9 (28.9%)	0.09	3053.9 (30.4%)	5985.0 (30.3%)	0.00
Calendar year
2016	244.3 (2.5%)	367.9 (1.8%)	0.02	217.7 (2.1%)	421 (2.1%)	0.00
2017–2019	5680.9 (56.6%)	10,030 (50.7%)	0.12	5301.5 (52.8%)	10417.7 (52.7%)	0.00
2020–2022	2638.1 (26.3%)	6421.6 (32.5%)	0.14	3056.8 (30.5%)	6029.1 (30.5%)	0.00
2023	1473.7 (14.7%)	2965.5 (15.0%)	0.01	1461.0 (14.6%)	2917.2 (14.7%)	0.01

Note: Continuous data were presented as means and standard deviations, while categorical data were reported as counts and percentages. These descriptive statistics were calculated by pooling the results from 100 imputed datasets generated through multiple imputation by chained equations. To assess the effectiveness of the inverse probability of treatment weighting based on baseline covariates, standardized mean differences (SMDs) were employed to evaluate the balance between the treatment groups in the weighted pseudo‐population.

Abbreviations: ARB, angiotensin receptor blocker; BMI, body mass index; CCB, calcium channel blocker; CHD, coronary heart disease; DBP, diastolic blood pressure; eGFR, estimated glomerular filtration rate; LDL, low‐density lipoprotein cholesterol; SBP, systolic blood pressure.

Before weighting, several baseline characteristics differed between treatment groups, reflecting routine prescribing patterns in older adults, including use of diuretics and the prevalence of dyslipidemia. After application of IPTW, all measured baseline covariates were well balanced between groups, with standardized mean differences below 0.10 for all variables (Table 2). Missing data were present for some baseline variables, particularly eGFR, smoking, and alcohol consumption. Baseline concomitant medications among patients with prior heart failure and prior myocardial infarction are summarized in Tables S3 and S4.

3.2. Primary Outcome: All‐Cause Mortality

During follow‐up, 3487 deaths occurred in the ITT analysis (1141 in the ARB group and 2346 in the CCB group). ARB therapy was associated with a lower risk of all‐cause mortality compared with CCB therapy (HR, 0.885; 95% CI, 0.823–0.951) (Table 3).

TABLE 3.

Comparative effectiveness of ARB‐ versus CCB‐based treatment strategies for mortality, cardiovascular, and renal outcomes.

Outcome	No. of events		Hazard ratio (95% CI)	5‐year risk (%)		5‐year risk difference (95% CI)
	ARB (N = 10,037)	CCB (N = 19,785)		ARB	CCB
All‐cause death
Intention‐to‐treat	1141	2346	0.885 (0.823 to 0.951)	0.127	0.148	−0.021 (−0.031 to −0.010)
Per‐protocol	37	213	0.385 (0.267 to 0.557)	0.015	0.033	−0.018 (−0.029 to −0.007)
Heart failure hospitalization
Intention‐to‐treat	837	1726	0.843 (0.774 to 0.918)	0.091	0.105	−0.014 (−0.022 to −0.005)
Per‐protocol	378	950	0.725 (0.640 to 0.821)	0.105	0.134	−0.029 (−0.054 to −0.004)
Myocardial infarction
Intention‐to‐treat	788	1686	0.867 (0.795 to 0.945)	0.090	0.101	−0.011 (−0.019 to −0.003)
Per‐protocol	398	1029	0.796 (0.706 to 0.897)	0.123	0.144	−0.021 (−0.047 to 0.005)
Stroke
Intention‐to‐treat	1234	2594	0.931 (0.869 to 0.998)	0.145	0.153	−0.008 (−0.018 to 0.002)
Per‐protocol	714	1771	0.880 (0.805 to 0.962)	0.197	0.209	−0.012 (−0.041 to 0.017)
MACE
Intention‐to‐treat	2666	5633	0.889 (0.848 to 0.931)	0.306	0.338	−0.032 (−0.045 to −0.018)
Per‐protocol	1068	2748	0.822 (0.764 to 0.883)	0.301	0.335	−0.034 (−0.070 to 0.002)
eGFR decline ≥ 40%
Intention‐to‐treat	52	79	1.110 (0.773 to 1.593)	0.006	0.006	0.000 (−0.002 to 0.003)
Per‐protocol	16	26	1.200 (0.634 to 2.272)	0.005	0.005	−0.001 (−0.005 to 0.004)
ESRD/Dialysis
Intention‐to‐treat	20	48	0.611 (0.354 to 1.056)	0.002	0.003	−0.001 (−0.002 to 0.001)
Per‐protocol	3	18	0.276 (0.079 to 0.962)	0.000	0.004	−0.004 (−0.007 to 0.000)
Hypotension
Intention‐to‐treat	38	132	0.515 (0.355 to 0.748)	0.005	0.008	−0.003 (−0.005 to −0.001)
Per‐protocol	8	89	0.166 (0.079 to 0.350)	0.001	0.010	−0.009 (−0.013 to −0.005)
Electrolyte disorders
Intention‐to‐treat	314	590	0.959 (0.834 to 1.104)	0.035	0.036	−0.001 (−0.006 to 0.004)
Per‐protocol	143	311	0.919 (0.749 to 1.127)	0.040	0.046	−0.007 (−0.021 to 0.009)

Note: Hazard ratios were estimated using weighted pooled logistic regression models with robust variance estimators. 5‐year risks were estimated from weighted cumulative incidence functions.

From an absolute risk perspective, the estimated 5‐year cumulative incidence of death was 12.7% in the ARB group and 14.8% in the CCB group, corresponding to an absolute risk difference of −2.1 percentage points (95% CI, −3.1 to −1.0), corresponding to a number needed to treat of 48. Weighted cumulative incidence curves demonstrated early and persistent separation between treatment strategies, with consistently lower mortality observed among ARB users throughout follow‐up (Figure 1A). In secondary PP analyses, effect estimates were directionally consistent with the primary ITT findings.

FIGURE 1.

Weighted cumulative incidence functions comparing ARB‐ and CCB‐based treatment strategies. Panels show cumulative incidence of (A) all‐cause death, (B) heart failure hospitalization, (C) myocardial infarction, and (D) stroke over follow‐up. Estimates were obtained using discrete‐time proportional hazards regression models with inverse probability of treatment and censoring weighting. Death was treated as a competing event for nonfatal outcomes. Abbreviations: ARB, angiotensin receptor blocker; CCB, calcium channel blocker.

3.3. Secondary Outcomes

3.3.1. Heart Failure Hospitalization

Hospitalization for heart failure occurred in 2563 participants in the ITT analysis (837 in the ARB group and 1726 in the CCB group). ARB therapy was associated with a lower risk of heart failure hospitalization compared with CCB therapy (HR, 0.843; 95% CI, 0.774–0.918) (Table 3).

The estimated 5‐year risk of heart failure hospitalization was 9.1% in the ARB group and 10.5% in the CCB group, yielding an absolute risk difference of −1.4 percentage points (95% CI, −2.2 to −0.5). Weighted cumulative incidence curves showed a lower cumulative incidence of heart failure hospitalization in the ARB group over time (Figure 1B).

3.3.2. Myocardial Infarction and Stroke

For myocardial infarction, 2474 events occurred in the ITT analysis (788 in the ARB group and 1686 in the CCB group). ARB therapy was associated with a lower risk compared with CCB therapy (HR, 0.867; 95% CI, 0.795–0.945), with a 5‐year absolute risk difference of −1.1 percentage points (95% CI, −1.9 to −0.3) (Table 3). Cumulative incidence curves showed lower incidences among ARB users (Figure 1C).

For stroke, 3828 events were observed (1234 in the ARB group and 2594 in the CCB group). ARB therapy was associated with a small reduction in risk compared with CCB therapy (HR, 0.931; 95% CI, 0.869–0.998), corresponding to a 5‐year absolute risk difference of −0.8 percentage points (95% CI, −1.8 to 0.2) (Table 3). Cumulative incidence curves were closely aligned between treatment groups (Figure 1D).

3.3.3. Major Adverse Cardiovascular Events

For MACE, 8299 events occurred in the ITT analysis (2666 in the ARB group and 5633 in the CCB group). ARB therapy was associated with a lower risk of MACE compared with CCB therapy (HR, 0.889; 95% CI, 0.848–0.931), with a 5‐year absolute risk difference of −3.2 percentage points (95% CI, −4.5 to −1.8) (Table 3).

3.3.4. Renal and Safety Outcomes

Renal outcomes were infrequent. For a sustained decline in eGFR of at least 40%, 131 events occurred in the ITT analysis, with no clear difference between treatment strategies (HR, 1.110; 95% CI, 0.773–1.593). Progression to end‐stage renal disease or dialysis was rare (68 events), and estimates favored ARB therapy but were imprecise (HR, 0.611; 95% CI, 0.354–1.056) (Table 3).

Hospitalization for hypotension was uncommon but occurred less frequently among ARB users than CCB users (HR, 0.515; 95% CI, 0.355–0.748). Electrolyte disorders occurred at similar rates in the two groups (HR, 0.959; 95% CI, 0.834 to 1.104), with low absolute risks (Table 3).

3.3.5. Longitudinal Trajectories of Systolic and Diastolic Blood Pressure

When evaluated in the weighted analytic population, longitudinal mean systolic and diastolic blood pressure values were closely aligned between treatment strategies, with consistently trivial between‐group differences (Table S5). Annual weighted mean differences in systolic blood pressure remained below 1 mmHg. These results indicate that mechanisms beyond achieved blood pressure levels likely contribute to the differences in heart failure and renal outcomes.

3.4. Subgroup Analyses

Prespecified subgroup analyses by age group (75–79, 80–84, and ≥ 85 years) demonstrated generally consistent associations between treatment strategy and outcomes (Figure 2). For all‐cause mortality, HRs favored ARB therapy across age strata, including adults aged 85 years or older (HR, 0.859; 95% CI, 0.778–0.949). Similar patterns were observed for heart failure hospitalization, myocardial infarction, stroke, and MACE, with no strong evidence of effect modification by age. Associations were broadly consistent across subgroups defined by prior heart failure, prior coronary heart disease, and baseline kidney function (Tables S6 and S7).

FIGURE 2.

Subgroup analyses comparing ARB‐ and CCB‐based treatment strategies for clinical outcomes across age groups. Hazard ratios (HRs) and 95% confidence intervals for the association between ARB versus CCB therapy and clinical outcomes across age subgroups (75–79, 80–84, and ≥ 85 years). HR estimates were obtained from intention‐to‐treat analyses using weighted pooled logistic regression, accounting for competing risk of death. Inverse probability of treatment weighting was applied to balance baseline covariates between treatment groups within each subgroup. Follow‐up began at treatment initiation and continued until outcome occurrence, loss to follow‐up, or the end of the study period. Abbreviations: ARB, angiotensin receptor blocker; CCB, calcium channel blocker; HR, hazard ratio.

4. Discussion

In this nationwide observational study emulating a target trial, ARB–based antihypertensive therapy was associated with lower risks of all‐cause mortality and heart failure hospitalization compared with CCB–based therapy among adults aged 75 years or older. These associations were observed under an ITT framework and were consistent across prespecified age subgroups, including adults aged 85 years or older. ARB therapy was also associated with reductions in myocardial infarction and stroke, although these effects were smaller in magnitude than those observed for mortality and heart failure, suggesting a differential pattern of treatment effects across clinical endpoints in late older age. Renal outcomes were infrequent. PP analyses yielded results directionally consistent with the primary ITT analyses, supporting the robustness of the findings under assumptions of sustained treatment adherence.

The most clinically important finding of this study is the association between ARB therapy and improved survival in a population characterized by advanced age and substantial multimorbidity. In older adults, mortality is driven not only by atherosclerotic cardiovascular disease but also by heart failure, arrhythmias, and noncardiovascular conditions, many of which are influenced by neurohormonal activation and alterations in volume homeostasis [31, 32]. ARBs inhibit the renin–angiotensin system and may mitigate maladaptive cardiac remodeling, fluid retention, and neurohormonal dysregulation—mechanisms that are particularly relevant in late older age [9, 10, 32]. In contrast, CCBs primarily exert vasodilatory effects and have been associated with peripheral edema, potentially triggering prescribing cascades involving loop diuretics [33]. Such cascades may increase susceptibility to volume depletion, renal dysfunction, and heart failure in older adults. The reduction in heart failure hospitalization observed among ARB users, together with lower mortality, is therefore biologically plausible. These observations are broadly consistent with prior evidence suggesting that renin–angiotensin system inhibition may be associated with improved cardiovascular outcomes and survival. However, most previous studies have focused on selected populations or have not directly compared commonly used first‐line antihypertensive drug classes in very old adults, and evidence remains limited in this population [34, 35]. Importantly, these associations were consistent across age subgroups, including adults aged 85 years or older, suggesting that the potential benefits of ARB‐based therapy may extend beyond blood pressure lowering even in very late life.

In addition to reductions in mortality and heart failure hospitalization, ARB–based therapy was associated with significant reductions in myocardial infarction and stroke. Although these effects were smaller in magnitude than those observed for heart failure–related outcomes, they are consistent with prior evidence suggesting that blood pressure reduction remains a key determinant of atherosclerotic cardiovascular risk across antihypertensive drug classes [1, 2]. CCBs are effective antihypertensive agents and have been widely used for stroke prevention, particularly in older populations. However, the observed reductions in myocardial infarction and stroke among ARB users suggest that the benefits of ARB‐based strategies in late older age may extend beyond heart failure–related pathways and include a modest but measurable reduction in atherosclerotic cardiovascular events.

An important strength of this study is its focus on treatment initiation decisions in late older age. Older adults are underrepresented in randomized clinical trials of antihypertensive therapy, and evidence guiding first‐line drug choice in those aged 75 years or older—particularly those aged 85 years or older—is limited [8, 12, 13]. By adopting a new‐user design within a target trial emulation framework, this study aligns eligibility assessment, treatment assignment, and start of follow‐up at the time of antihypertensive initiation, thereby addressing a clinically relevant question that is distinct from treatment continuation or switching. The consistency of findings across age strata further suggests that the observed associations are not driven solely by younger subsets of the older population.

From a clinical perspective, these findings have potential implications for antihypertensive drug selection in late older age. Current guidelines generally position ARBs and CCBs as acceptable first‐line options for older adults without specific indications such as chronic kidney disease or proteinuria [8, 36]. The present results suggest that, when initiating antihypertensive therapy in adults aged 75 years or older, ARB‐based strategies may offer advantages in terms of survival and heart failure prevention, without evidence of increased harm. Importantly, this interpretation does not imply that CCBs are inappropriate or ineffective, but rather that ARBs may confer additional prognostic benefits in this age group that merit consideration alongside blood pressure targets, comorbidities, and patient preferences.

This study has several limitations. As with all observational analyses, residual confounding due to unmeasured factors cannot be completely excluded, despite extensive adjustment using inverse probability weighting and careful specification of the target trial [22, 23, 28]. Information on frailty, functional status, and blood pressure variability was limited in the available data and may have influenced treatment selection and outcomes. Misclassification of diagnoses and outcomes is possible when using administrative claims data, although such misclassification is unlikely to differ systematically between treatment groups [37]. Renal outcomes were infrequent, resulting in limited statistical precision for these endpoints. Adherence to antihypertensive therapy could not be directly measured in this study; however, in Japan, these medications are generally covered under the universal health insurance system, which may facilitate access and sustained use, particularly among older adults. Finally, the study population consisted of older adults receiving care within the Japanese healthcare system, and generalizability to other healthcare settings may be limited.

In conclusion, in this target trial emulation using nationwide real‐world data, ARB‐based antihypertensive therapy in adults aged 75 years or older was associated with lower risks of all‐cause mortality and heart failure hospitalization compared with CCB‐based therapy, with smaller but significant reductions in risks of myocardial infarction and stroke. These findings provide comparative effectiveness evidence directly relevant to antihypertensive treatment decisions in late older age and support consideration of ARBs as a preferred option when initiating therapy in this rapidly growing population.

Author Contributions

H.N. conducted statistical analyses and wrote the initial draft of the manuscript. H.N. and H.F. conceptualized the research question and study design. F.O., M.M. and H.F. conducted data management and database development. H.S., T.S. and K.‐E.S. supervised study design, statistical methods and data interpretation, and wrote/revised the manuscript. All authors approved the final manuscript.

Funding

This study was supported by the Japan Society for the Promotion of Science (JP24K21306, JP23K11931 and JP22H03554) and the Japan Science and Technology Agency (JPMJFR205J).

Disclosure

The funders had no role in considering the study design or in the collection, analysis, interpretation of data, writing of the report, or decision to submit the article for publication.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Table S1: Distribution of calcium channel blocker subclasses and individual agents.

Table S2: Adjustment variables involved in the IPTW and IPCW estimation models.

Table S3: Baseline concomitant medications among patients with prior heart failure.

Table S4: Baseline use of antiplatelet agents, statins, and beta‐blockers among patients with prior myocardial infarction.

Table S5: Longitudinal trajectories of systolic and diastolic blood pressure for IPTW–IPCW–adjusted pseudo‐populations.

Table S6: Subgroup analyses comparing ARB and CCB treatment strategies (outcome: all‐cause death).

Table S7: Subgroup analyses comparing ARB and CCB treatment strategies (outcome: heart failure hospitalization).