Introduction

Hypertension is a significant global health concern, particularly in China and is a major risk factor for cardiovascular diseases (CVDs) and mortality [1, 2]. Systematic reviews and meta-analyses have shown that lowering blood pressure with antihypertensive drugs effectively reduces major cardiovascular events and mortality [3,4,5]; this benefit is primarily attributed to the extent of blood pressure reduction rather than the specific properties of individual drug classes [6]. However, different classes of antihypertensive drugs have been reported to exert distinct effects on specific cardiovascular outcomes [7,8,9,10,11,12].

Despite extensive research, evidence on comparative effectiveness of different antihypertensive drug classes remains inconsistent, and clinical guidelines vary significantly. The 2017 American College of Cardiology/American Heart Association Blood Pressure Treatment Guidelines [13] recommended thiazide or thiazide-like diuretics, angiotensin-converting enzyme inhibitors (ACEIs), angiotensin II receptor blockers (ARBs), or calcium channel blockers (CCBs) as first-line agents. The 2023 European Society of Hypertension Guidelines provide a comparable broad recommendation, but also include beta-blockers as first-line option [14]. The NICE (National Institute for Health and Care Excellence) guideline recommends ACEIs or ARBs as first-line treatment for patients with type 2 diabetes, or for adults younger than 55 years who are not of Black African or African Caribbean family origin; for all other patient groups, CCBs are recommended [15]. These inconsistencies highlight the need for further research to identify which antihypertensive drug class provides the most favourable cardiovascular outcomes and should be prioritised as the optimal first-line antihypertensive therapy. Therefore, this study aimed to investigate whether prolonged exposure to specific antihypertensive drug classes is associated with a reduced risk of composite cardiac outcome in individuals with well-controlled blood pressure, using data from the STEP (Strategy of Blood Pressure Intervention in the Elderly Hypertensive Patients) trial.

Methods

Study population

We conducted a post hoc analysis of the STEP trial, as described in more detail in previously published studies [16, 17]. Briefly, the STEP trial was an open-label, multicentre, randomised controlled trial conducted between January 10, 2017 and December 31, 2020. The trial included 8511 participants with hypertension aged 60–80 years in China, all of whom had no history of ischaemic or haemorrhagic stroke. Eligible individuals were randomly assigned to either the intensive or standard treatment group. The study defined two systolic blood pressure (SBP) targets: an intensive treatment range of 110 to less than 130 mm Hg and a standard treatment range of 130 to less than 150 mm Hg. All participants were scheduled for monthly follow-up visits during the first 3 months and subsequently every 3 months until the final visit or death. For the present analysis, we excluded participants who were lost to follow-up during the study (N = 234) and those with missing post-randomisation blood pressure records (N = 20) (Additional file 1: Fig. S1).

Exposure and clinical outcomes

In the STEP trial, participants received free antihypertensive medications upon enrolment, including olmesartan (an ARB), amlodipine (a CCB), and hydrochlorothiazide (a diuretic). At the start of the trial, participants were initiated on olmesartan-medoxomil tablets as the preferred ARB or amlodipine tablets as the preferred CCB. Hydrochlorothiazide was not recommended as a primary monotherapy for the initial treatment. Other antihypertensive medications, such as beta-blockers, were allowed at the discretion of the investigators when deemed necessary and appropriate. Medication adjustments (dose increases and additional antihypertensive therapy) to achieve target SBP were based on office blood pressure measurements, which were evaluated monthly for the first 3 months and every 3 months thereafter. In the present study, we focused on four main classes of antihypertensive medications, categorised based on their pharmacological mechanisms: ARBs, CCBs, thiazide-type diuretics, and beta-blockers.

The primary exposure of interest was the relative time for each class of antihypertensive medication, defined as the ratio of medication exposure time to event time. Medication exposure time was calculated as the number of days from the first prescription date to the discontinuation date during the follow-up period, the first event, or the end of the study, whichever came first. If a particular class of antihypertensive drug was discontinued at a clinical visit but reinitiated at a later visit, the days of exposure were added to the previous count. The visit data were used to calculate the medication time for the four medication classes.

The primary outcome of this analysis was consistent with that used in the STEP trial [17] and included a composite of the first occurrence of stroke (ischaemic or haemorrhagic), acute coronary syndrome (ACS; including acute myocardial infarction and hospitalisation for unstable angina), acute decompensated heart failure, coronary revascularisation, atrial fibrillation, and death from cardiovascular causes. Secondary outcomes included the individual components of the primary outcome, all-cause mortality, and major adverse cardiac events (MACE) (a composite of the individual components of the primary outcome, excluding stroke).

Covariate measurements

Participants’ office blood pressure measurements were taken by trained staff after a 5-min rest, with an average of three measurements recorded per visit. To ensure standardised measurement procedures, the same validated office sphygmomanometer (Omron Healthcare) was used across all study centres during enrolment and follow-up. Antihypertensive regimens before and after each visit were recorded in the STEP database. Laboratory examinations were performed at baseline and annually thereafter. Socio-demographic characteristics were collected through face-to-face interviews using standardised questionnaires administered by trained STEP physicians. Current smokers were defined as individuals who smoked at least one cigarette per day for more than 6 months. Current alcohol drinkers were defined as those who consumed alcohol at least once per month for more than 6 months. Body mass index was calculated as weight (kg) divided by height (m) squared. The estimated glomerular filtration rate (eGFR) was calculated using the Modification of Diet in Renal Disease equation, with renal dysfunction defined as an eGFR < 60 ml/min/1.73 m2 [18].

Statistical methods

Continuous variables were assessed for normality and are presented as mean ± standard deviation, while categorical variables are presented as n (%). Baseline characteristics of the study sample were summarised by specific antihypertensive medication use and compared using a one-way analysis of variance for continuous variables and the chi-square (χ2) test for categorical variables. The relative time to the event of interest (medication time/event time) was calculated to account for the different survival times of the participants. Subsequently, the relative time was included as a continuous variable in the Cox regression analysis. Cox proportional hazards regression models were used to calculate the hazard ratio (HR) and 95% confidence interval (CI) for the primary and secondary outcomes of relative time for each antihypertensive drug class. The HRs were calculated per 1 unit increase (representing full-time drug use vs. not use) and per 0.1 unit increase (representing a 10% longer duration of use) in relative time to assess the impact of changes in relative time on CVD risk. Three models were used in the analysis: model 1: unadjusted (crude) model. Model 2: adjusted for randomisation group, age, sex, body mass index, serum glucose, low-density lipoprotein cholesterol, baseline SBP, and cumulative SBP. Model 3: extended model 2 by further adjusting for history of CVD and baseline renal dysfunction. The HRs for ARBs and CCBs were compared using a two-independent-samples t-test based on bootstrapped estimates (n = 100) [19]. Sensitivity analyses were conducted to assess the robustness of the findings by excluding participants with a history of CVD or an eGFR < 60 ml/min/1.73 m2 at baseline. Considering that beta-blockers are often prescribed to patients with pre-existing cardiovascular conditions, a propensity score-matched analysis was applied to the relative time of beta-blockers to control for residual confounders and examine cardiovascular risk. All the analyses were performed using SAS (version 9.4; SAS Institute Inc., Cary, NC, USA). Statistical significance was set at P < 0.05. We did not adjust the CI or P values for multiple comparisons, and the findings were considered exploratory.

Results

A total of 8257 patients were included in the analysis, and their baseline characteristics across different antihypertensive drug classes are summarised in Table 1. The percentage of each antihypertensive drug class regimen among study participants at baseline and follow-up visits for both the intensive and standard treatment groups was calculated and is presented in Additional file 1: Fig. S2. Among the participants in the standard group, major adjustments to antihypertensive drugs were completed within 6 months after randomisation and remained stable throughout the study period. In contrast, among the participants in the intensive treatment group, the use of ARBs, CCBs, and thiazide-type diuretics continued to increase beyond the initial 6-month adjustment period. Between baseline and 36 months of follow-up, the proportion of ARB use in the intensive group increased by 19.48%, while CCB use increased by 12.63%, and diuretic use by 10.86%, respectively. In the standard treatment group, the use of ARBs increased by 9.58%, whereas the use of thiazide-type diuretics and CCBs remained relatively stable. The frequency of beta-blocker regimens used in both the intensive and standard treatment groups remained stable from the baseline to the end of the study.

Table 1 Baseline characteristics of participants using different antihypertensive drug classes
Full size table

Primary outcome

During a median follow-up of 3.34 years, 336 primary outcome events occurred. We analysed cardiovascular risk using the relative time for each drug class until the occurrence of the primary outcome (Fig. 1 and Table 2). After adjusting for potential confounders, multivariate Cox proportional hazard regression analyses indicated that a longer relative time to ARB and CCB treatment was associated with a significantly lower risk of the primary outcome. Each unit increase in the relative time of ARB corresponded to a 45% reduction in CVD risk (HR 0.55, 95% CI 0.43–0.70), compared to a 30% reduction with CCBs use (HR 0.70, 95% CI 0.54–0.92). The risk reduction associated with ARBs was significantly higher than that observed with CCBs (Fig. 1; P < 0.001). Moderate associations were observed between the relative duration of thiazide-type diuretic use and primary outcome. In contrast, each unit increase in the relative time on beta-blockers was associated with a higher risk of primary events (HR 2.20, 95% CI 1.81–2.68). To further interpret the result and assess their applicability, we calculated the HR per 0.1 unit change in relative time (Fig. 1 and Table 2) and found that an increase of 0.1 unit in ARBs’ relative time was associated with a 6% reduction in the risk of primary outcome (HR 0.94, 95% CI 0.92–0.96), whereas CCBs reduced the risk by only 3% (HR 0.97, 95% CI 0.94–0.99). Additionally, each 0.1-unit increase in the relative time on beta-blockers was associated with an 8% increase in the risk of the primary outcome (HR 1.08, 95% CI 1.06–1.10). A significant increase in cardiovascular risk associated with beta-blockers was also observed in the propensity score-matched analysis after controlling for several residual confounders (Additional file 1: Table S1).

Fig. 1
figure 1

Association of antihypertensive drug class exposure with primary composite outcome† in the entire study population. The relative exposure time was calculated for each antihypertensive drug class. This variable represents the proportion of time a participant was exposed to a specific drug class relative to their total event or censoring time. The relative exposure time was subsequently included as a continuous independent variable in the Cox regression models. The figure presents HRs derived from these models, illustrating the impact of changes in the relative time on the risk of the primary composite outcome: left panel (HR per 1 unit): displays HRs associated with each 1-unit increase in the relative exposure time. A 1-unit increase in this variable signifies a transition from no exposure (relative time = 0) to continuous exposure throughout the participant’s entire follow-up period (relative time = 1). Right panel (HR per 0.1 unit): displays HRs associated with each 0.1-unit increase in the relative exposure time. This represents the change in hazard for every 10% increase in the proportion of relative time which the participant was exposed to the drug class (e.g. an increase in relative exposure time from 0.5 to 0.6). The adjusted model accounted for the randomisation group, age, sex, baseline systolic blood pressure, cumulative systolic blood pressure, body mass index, low-density lipoprotein cholesterol, glucose level, history of cardiovascular disease, and baseline renal dysfunction. *The P value was obtained by comparing the hazard ratios of ARBs and CCBs using a two-independent-samples t-test, based on bootstrapped estimates (n = 100). †Primary composite outcome includes the first occurrence of stroke (ischaemic or haemorrhagic), acute coronary syndrome (myocardial infarction and hospitalisation for unstable angina), acute decompensated heart failure, coronary revascularisation, atrial fibrillation, or death from cardiovascular causes. Abbreviations: ARB, angiotensin II receptor blocker; CCB, calcium channel blocker; CI, confidence interval; HR, hazard ratio

Full size image
Table 2 Association of antihypertensive drug class exposure with primary composite outcome*
Full size table

Sensitivity analyses demonstrated that the survival analyses remained robust after excluding participants with a history of CVD (n = 528) or those with a baseline eGFR < 60 ml/min/1.73 m2 (n = 141) (Additional file 1: Fig. S3 and Table S2). A reduced risk of the primary outcome was consistently observed in participants with a longer exposure to ARBs and CCBs, while the increased risk persisted in participants who used beta-blockers. The HRs associated with per 1 unit and per 0.1 unit increase in the relative time of ARB use remained significantly and consistently lower than those of CCB across models 1–3 and sensitivity analyses (Fig. 1 and Additional file 1: Fig. S3, P < 0.001).

Secondary outcomes

Secondary outcomes included 234 MACE, 129 all-cause mortalities, 42 cardiovascular mortalities, 116 strokes, 135 ACSs, 54 coronary revascularisations, 49 atrial fibrillations, and 13 heart failures. In separate analyses, longer relative exposure times to ARBs or CCBs were associated with a significantly lower risk of all-cause mortality and cardiovascular mortality (Fig. 2 and Additional file 1: Table S3). For all-cause mortality, each unit increase in the relative time on ARB treatment resulted in a 54% risk reduction (HR 0.46, 95% CI 0.31–0.67), while CCB use led to a 51% risk reduction (HR 0.49, 95% CI 0.32–0.75). However, this relationship was not observed for diuretics or beta-blockers. Additionally, each unit increase in the relative time on ARBs was associated with a 56% lower risk of stroke (HR 0.44, 95% CI 0.29–0.66), a 41% lower risk of ACS (HR 0.59, 95% CI 0.41–0.87), and a 39% lower risk of MACE (HR 0.61, 95% CI 0.46–0.81), whereas these benefits were not observed with other types of drugs. Conversely, each unit increase in the relative time on beta-blocker treatment was associated with an increased risk of MACE (HR 2.31, 95% CI 1.82–2.94), stroke (HR 2.04, 95% CI 1.42–2.93), ACS (HR 2.28, 95% CI 1.70–3.05), coronary revascularisation (HR 2.49, 95% CI 1.56–3.96), and atrial fibrillation (HR 2.30, 95% CI 1.63–3.25). The propensity score-matched analysis showed consistent results for MACE, stroke, and ACS (Additional file 1: Table S1). Furthermore, longer exposure to ARBs demonstrated more pronounced benefits in reducing the risks of MACE and stroke compared to CCBs (Fig. 2, all P < 0.05). A similar trend was observed when calculating the HR per 0.1 unit change in the relative time for different antihypertensive drug classes (Additional file 1: Fig. S4 and Table S3).

Fig. 2
figure 2

Association of antihypertensive drug class exposure with secondary outcomes demonstrating the hazard ratio per unit change of relative time. The adjusted model accounted for the randomisation group, age, sex, baseline systolic blood pressure, cumulative systolic blood pressure, body mass index, low-density lipoprotein cholesterol, glucose level, history of cardiovascular disease, and baseline renal dysfunction. *The P value was obtained by comparing the hazard ratios of ARBs and CCBs using a two-independent-samples t-test, based on bootstrapped estimates (n = 100). Abbreviations: ACS, acute coronary syndrome; ARB, angiotensin II receptor blocker; CCB, calcium channel blocker; CI, confidence interval; CV, cardiovascular; MACE, major adverse cardiac events

Full size image

Discussion

In this post hoc analysis of the STEP trial conducted among elderly patients with hypertension, our findings reveal that prolonged treatments with ARB and CCB are significantly associated with a lower risk of composite cardiac outcomes, all-cause mortality, and cardiovascular mortality. In addition, extended ARB exposure further reduces the risk of stroke, ACS, and MACE; these benefits are not observed with other types of antihypertensive drugs. Compared to CCBs, ARBs demonstrate more pronounced benefits in reducing the risk of various cardiovascular outcomes.

Given these findings, ARBs may be preferable to CCBs in providing additional cardiovascular benefits in hypertensive patients with well-controlled blood pressure. Clinically, the favourable antihypertensive effects of ARBs in reducing cardiovascular events have been extensively documented [20,21,22]. Several recent studies have compared the long-term benefits of ARBs and CCBs, the two most commonly used antihypertensive drugs. However, the results have been contradictory. Some studies have reported neutral findings [23, 24], whereas some studies have indicated that renin-angiotensin system (RAS) inhibitors provide more favourable clinical outcomes than CCBs. A pooled meta-analysis revealed that RAS inhibitors resulted in a significant absolute reduction in heart failure-related deaths or hospitalisations compared with CCBs, suggesting that RAS inhibitors might be a better choice for first-line medications [9]. Furthermore, a study investigating the outcomes in 7080 high-risk hypertensive patients using ARBs or CCBs with comparable blood pressure levels found that ARBs were associated with a significantly lower risk of heart failure and new-onset diabetes [25]. Another randomised controlled trial reported that the primary composite cardiac endpoints did not differ significantly between the ARB and CCB groups; however, ARB recipients had a 35% lower risk of congestive heart failure than participants receiving CCBs [26]. Our study further supports these findings and extends the existing evidence by demonstrating that ARBs provide remarkable benefits than CCBs across multiple cardiovascular outcomes.

Except for ARBs and CCBs, our findings also revealed a strong correlation between prolonged beta-blocker exposure and cardiovascular events, including primary outcomes, MACE, stroke, and ACS, in both unmatched and propensity score-matched analyses. Although these associations may be influenced by selection bias due to medical indications, it could also be attributed to the less efficient and less stable blood pressure-lowering effects of beta-blockers as well as their inferior ability to reduce the risk of CVD events. A recent randomised controlled trial, the Randomised Evaluation of Decreased Usage of Beta-Blockers after Acute Myocardial Infarction (REDUCE-AMI), reported that long-term use of beta-blockers in patients with acute myocardial infarction and preserved left ventricular ejection fraction did not reduce all-cause mortality or the risk of recurrent myocardial infarction [27], raising concerns regarding the efficacy of beta-blockers [28]. Similarly, a previous study indicated that combining ARBs with beta-blockers provided little additional blood pressure reduction compared with ARB monotherapy [29]. Solanki et al. also reported that beta-blockers have a lower blood pressure-lowering effect than CCBs [30]. Several studies have further demonstrated that beta-blockers are less effective compared to other antihypertensive drugs in reducing cardiovascular risk [31,32,33]. These findings underscore the need for further high-quality research to assess the cardiovascular outcomes of different beta-blocker subtypes compared with other antihypertensive medications.

A notable strength of the present study is its examination of various antihypertensive drug classes and their impact on future CVD events in a real-world clinical trial that enrolled participants with well-controlled blood pressure. Additionally, in the present study, the HR estimates derived from the ‘per 1 unit’ and ‘per 0.1 unit’ increase in relative time offer complementary perspectives. The ‘HR per 1 unit’ provides an estimate of the overall potential impact of sustained drug exposure relative to non-exposure over the observed period. The magnitude of these HRs reflects a larger effect size because it encompasses the entire spectrum from no use to full use. In contrast, the ‘HR per 0.1 unit’ quantifies the effect of a 10% increase in proportional exposure duration. These HRs are naturally smaller as they represent a smaller change in relative exposure duration. Critically, the direction of the associations (protective for ARBs/CCBs, negative for beta-blockers, and neutral for diuretics) remained consistent across both analytical approaches. This consistency strengthens the confidence in our findings. The current results allow for a more nuanced interpretation: not only is sustained use of ARBs and CCBs beneficial, but longer proportional duration of use appears to contribute incremental benefit.

However, this study has certain limitations. First, it is important to acknowledge that this study is a post hoc analysis of the STEP trial, which was not prespecified designed to compare the effects of different antihypertensive drug classes. Thus, the observed associations could not establish causality. Prespecified designed randomised controlled trials might further verify these associations. To mitigate potential bias, adjustments were made for the randomisation group, baseline SBP, and cumulative SBP values. Additionally, variations in survival among participants were accounted for by utilising relative time, which considered the occurrence of events and helped avoid biases associated with these differences. Second, the higher event rates observed in patients receiving beta-blockers for a longer duration may be due to selection bias. Beta-blockers are often prescribed to patients with known CVDs, suspected ischaemic heart disease, paroxysmal or permanent tachycardia, or reduced or suspected reduced left ventricular ejection fraction, as well as to patients in need of more than two antihypertensive drugs. Efforts were made to reduce this bias and validate the results by excluding patients with previous CVD history or reduced eGFR in the sensitivity analysis, as well as to incorporate propensity score-matched analyses. Third, the STEP trial exclusively included Chinese participants, which may have limited the ethnic diversity and generalisability of the findings. For example, previous evidence suggests that CCBs may produce better outcomes in African and other heterogeneous populations [34,35,36]. Conversely, RAS inhibitors have been found to be more effective in Asian populations [37, 38]. Given that a head-to-head study has reported comparable efficacy between ARBs and ACEIs [39], and considering factors such as adherence, antihypertensive efficacy, and cardiovascular benefits, ARBs may be more suitable for Asian populations [40], particularly for those who experience dry cough associated with ACEIs [41]. Therefore, regional and ethnic differences should be considered when interpreting the effects of antihypertensive medications. Further research is required to validate these findings in the general population. More well-designed studies are also warranted to provide deeper insight into the comparative effectiveness of different antihypertensive drug classes across diverse patient populations.

Conclusions

The findings from this post hoc analysis suggested that longer treatments with ARBs and CCBs might be associated with a more favourable long-term prognosis for multiple cardiovascular events in elderly patients with hypertension than diuretics and beta-blockers. Specifically, prolonged exposure to ARBs appeared to be correlated with a lower incidence of composite cardiac outcomes, MACE, and stroke when compared to CCBs. In contrast, while longer beta-blocker use was observed to be associated with an increased risk of various CVD events, this finding could potentially be influenced by selection bias based on medical indications.