Association of long-term time in target range for systolic blood pressure with cardiovascular risk in the elderly: a Chinese veteran cohort study

Abstract

Aims

Short-term blood pressure (BP) time in target range (TTR) independently predicts cardiovascular (CV) outcomes in adults. However, there are limited data regarding long-term TTR for BP among elderly participants. We aimed to determine whether future CV risk varies for those who can maintain a long-term systolic BP (SBP) target range by assessing TTR in elderly individuals with hypertension.

Methods and results

The Chinese veteran cohort study included 943 elderly participants with hypertension aged over 75 years. The primary outcome was the first occurrence of CV events during annual visits. Time in target range was estimated over 15 years of follow-up using linear interpolation. The target range was defined as 120–140 mmHg according to guidelines. The association between SBP TTR and CV outcomes was estimated using multivariable Cox proportional hazards models. During the 15 year follow-up, the probability of CV events gradually decreased with increasing TTR for SBP. After multivariable adjustment for traditional CV risk factors and mean BP, comparing the highest vs. lowest quartiles of TTR for SBP, the hazard ratios (HRs) [95% confidence intervals (CIs)] were 0.424 (0.289–0.624) for the primary outcome. For each 1 SD increase in TTR, the risk of the primary outcome decreased by 25.4% (HR: 0.746; 95% CI: 0.666–0.834). Consistent findings were observed in sensitivity analyses.

Conclusion

Greater long-term TTR for SBP was associated with a decreased risk of CV events in elderly individuals independent of mean BP, suggesting that SBP TTR might serve as a modifiable risk factor for future CV health in elderly patients with hypertension.

Lay Summary

This ongoing Chinese veteran cohort study adds to the understanding of the relationship between higher long-term systolic blood pressure (SBP) time in target range (TTR) and cardiovascular benefits among elderly individuals with hypertension.

Graphical Abstract

Higher long-term TTR for SBP is associated with a decreased risk of CV events in the elderly. TTR, time in target range; SBP, systolic blood pressure; CV, cardiovascular; MI, myocardial infarction; HR, hazard ratio; CI, confidence interval.

Open in new tabDownload slide

Introduction

Hypertension, one of the established risk factors for cardiovascular disease (CVD), is highly prevalent (>70%) among adults over 75 years. Effective blood pressure (BP) management can reduce the incidence of cardiovascular (CV) events.^1,2 Blood pressure–relevant CV risk may depend not only on mean BP but also on the degree of BP variability.³ Despite antihypertensive therapy, elderly hypertensive patients frequently experience pseudo hypertension, postural hypotension, and whitecoat or masked hypertension,⁴ which typically manifest as high BP fluctuation and increased systolic BP (SBP) and pulse pressure.⁵ Given that BP is dynamic and that there is inter-measurement variability in BP readings, this reinforces the need for widely available and reliable measures of BP control that could assess BP variability as well as BP levels in elderly individuals.

Time in target range (TTR) for BP could incorporate both the degree of BP variability and the mean BP level, which reflect variation over time both within and outside of the target range.⁶ A few studies have previously demonstrated that high TTR for BP reduces the incidence of CV events in people who achieve and maintain their BP target range,^7–9 but these benefits may be obscured in the elderly population by individuals who have CVD and a short life expectancy. The related evidence derived from elderly participants over the age of 75 years is limited. Moreover, prior studies investigated TTR using BP data typically from a short-term viewpoint. For example, in the Systolic Blood Pressure Intervention Trial (SPRINT) study, TTR for BP was only estimated over the first 3 months of follow-up.⁷ However, there are no data linking long-term BP TTR to CV outcomes. Therefore, we hypothesized that BP control assessed by long-term TTR might help predict CV outcomes in elderly individuals.

To test this hypothesis, we conducted a cohort study of Chinese veterans aged over 75 years. We aimed to estimate the association between long-term TTR for SBP of 120–140 mmHg and CV outcomes among elderly hypertensive patients with a 15 year follow-up and explore the prognostic value of TTR for these patients in real-life clinical practice.

Methods

Study population

As an ongoing retrospective military veteran cohort study initiated in Guangzhou, China, all participants in our analysis were veterans who were regularly hospitalized or medically examined in the General Hospital of Southern Theater Command. From January 2006 to June 2006, a total of 1154 elderly participants with hypertension aged ≥75 years were enrolled in this study. We excluded participants who had less than three valid BP values within the first 3 years (n = 139) and those with missing covariate data (n = 65) or a history of severe renal dysfunction [estimated glomerular filtration rate (eGFR) <15 mL/(min•1.73 m²)] (n = 7). The final sample size for the analysis of TTR for SBP and CV outcomes was 943 participants (see Supplementary material online, Figure S1). This study was conducted according to the Declaration of Helsinki and approved by the Institutional Review Board of the General Hospital of Southern Theater Command. The detailed study design has been described in the Chinese Clinical Trial Registry (URL: http://www.chictr.org.cn; Unique identifier: ChiCTR2200059475).

Blood pressure measurement and definition of time in target range

Blood pressure was measured by trained physicians three times at 30 s intervals using a validated BP monitor and followed standard recommended procedures. Manual mercury sphygmomanometers were used in the period between 2006 and 2016, and electronic BP metre (HEM-7136; Omron Healthcare, Matsusaka, Japan) was used in the 2016 and thereafter. The BP measurement devices have been changed synchronously during long-term follow-up. The numerical correction of the BP measuring device was usually carried out every 6–12 months. To ensure that measurements are comparable, the two devices would be calibrated to each other when a device change occurs. Participants were required to rest for 3–5 min in a seated position. An appropriate cuff size that covered two-thirds of the left upper arm circumference was used for accurate reading. Blood pressure values were obtained from a yearly visit during the study period. Mean SBP was calculated using all available SBP readings from baseline to the end of follow-up or to the year before outcomes occurred. Hypertension was defined as office SBP values ≥140 mmHg and/or diastolic BP (DBP) values ≥90 mmHg and any self-reported or treated hypertension.¹⁰ Blood pressure values from yearly visit before the onset of CVD events were used. At least three visits of SBP measurements from 2006 to 2021 were used to calculate TTR. Time in target range for SBP was defined as the percentage of time during which the SBP was within the target range and was estimated with linear interpolation¹¹ using at least three valid measurements of SBP during years 1–16. The SBP target range was determined as 120–40 mmHg to match the current guideline-recommended target^12–14 and determined as <120 mmHg¹⁵ (or <130 mmHg,¹⁶ < 140 mmHg¹⁰, < 150 mmHg) for additional analysis. Greater TTR (%) indicated that the participants achieved and maintained their target SBP range for a longer duration within the 15 year period.

Covariates

Data on baseline demographic (age, sex), smoking status, medical history [diabetes, stroke, myocardial infarction (MI), or angina pectoris], and concomitant treatment (antihypertensive medication) characteristics were obtained through standardized questionnaires and hospital electronic medical record systems. The use of antihypertensive medication was based on guidelines and the participant’s needs. Height was measured using a portable stadiometer while participants were barefoot. Weight was measured using an electronic scale with participants in light indoor clothing. Body mass index (BMI) was calculated as weight in kilograms divided by height in metres squared. Blood samples were drawn for testing after an overnight fast and were analysed using an autoanalyzer (Roche Diagnostics). Biochemical indicators, including triglyceride (TG), low-density lipoprotein cholesterol (LDL-C), and serum creatinine, were measured. We used the simplified Modification of Diet in Renal Disease formula to calculate the eGFR.¹⁷

Assessment of outcomes

The study participants were followed up from the baseline investigation to the end of follow-up (1 October 2021) or to the date of death. Only the first CV event was used in the analysis for participants who reached multiple endpoints. The primary composite CV outcome was the first occurrence of stroke, MI, angina pectoris, or cardiac death. The following three secondary composite CV outcomes were also considered: stroke, MI, or cardiac death (secondary outcome 1); stroke, MI, or angina pectoris (secondary outcome 2); and stroke, MI, angina pectoris, or all-cause mortality (secondary outcome 3).

Information on CV event diagnoses and causes of death was obtained from individuals’ discharge records from the General Hospital of Southern Theater Command and annual questionnaires during the follow-up period. Two cardiologists reviewed the records to identify participants with suspected CVD. The diagnosis of events was ascertained by clinical symptoms, blood tests, and examinations including electrocardiograms, magnetic resonance imaging, and computed tomography. Definitions and ascertainment criteria for stroke,¹⁸ MI,¹⁹ and angina pectoris²⁰ were based on the current guidelines. Data were also retrieved from the Medical Record Management Department to verify the events and accurate dates.

Statistical analysis

Continuous variables are expressed as the medians (interquartile range) or means ± standard deviations (SD). Categorical variables are summarized as frequencies and percentages (%). Continuous baseline data were compared across the four TTR groups for SBP (0% to <25%, 25% to <50%, 50% to <75%, 75–100%) by using a nonparametric test or variance test, and categorical baseline data were compared using the χ² test. The number of events per outcome and incidence rate per 1000 person-years were also calculated. Time in target range was estimated over 15 years of follow-up using linear interpolation and was transformed into categorical variables and continuous variables for analysis. We graphically displayed the mean annual SBP changes in the four TTR groups during years 1–16. Cumulative incidences were estimated for outcomes for the four TTR groups using the Kaplan–Meier method. The log-rank test was used for survival distributions. The associations of TTR with the first occurrence of an efficacy outcome were estimated using hazard ratios (HRs) and 95% confidence intervals (CIs) derived from unadjusted and adjusted Cox proportional hazard models. Model 1 was fully adjusted for baseline demographic variables (age, sex), traditional risk factors (smoking status, BMI, SBP, TG, LDL-C, eGFR), use of antihypertensive medication, and baseline diabetes and CVD status. Model 2 was adjusted by the variables from Model 1 plus the last SBP measurement. Model 3 was adjusted by the variables from Model 1 plus mean SBP. To consider the effects of age, BMI, diabetes status, and LDL-C on CV events, multiplicative interaction terms were used to assess the interactions of age, BMI, diabetes, and LDL-C and the association of TTR with CV outcomes. To assess the potential influence of BP measurement number on outcome, we performed a series of sensitivity analyses according to the number of BP measurement (3–4, 5–8, or 6–16 BP readings) during follow-up. Statistical analyses were performed using the IBM SPSS Statistics 22 program (IBM SPSS) and Stata 16 software (Stata Corp). A result associated with a two-sided P value of <0.05 was considered statistically significant.

Results

Study population characteristics

A total of 943 participants with a median age of 78 (76–81) years at baseline were enrolled; 927 (98.3%) were male, and the median follow-up period was 10 (6–14) years (Table 1). A total of 194 (20.6%), 75 (8.0%), and 153 (16.2%) participants had a history of stroke, MI, and angina pectoris, respectively. Participants in the highest TTR quartile had lower baseline SBP than those in the lowest quartile [130 (123–132) vs. 140 (129–150) mmHg, P < 0.001]. Laboratory examinations (eGFR, TG, LDL-C), the use of antihypertensive medication, and the percentage of participants with CVD were not significantly different among the four TTR groups at baseline. The baseline characteristics of the included and excluded populations are shown in Supplementary material online, Table S1. As shown in Supplementary material online, Table S2, the numbers of BP readings in the cohort ranged from 3 to 16. The participants received BP measurements with 3–8 accounted for 15.2%, 20.6%, 18.1%, 15.8%, 9.4%, and 5.8% of the included participants, respectively. The use of antihypertensive drugs was classified as angiotensin II receptor blocker (ARB), angiotensin-converting enzyme inhibitor (ACEI), calcium channel blockers (CCB), beta-blockers, or diuretics (see Supplementary material online, Table S3). The numbers of antihypertensive medication ranged from 0 to 4 (see Supplementary material online, Table S4). There was no significant difference in the categories and the numbers of antihypertensive drugs among the four TTR groups. In addition, the mean annual SBP changes that occurred in the four TTR groups during the 15 years of follow-up are shown in Figure 1. As expected, participants with greater TTR achieved and maintained their SBP in the target range for a longer period of time.

Associations of time in target range for systolic blood pressure and cardiovascular outcomes

The Kaplan–Meier survival function curves showed a lower cumulative incidence of primary and secondary outcomes in participants with the highest quartiles of TTR levels than in those with the lowest quartiles (all log-rank test P < 0.05, Figure 2). The effect estimates of the unadjusted and adjusted models for the association of TTR with CV outcomes are shown in Table 2. In the unadjusted models, the highest TTR quartile group was associated with a decreased risk of primary outcome. After adjusting for baseline demographics, traditional CV risk factors, and mean SBP, compared to the lowest TTR group, the HRs (95% CIs) were 0.424 (0.289–0.624) for the primary outcome in the highest TTR group (Table 2). Participants with a TTR of 75–100% had a 60.2% lower risk of secondary outcome 1 (HR: 0.398; 95% CI: 0.239–0.664), a 55.1% lower risk of secondary outcome 2 (HR: 0.449; 95% CI: 0.304–0.662), and a 40.4% lower risk of secondary outcome 3 (HR: 0.596; 95% CI: 0.445–0.798) than those with a TTR of 0% to <25%.

When TTR was converted to a continuous variable, the association of the CV outcomes and TTR is shown in Table 3. For each 1 SD increase in TTR, the risk of the primary outcome decreased by 25.4% (HR: 0.746; 95% CI: 0.666–0.834) in the fully adjusted models. For secondary outcome 1, secondary outcome 2, and secondary outcome 3, the risk was reduced by 26.5% (HR: 0.735; 95% CI: 0.636–0.848), 24.5% (HR: 0.755; 95% CI: 0.674–0.846), and 15.5% (HR: 0.845; 95% CI: 0.775–0.922), respectively, with each 1 SD increase in TTR after full adjustment.

Subgroup analyses

Subgroup analyses were performed to stratify the association between TTR and CV events by baseline age, BMI, diabetes status, and LDL-C. With increased TTR, the risk of CV events showed a similar decrease in participants younger or older than 80 years in the age subgroup [e.g. for each 1 SD increase in TTR, the HRs for the primary outcome were 0.783 (95% CI, 0.681–0.900) for participants younger than 80 years vs. 0.694 (95% CI, 0.570–0.846) for participants aged at least 80 years (P for interaction ≥ 0.05) (see Supplementary material online, Figure S2)]. The effects of BMI and baseline diabetes status on the association of TTR with cardiovascular outcomes were also similar in subgroups (P for interaction ≥ 0.05) (see Supplementary material online, Figures S3 and S4). When stratified by LDL-C, the risk of CV events also gradually decreased with increasing levels of TTR. The HRs for the primary outcome in the adjusted analyses were 0.738 (95% CI, 0.621–0.877) for participants with LDL-C < 100 mg/dL and 0.775 (95% CI, 0.667–0.901) for participants with LDL-C ≥ 100 mg/dL (P for interaction = 0.015) (see Supplementary material online, Figure S5). The risk of CV events gradually decreased with increasing levels of TTR.

Sensitivity analyses

Consistent significant associations were observed in the multivariable Cox regression model that adjusted for the number of antihypertensive drugs (see Supplementary material online, Table S5). To assess the influence of wrong measurement of BP due to atrial fibrillation (AF), we restricted individuals to participants without AF for sensitivity analysis. The results of the sensitivity analysis that restricted individuals to participants without AF were also consistent with the main findings (see Supplementary material online, Table S6). In a sensitivity analysis that restricted the analysis to individuals with 3–4 SBP readings, participants with higher TTRs were still associated with lower risks of primary and secondary outcomes (see Supplementary material online, Table S7). In the unadjusted and adjusted models, the associations of greater TTR with a decreased risk of primary outcome remained significant when restricting the analysis to individuals with 5–8 SBP readings (see Supplementary material online, Table S8). Restricting the analysis to participants with 6–16 SBP readings was also significantly associated with CV events (see Supplementary material online, Table S9). This showed that it would be unlikely that the number of SBP measurements could account for the identified association between TTR and CV outcomes. In the redefined SBP target range of <140 mmHg, the results were similar to the primary findings from the target range of 120–140 mmHg (see Supplementary material online, Table S10).

Discussion

In this analysis of a Chinese veteran cohort study of elderly individuals with hypertension at the 15 year follow-up, more time with SBP levels within the proposed optimal target range of 120–140 mmHg was associated with a lower risk of CV events, and this association remained significant despite adjustment for traditional CV risk factors and mean SBP. These results indicate that long-term TTR for SBP provides considerable value for elderly individuals, suggesting that the management of SBP should focus not only on achieving the target range but also on maintaining most of the time achieved.

Given that the newly introduced TTR measurement considers BP variability along with the target range over time, it might be suitable and convenient for the long-term management of hypertension. A large study demonstrated that sustainable control of SBP in the range of 120–140 mmHg minimizes the risk of all-cause mortality in hypertensive patients.¹³ A recent post hoc analysis of the SPRINT study showed that TTR for SBP independently predicted CV outcomes during a median follow-up of 3.3 years. Few studies have previously used TTR for BP to assess the CV risk profile in populations with hypertension, but data from elderly individuals, including long-term TTR data and an appropriate control group, are lacking.^7–9 In contrast to previous studies, our study provides new evidence from a long-term cohort study of veterans with well-balanced baseline characteristics. Here, our study focused on hypertensive patients over 75 years of age and showed that a greater long-term TTR was associated with a reduced risk of CV events at the 15 year follow-up. The results remained consistent even when they were fully adjusted by traditional CV risk factors and mean SBP. A significant interaction existed between LDL-C and the association of TTR with CV outcomes and LDL-C alone might be one of the determinant factors for CVD.²¹ Despite all this, TTR could predict the CV outcomes, independently of LDL-C levels. These findings highlight the need to maintain the SBP therapeutic range achieved through effective antihypertension therapy to reduce CV events in elderly individuals,²² emphasizing the CV benefits of long-term TTR for SBP management.

Current guideline-based recommendations for the optimal SBP target for elderly patients remain inconsistent. For example, the SBP target is <130 mmHg in the 2017 American College of Cardiology/American Heart Association (ACC/AHA) guidelines¹⁶ and 130–139 mmHg in the 2018 European Society of Cardiology/European Society of Hypertension (ESC/ESH) guidelines.¹⁰ In recent years, the accumulation of evidence from the SPRINT study and the Strategy of Blood Pressure Intervention in the Elderly Hypertensive Patients (STEP) trial has gradually established the tenet ‘the lower, the better’ in the hypertension field; however, the higher incidence of adverse events should be noted,²³ and the exclusion criteria of randomized controlled trials and the approach to BP monitoring have caused concerns in relation to the generalizability of the results in real-world clinical practice.²⁴ Furthermore, a large-scale observational study showed that BP <140/90 mmHg may be associated with an increased risk of mortality in octogenarians.²⁵ The SBP target range used in this study was 120–140 mmHg, which was in line with most BP-lowering trials focusing on the elderly population. Interestingly, our study showed significant benefits for CV outcomes at SBP targets of 120–140 mmHg but not significant at targets of <130 mmHg, while negative effects may have been observed at targets of <120 mmHg (see Supplementary material online, Tables S11–S13). This finding indicated that a target of reducing SBP to <120 mmHg needs to be applied with caution in elderly patients aged over 75 years. These results do not support the speculation that ‘lower BP is better’, especially among older populations with coexisting CVD.²⁶ The reduction in therapy compliance, individual differences, and adverse reactions relevant to low SBP should be carefully assessed.²⁷

A higher TTR reflects lower variability and steadier BP during a specific period. There are several possible mechanisms involved in the process of high TTR and CV outcomes. First, the consistency in shear stress tends to show less tears and fragmentation in the internal elastic lamina of the arteries,²⁸ which may reduce the development of atherosclerotic plaque.²⁹ Second, the activation of intracellular signal transduction pathways induced by mechanical factors is reduced, stabilizing intracellular function.^30,31 With the aging process, the elasticity of the vessel is decreased due to the degradation of elastin and subsequent accumulation of collagen.^32,33 As impaired arterial function during aging is related to arterial remodeling, a higher TTR may play an important role in preventing the initial adaptive remodeling from becoming maladaptive remodeling.

The strengths of our study included having a sample of participants aged 75 years and older, a relatively long-term follow-up, standardized data collection protocols, and rigorous quality control. In this regard, several limitations should be considered. First, we mainly focused on SBP. Although prior studies have analysed both SBP and DBP, their findings demonstrated that SBP rather than DBP is a stronger determinant of CV outcomes in hypertensive patients.³⁴ Second, our study was a racially homogeneous, single-centre, male-dominated study; however, this could be because of the specificity of our cohort of veterans. Third, we used at least three valid BP values to calculate TTR, which have the potential influence on the outcome. Although the sensitivity analyses reveal that the number of BP measurement did not influence the outcome, more evidence by using frequent BP measurement to calculate TTR is needed in the future. Fourth, although we performed analyses to adjust for baseline differences, other potential confounders may not have been considered. In addition, some valuable parameters, such as 24 h BP measurements and assessment of orthostatic hypotension, were not contained in this study. Future studies using ambulatory BP monitoring to assess TTR could be conducted to verify this finding. Fifth, there might be erroneous measurements due to increased stiffness. The current study enrolled only the Chinese veteran population mainly in south China; the homogeneous nature of our cohort could help to reduce differences in vascular stiffness because of racial and healthcare disparities. The use of a standardized protocol by trained technicians and calibrated equipment to measure BP could ensure the accuracy of results. Finally, we cannot explain the cause-and-effect relationships between variables because of the study’s observational nature.

Conclusions

In conclusion, among elderly individuals with hypertension, higher long-term TTR for SBP is associated with a significantly decreased risk of CV events independent of mean SBP. This study identifies TTR as a convenient and reliable metric of BP control for elderly hypertensive patients in real-life clinical practice, potentially emphasizing the importance of maintaining SBP within the therapeutic range of 120–140 mmHg most of the time to obtain better CV outcomes in the elderly population.

Author contributions

Z.L., Zhiw.X., W.F., Y.C., and J.B. conceived and designed the study. Zhiw.X., W.F., and Y.C. interpreted the data and wrote the manuscript. W.C., W.X., C.H., J.X., M.J., X.W., and S.H. interpreted the data and revised the manuscript draft for important intellectual content. Zhiq.X., W.L., and Y.L. provided supervision. All authors read and approved the final manuscript. W.F., Y.C., and J.B. assumed final responsibility for the decision to submit the manuscript for publication. All gave final approval and agree to be accountable for all aspects of work ensuring integrity and accuracy.

Supplementary material

Supplementary material is available at European Journal of Preventive Cardiology.

Acknowledgements

We thank the staff and participants of Guangdong Branch Center, National Clinical Research Center for Geriatric Diseases, for their significant contributions.

Funding

This work was supported by the National Natural Science Foundation of China (grant 82070315 to J.B., 81970239 to Y.C., and 82200900 to W.F.), GuangDong Basic and Applied Basic Research Foundation (grant 2023A1515010381 to Y.C.), China Postdoctoral Science Foundation (grant 2022M721495 to W.F.), Foundation of General Hospital of Southern Theater Command (grant 2021NZC028 to Z.L.), and National Key Research and Development Program of China (grant 2021YFC2500600/2021YFC2500601 to Z.L.). The funders had no role in the study design, data collection and analysis, interpretation, or writing of the report.

Ethics approval and consent to participate

This study was conducted according to the Declaration of Helsinki and was approved by the Institutional Review Board of the General Hospital of Southern Theater Command.

Data availability

The data that support the findings of this study are available from the corresponding author upon request.

References

Author notes