Skip main navigation

Influence of Baseline Diastolic Blood Pressure on Effects of Intensive Compared With Standard Blood Pressure Control

and for the SPRINT Research Group
Originally publishedhttps://doi.org/10.1161/CIRCULATIONAHA.117.030848Circulation. 2018;137:134–143

Abstract

Background:

In individuals with a low diastolic blood pressure (DBP), the potential benefits or risks of intensive systolic blood pressure (SBP) lowering are unclear.

Methods:

SPRINT (Systolic Blood Pressure Intervention Trial) was a randomized controlled trial that compared the effects of intensive (target <120 mm Hg) and standard (target <140 mm Hg) SBP control in 9361 older adults with high blood pressure at increased risk of cardiovascular disease. The primary outcome was a composite of cardiovascular disease events. All-cause death and incident chronic kidney disease were secondary outcomes. This post hoc analysis examined whether the effects of the SBP intervention differed by baseline DBP.

Results:

Mean baseline SBP and DBP were 139.7±15.6 and 78.1±11.9 mm Hg, respectively. Regardless of the randomized treatment, baseline DBP had a U-shaped association with the hazard of the primary cardiovascular disease outcome. However, the effects of the intensive SBP intervention on the primary outcome were not influenced by baseline DBP level (P for interaction=0.83). The primary outcome hazard ratio for intensive versus standard treatment was 0.78 (95% confidence interval, 0.57–1.07) in the lowest DBP quintile (mean baseline DBP, 61±5 mm Hg) and 0.74 (95% confidence interval, 0.61–0.90) in the upper 4 DBP quintiles (mean baseline DBP, 82±9 mm Hg), with an interaction P value of 0.78. Results were similar for all-cause death and kidney events.

Conclusions:

Low baseline DBP was associated with increased risk of cardiovascular disease events, but there was no evidence that the benefit of the intensive SBP lowering differed by baseline DBP.

Clinical Trial Registration:

URL: https://www.clinicaltrials.gov. Unique identifier: NCT01206062.

Introduction

Editorial, see p 144

Clinical Perspective

What Is New?

  • There were U-shaped relationships of baseline diastolic blood pressure (DBP) with the primary cardiovascular disease outcome and all-cause death in SPRINT (Systolic Blood Pressure Intervention Trial).

  • However, the beneficial effects of intensive systolic blood pressure lowering (intensive systolic blood pressure goal <120 mm Hg versus standard systolic blood pressure goal <140 mm Hg) on the primary cardiovascular disease outcome and all-cause death were not modified by baseline level of DBP.

  • Increased risk of kidney events and serious adverse effects of the intervention was consistent across baseline DBP quintiles.

  • Therefore, there was no evidence that the benefit of the intensive systolic blood pressure lowering differed by baseline DBP level.

What Are the Clinical Implications?

  • Some cohort observational studies and nonrandomized secondary analyses of achieved blood pressures suggested a J-curve relationship of DBP with cardiovascular events.

  • Results of current analyses of SPRINT data suggest that underlying processes (such as increased arterial stiffness) that lead to a decline in DBP rather than the level of DBP per se might be the reason for the observed associations of worse outcomes with lower DBP.

  • Low levels of DBP within the ranges examined here in SPRINT should not be an impediment to intensive treatment of hypertension, at least in those without diabetes mellitus or stroke.

Elevated blood pressure (BP) is an important risk factor for cardiovascular disease (CVD),1,2 end-stage kidney disease,3,4 and all-cause mortality.2,5 Beginning in the 1960s, randomized controlled trials demonstrated the value of treating high diastolic BP (DBP) and subsequently high systolic BP (SBP).6,7 Recently, SPRINT (Systolic Blood Pressure Intervention Trial) demonstrated that intensive SBP lowering (SBP target <120 versus <140 mm Hg) improved CVD outcomes and all-cause mortality in adults at high risk for CVD events,8 even in those ≥75 years of age.9

Despite the documented value of traditional treatment in adults with a high DBP,6 intensive therapy to low levels of DBP is controversial. Nearly 30 years ago, a J-shaped relationship was observed between on-treatment DBP and death resulting from myocardial infarction, with the risk being lowest in those with an achieved DBP between 85 and 90 mm Hg and higher at achieved DBP levels on either side of this range.10,11

We examined the hypothesis that low baseline DBP adversely modifies the effect of intensive SBP lowering on CVD, kidney disease, and all-cause mortality in SPRINT.8 In addition, we examined whether baseline pulse pressure (PP) or mean arterial pressure (MAP) modified the effects of the SPRINT intervention.

Methods

Limited SPRINT data are available through the National Heart, Lung, and Blood Institute at https://biolincc.nhlbi.nih.gov/studies/sprint_pop for reproducing/replicating the results of this analysis. The Methods section and online-only Data Supplement provide details of the analytical procedures.

SPRINT was a randomized controlled open-label trial that compared the effects of intensive (SBP target <120 mm Hg) andstandard (SBP target <140 mm Hg) BP control in 9361 participants from the United States, including Puerto Rico.8 Details of the SPRINT protocol have been published.12,13 Institutional review boards at each of the participating study sites approved the protocol, and all participants provided informed consent.

Study Population

Participants had to be ≥50 years of age with an SBP of 130 to 180 mm Hg and an increased risk of CVD (defined as having at least 1 of the following: clinical or subclinical CVD other than stroke, 10-year risk of CVD ≥15% based on the Framingham global risk indicator,14 age ≥75 years, or estimated glomerular filtration rate [eGFR] of 20–<60 mL·min−1·1.73 m−2). Major exclusion criteria included diabetes mellitus, prior stroke, advanced chronic kidney disease (CKD; eGFR <20 mL·min−1·1.73 m−2), proteinuria >1 g/d, polycystic kidney disease, congestive heart failure, dementia, or residence in a nursing home.

Intervention, Follow-Up, and Measurements

Participants were randomly assigned to intensive or standard SBP control, stratified by clinical site. Details of the SPRINT intervention algorithm and medication formulary are provided elsewhere.12,13 Participants were seen monthly for 3 months and quarterly thereafter for standardized study visits by trained study staff following protocol requirements. An automated measurement system (model 907XL, Omron Healthcare) was used to record BP at the clinic visit after the participant had been seated for 5 minutes of quiet rest. The mean of 3 BP readings, each 1 minute apart, was used to estimate BP.

Medications were adjusted to target an SBP <120 mm Hg in the intensive treatment group and an SBP of 135 to 139 mm Hg in the standard treatment group. Blood specimens were obtained at each visit for the first 3 months and quarterly thereafter for measurement of serum creatinine. The 4-variable Modification of Diet in Renal Disease equation was used to estimate glomerular filtration rate.15

Event ascertainment and safety assessments were performed per protocol.12,13

Sprint Outcomes

The primary outcome was a composite of nonfatal myocardial infarction, acute coronary syndrome not resulting in myocardial infarction, stroke, acute decompensated heart failure, or death resulting from CVD. Death resulting from any cause was a predefined secondary outcome in SPRINT. All outcome events were adjudicated by a committee blinded to treatment assignment. In the present analysis, we also explored a composite CVD outcome that excluded stroke.

The main secondary kidney outcome was a composite of ≥50% decrease in eGFR or development of end-stage renal disease in participants with baseline CKD (eGFR <60 mL·min−1·1.73 m−2). A secondary kidney outcome was incident CKD defined as >30% decrease in eGFR (with a value <60 mL·min−1·1.73 m−2, confirmed at the next available SPRINT blood draw) in participants without CKD at baseline. In addition, we monitored for serious adverse events as reported earlier.8

A decision to discontinue the SPRINT BP intervention was made on August 20, 2015, after interim analyses showed that the primary outcome had exceeded preset monitoring boundaries on 2 consecutive occasions.8 Our analysis is based on information provided in the SPRINT public access BioLINCC database.16 It includes events that occurred on or before the trial was stopped on August 20, 2015, and were recognized with the use of a data freeze date of October 14, 2015.

Statistical Methods

We performed all analyses in STATA version MP 14.0 or SAS version 9.4, and we used a 2-sided α=0.05 for hypothesis testing without adjustment for multiple comparisons. We compared baseline characteristics between DBP quintiles using 1-way ANOVA for numeric variables (after log transformation for the albumin-to-creatinine ratio) and used χ2 tests for categorical variables.

We computed the mean follow-up DBP for each patient by averaging their BP measurements from month 3 to the last reading. We used boxplots to display the patients’ mean follow-up DBP values by quintile of baseline DBP within the intensive and standard groups, and we applied 2-sample tests to compare mean follow-up DBP between the lowest and highest quintile of baseline DBP.

We analyzed the association of baseline DBP with the primary and secondary outcomes by fitting a Cox regression model with the randomized SBP intervention and cubic spline terms in baseline DBP as predictor variables, with covariable adjustment for age, sex, and race. We then performed 3 types of analyses based on Cox proportional hazards regression to investigate whether the effects of intensive SBP intervention on the primary and secondary outcomes differed depending on baseline level of DBP. The primary analysis of the interaction between the intensive SBP intervention and baseline DBP, which was specified before initiation of these post hoc analyses, compared the hazard ratio for the effect of intensive SBP intervention on the primary CVD composite outcome between the lowest baseline DBP quintile and the upper 4 quintiles. Second, we investigated the possibility of a more general interaction by fitting a Cox regression for the primary outcome with main effects for the SBP intervention and cubic spline terms for baseline DBP, plus multiplicative interactions between the SBP intervention and the cubic spline terms. In the absence of evidence of a nonlinear interaction (indicated by an interaction P>0.10), we refit the Cox regression using a linear interaction between the SBP intervention and baseline DPB. Third, we provided hazard ratios with 95% confidence intervals to compare the intensive and usual SBP goals within each baseline DPB quintile. These hazard ratios are presented to provide a comprehensive presentation of the results; however, it is important to note that much or all of the reported variation in hazard ratios between the quintile subgroups is caused by chance and that, in the absence of a statistically significant interaction, the best estimate of the effect of the intervention is given by the study-wide effect estimate, including all patients.

We performed similar analyses of the interaction between the SBP intervention and baseline DBP for the secondary outcomes, except that we categorized baseline DBP by a median split rather than quintiles for the kidney composite outcome because of the small number of events for this outcome.

We repeated each of these 3 analyses to evaluate interactions of the BP intervention with baseline MAP and baseline PP. We also provided hazard ratios from Cox regressions comparing the intensive and usual SBP interventions by baseline DBP quintile for the safety outcomes: all serious adverse events, hypotension, syncope, electrolyte abnormality, and acute kidney injury or acute kidney failure.

In participants with and without baseline clinical/subclinical CVD, we performed additional sensitivity analyses in separate Cox models to evaluate the linear interaction of the SPRINT intervention with baseline DBP.

Additional details of the Cox regression models are provided in the online-only Data Supplement.

Results

The mean age of the study population (n=9361) was 67.9±9.4 years, with 35.6% women and 31.5% black. Mean±SD baseline SBP and DBP were 139.7±15.6 and 78.1±11.9 mm Hg, respectively. Baseline demographic, clinical, and laboratory characteristics of the study population by DBP quintile are summarized in Table 1. In general, participants with lower DBP tended to be older, have a higher baseline prevalence of CVD and CKD, be on more antihypertensive medications, have a lower baseline SBP and MAP and a higher PP, and have a lower eGFR.

Table 1. Baseline Characteristics by Baseline Quintiles of DBP (n=9361)

Quintile 1<68 mm Hg (n=1749)Quintile 268–74 mm Hg (n=1874)Quintile 375–80 mm Hg (n=1816)Quintile 481–87 mm Hg (n=1934)Quintile 5≥88 mm Hg (n=1988)
DBP, mm Hg61±571±278±284±295±6
Age, y74.7±8.270.3±8.868.0±8.565.2±8.362.3±8.3
Female sex, %39.536.035.832.334.9
Black race, %23.225.129.733.644.4
History of CVD, %29.124.118.015.314.9
CKD, %42.329.827.623.220.1
Framingham 10-y CVD risk score ≥15%, %60.359.560.360.867.2
Never smoked, %43.243.645.544.643.3
Antihypertensive agents, n/patient2.1±1.01.9±1.01.8±1.01.7±1.01.6±1.1
Systolic blood pressure, mm Hg*131±15134±13138±13142±13152±15
PP, mm Hg70±1563±1461±1358±1357±13
MAP, mm Hg85±692±598±5103±5114±8
Body mass index, kg/m228.3±5.329.4±5.730.0±5.730.5±5.830.8±6.0
eGFR, mL·min−1·1.73 m−265±2070±2072±2075±2076±21
Urine ACR, mg/g10.7 (6.2, 24.8)9.4 (5.6, 20.3)8.5 (5.2, 18.7)8.9 (5.4, 20.5)10.2 (6.1, 24.6)

Results are presented as percents for binary variables, as mean±SD for continuous variables other than ACR, and as median (interquartile range) for ACR. For comparison of differences between the quintiles, P<0.001 for all except never smoked (P=0.57).

ACR indicates albumin-to-creatinine ratio;

CKD, chronic kidney disease; CVD, cardiovascular disease; DBP, diastolic blood pressure; eGFR, estimated glomerular filtration rate; MAP, mean arterial pressure; and PP, pulse pressure.

*Systolic blood pressure at screening visit was used to determine trial eligibility. Baseline visit values are presented in this table.

Boxplots displaying the medians and 25th and 75th percentiles of mean follow-up SBP, DBP, PP, and MAP levels by baseline DBP quintile for participants in the intensive and standard arms are presented in Figure 1. Because the intervention targeted SBP, regardless of baseline DBP, the distribution of achieved mean follow-up SBP in the intensive arm was similar across baseline quintiles of DBP (Figure 1A). Similar findings for achieved SBP across baseline quintiles of DBP were noted in the standard arm (Figure 1A). However, the achieved mean follow-up DBP was significantly lower among participants in the lowest compared with the highest quintile of baseline DBP within both the intensive (59.5±6.9 versus 74.9±7.0 mm Hg; P<0.001) and standard (65.0±7.6 versus 83.3±6.5 mm Hg, P<0.001) groups (Figure 1B). Within each baseline quintile of DBP, achieved DBP was lower in the intensive compared with the standard group (Figure 1B). Achieved MAP mirrored the pattern noted for achieved DBP (Figure 1C). Achieved PP was the highest in the lowest baseline DBP quintile in both the intensive and standard groups (Figure 1D).

Figure 1.

Figure 1. Mean follow-up blood pressures by randomized SBP intervention and quintile of baseline DBP (n=9119). The boxplots display the median, 25th and 75th percentiles of the patients’ mean follow-up values for systolic blood pressure (SBP; A), diastolic blood pressure (DBP; B), mean arterial pressure (MAP; C) and pulse pressure (PP; D), by randomized SBP intervention and quintile of baseline DBP. 242 of 9361 subjects (2.6%) (140 in the standard group and 102 in the intensive group) had missing blood pressure measurements after month 2 and are not included.

In the entire cohort, there were 562 primary outcome events over 29 278 person-years of follow-up and 365 all-cause deaths over 30 158 person-years of follow-up. In the subgroup with CKD at baseline, there were 29 kidney composite outcomes over 8490 person-years of follow-up. In the subgroup of participants without CKD at baseline, there were 164 incident CKD events over 21 155 person-years of follow-up.

Adjusted for age, sex, race, and the intervention arm, there was a U-shaped association of baseline DBP with the primary outcome, all-cause deaths, and incident CKD in cubic spline regression analyses (Figure I in the online-only Data Supplement).

Interactions of Baseline DBP and SBP Intervention for Prespecified Outcomes

In our primary assessment of the interaction between the intensive SBP intervention treatment effect and baseline DBP (Table 2), the hazard ratio for the primary outcome was 0.78 (95% confidence interval, 0.57–1.07) within the lowest DBP quintile and 0.74 (95% confidence interval, 0.61–0.90) within the upper 4 DBP quintiles (interaction P=0.78). Similarly, there was no evidence of an interaction between intensive SBP intervention and baseline DBP for all-cause death, composite kidney outcome, or incident CKD events (Table 2).

Table 2. Effects of Intensive SBP Control on the Primary and Secondary Outcomes in the Lowest DBP Quintile Compared With the Upper 4 Quintiles of Baseline DBP, Based on Intention-to-Treat Analysis

Intensive vs. Standard in Lowest DBP Quintile, HR (95% CI)Intensive vs. Standard in Top 4 DBP Quintiles, HR (95% CI)Interaction P*
Primary CVD outcome (n=9361)0.78 (0.57–1.07)0.74 (0.61–0.90)0.78
All-cause death (n=9361)0.88 (0.60–1.29)0.68 (0.53–0.87)0.29
Composite kidney outcome in CKD subgroup (n=2646)1.17 (0.36–3.84)0.79 (0.31–2.00)0.61
Incident CKD in non-CKD subgroup (n=6677)3.16 (1.42–7.00)3.58 (2.37–5.41)0.79

CI indicates confidence interval; CKD, chronic kidney disease; CVD, cardiovascular disease; DBP, diastolic blood pressure; HR, hazard ratio; and SBP, systolic blood pressure.

*HRs comparing the intensive and standard SBP interventions are presented for patients in the lowest baseline DBP quintile subgroup (left) and for patients in the upper 4 baseline DBP quintiles (right). Interaction P values evaluate whether the HRs differed between the 2 baseline DBP subgroups and were computed with likelihood ratio tests for the interaction between the randomized SBP intervention and baseline DBP subgroup in Cox regressions with separate baseline hazards for the 2 baseline DBP subgroups.

The incidence of primary outcome events, all-cause death, and incident CKD (in participants without CKD at baseline) by quintile of baseline DBP is presented in Figure 2. Within each baseline DBP quintile, participants randomized to the intensive arm had a lower incidence of the primary outcome and all-cause death and a higher incidence of CKD. There was no suggestion of heterogeneity of the hazard ratios for intensive versus standard SBP treatment effect across DBP quintiles for the 3 outcomes studied (Figure 3). The P values for interaction between treatment effect and baseline quintile of DBP were 0.92, 0.57, and 0.91 for the primary cardiovascular outcome, all-cause mortality, and incident CKD, respectively, in the subgroup without CKD at baseline.

Figure 2.

Figure 2. Incidence rates of events of interest by randomized SBP intervention and quintile of baseline DBP. Shown are incidence rates and pointwise 95% CIs for the primary cardiovascular disease (CVD) outcome (A), all-cause death (B), and incident chronic kidney disease (CKD; C) in the standard and intensive systolic blood pressure (SBP) groups by quintile of baseline diastolic blood pressure (DBP). The 95% CIs were calculated for incidence rates using the quadratic approximation to the Poisson log likelihood for the log-rate parameter. Lines are drawn between the incidence rates quintiles for the different quintiles for visual clarity, and do not represent fitted regression curves. The analysis of incident CKD patients was performed for patients with baseline estimated glomerular filtration rate ≥60 mL·min−1·1.73 m−2.There were too few events to provide a meaningful similar analysis for the composite kidney outcome.

Figure 3.

Figure 3. Forest plots with hazard ratios for the effects of SBP intervention on the events of interest by baseline DBP quintiles. Shown are forest plots with hazard ratios (HRs) for the effect of intensive vs. standard systolic blood pressure intervention by quintile of baseline diastolic blood pressure (DBP) for the primary cardiovascular disease (CVD) outcome (A), all-cause death (B), and incident chronic kidney disease (CKD; C). In joint Cox regression models with separate baseline hazards for each baseline DBP quintile, likelihood ratio tests comparing the HRs for the intensive vs. standard systolic blood pressure interventions between the 5 baseline DBP quintiles were nonsignificant (primary CVD outcome interaction, P=0.92; all-cause death interaction, P=0.57; incident CKD interaction, P=0.91; composite kidney outcome interaction, P=0.71). Because of a small number of events, the interaction test for composite kidney outcome compared HRs below and above the median baseline DBP instead of by baseline DBP quintile, and the HRs are not displayed in the figure. The analyses of patients with incident CKD and the composite kidney outcome were performed for patients with baseline estimated glomerular filtration rate ≥60 and <60 mL·min−1·1.73 m−2, respectively. CI indicates confidence interval.

After adjustment for baseline DBP, intensive versus standard SBP treatment had a lower hazard ratio for the primary CVD outcome and all-cause death but a higher hazard for incident CKD (Table 3). After adjustment for the intervention, participants with a baseline DBP of 61 mm Hg (mean DBP in the lowest baseline DBP quintile) had a higher hazard of the primary outcome, all-cause death, and incident CKD compared with those with a baseline DBP of 78 mm Hg (mean baseline DBP of the entire cohort; Table 3). There was no evidence of a nonlinear treatment by baseline DBP interaction for any of the outcomes, and the P values for the linear treatment by baseline DBP interaction did not approach statistical significance for the primary outcome (P=0.85), all-cause death (P=0.37), composite kidney outcome (P=0.57), or incident CKD events (P=0.94; Table 3). Similarly, within the subgroups with or without CVD at baseline, there was no evidence of interaction between the intervention and baseline DBP (Tables I and II in the online-only Data Supplement).

Table 3. Effects of the SBP Intervention, Baseline DBP, and the Linear Interaction Between the SBP Intervention and Baseline DBP for the Primary and Secondary Outcomes

Model 1*Model 2
Intensive vs. StandardComparison of Baseline DBP of 61 vs. 78 mm HgInteraction Term (Change in Intensive vs. Standard HR for Each 5–mm Hg Increase in DBP)
HR (95% CI)P ValueHR (95% CI)P ValueHR (95% CI)P Value
Primary CVD outcome (n=9361)0.76 (0.64–0.89)0.0011.27 (1.05–1.54)0.011.01 (0.94–1.07)0.85
All-cause death (n=9361)0.74 (0.60–0.92)0.0051.17 (0.93–1.48)0.180.96 (0.89–1.04)0.37
Composite kidney outcome in CKD subgroup (n=2646)0.92 (0.44–1.91)0.831.71 (0.78–3.75)0.180.92 (0.69–1.23)0.57
Incident CKD in non-CKD subgroup (n=6677)3.52 (2.44–5.08)<0.0011.30 (0.85–1.99)0.230.99 (0.87–1.14)0.94

CI indicates confidence interval; CKD, chronic kidney disease; CVD, cardiovascular disease; DBP, diastolic blood pressure; HR, hazard ratio; and SBP, systolic blood pressure.

*The second and third columns under model 1 display the results of Cox regression analyses relating the primary and secondary outcomes to the randomized SBP intervention (HRs in the second column) and to a cubic spline in the level of baseline DBP with knots at each baseline DBP quintile (HRs in the fourth column compare the hazards of each outcome between a baseline DBPs of 61 and 78 mm Hg), with covariable adjustment for age, sex, and race.

The fifth and sixth columns display the proportional change in the HR comparing the intensive and standard SBP interventions for each 5–mm Hg increase in DBP under model 2, which includes main effects for the randomized SBP intervention and cubic splines in baseline DBP, plus linear interactions between the randomized SBP intervention and baseline DBP. We present linear interactions between the randomized SBP intervention and baseline DBP because likelihood ratio tests evaluating interactions between the SBP intervention and cubic splines in baseline DBP indicated no evidence of nonlinear interactions (P>0.10 for each outcome).

Interactions of Baseline DBP and SBP Intervention for the Primary CVD End Point Excluding Stroke

There were 467 nonstroke CVD outcome events over 29 434 person-years of follow-up. There was a U-shaped relation between DBP and the nonstroke CVD outcome (Figure IIA in the online-only Data Supplement). As for the primary CVD outcome, there was no suggestion of heterogeneity of the hazard ratios for intensive versus standard SBP treatment effect across DBP quintiles when considering the nonstroke CVD outcome (Figure IIB in the online-only Data Supplement). There was no evidence of an interaction between baseline DBP- and the SBP-lowering intervention.

Interactions of Baseline DBP and SBP Intervention for Safety Outcomes

The incidence of safety outcomes (any serious adverse event and serious adverse events associated with hypotension, syncope, electrolyte abnormality, acute kidney injury, or acute kidney failure) is summarized in Table 4. Those in the lowest quintile of baseline DBP had the highest incidence of these serious adverse events, but there was no evidence for heterogeneity of the effects of the intervention by baseline quintile of DBP.

Table 4. Incidence (per 100 Person-Years) of Serious Adverse Events in Participants Randomized to the Intensive and Standard Treatment Groups, by Quintile of Baseline DBP (n=9361)1*

Quintile 1<68 mm HgQuintile 268–74 mm HgQuintile 375–80 mm HgQuintile 481–87 mm HgQuintile 5≥88 mm Hg
IntensiveStandardIntensiveStandardIntensiveStandardIntensiveStandardIntensiveStandard
Any serious adverse event20.5319.5515.6715.2115.3914.2712.9912.5012.6012.48
Serious adverse events associated with:
 Hypotension0.790.530.980.470.750.420.620.490.660.38
 Syncope0.980.650.700.700.930.590.660.520.460.32
 Electrolyte abnormality1.451.121.270.600.860.630.850.520.630.86
 Acute kidney injury or acute kidney failure1.831.341.450.631.550.590.880.811.100.73

DBP indicates diastolic blood pressure.

*The table presents incidence rates of the indicated adverse events expressed as number of events per 100 person-years of follow-up. In corresponding Cox regression models with separate baseline hazards for each baseline DBP quintile, likelihood ratio tests comparing the hazard ratios for the intensive and standard systolic blood pressure intervention between the 5 baseline DBP quintiles were nonsignificant (serious adverse event, interaction P=0.98; hypotension, interaction P=0.87; syncope, interaction P=0.83; electrolyte abnormality interaction, P=0.13; acute kidney injury/acute kidney failure, interaction P=0.12).

Interactions of Baseline MAP or PP and SBP Intervention for Prespecified Outcomes and Safety Outcomes

With 1 exception (acute kidney injury by baseline PP), the results were similar for baseline quintiles of MAP (Figure III and Table III in the online-only Data Supplement) and PP (Figure IV and Table IV in the online-only Data Supplement). In other words, intention-to-treat analyses yielded almost no evidence for heterogeneity in the effect of SBP lowering by baseline DBP, MAP, or PP.

Discussion

The results of the present study indicate that low baseline DBP was associated with increased risk of primary CVD outcome, but an intervention that actively lowered SBP consistently reduced the risk of the primary CVD outcome across baseline quintiles of DBP.

At some level of low BP, perfusion of organs must become inadequate. On the basis of the on-treatment reports,1723 one might expect individuals with a lower DBP to be at greater risk for adverse outcomes during intensive BP lowering. Because most ventricular myocardial perfusion occurs during diastole, a lower DBP could potentially lead to myocardial hypoperfusion and associated damage, especially in individuals with left ventricular hypertrophy (which increases oxygen demand) or coronary artery disease (in which oxygen supply is already compromised). In the ARIC (Atherosclerosis Risk In Communities) cohort, lower DBP was associated with higher serum concentrations of cardiac troponin T, a marker of myocardial injury.24

Almost all SPRINT participants were being treated for hypertension at baseline. Consistent with previous on-treatment reports, our study identified a U-shaped relationship between baseline DBP and the SPRINT primary CVD composite outcome.

Use of an intention-to-treat analysis provides a better way to determine whether the beneficial effects of intensive BP control are modified by level of baseline DBP because it takes advantage of the randomized design. In SPRINT, intensive SBP lowering that also lowered DBP was beneficial rather than hazardous even for those within the lowest quintile of baseline DBP (<68 mm Hg), where the average achieved DBP during follow-up in the intensive arm was <60 mm Hg. Our findings suggest that the association of a higher CVD event rate with lower levels of DBP is more likely to be a result of the clinical characteristics associated with a lower DBP such as age and comorbidities than a response to lowering of DBP per se.

Our findings are consistent with experience in the HOT trial (Hypertension Optimal Treatment),25 in which 6264 patients were randomly allocated to a target DBP ≤90 mm Hg, 6264 to ≤85 mm Hg, and 6262 to ≤80 mm Hg, and DBP was reduced by 20.3, 22.3, and 24.3 mm Hg, respectively. An intention-to-treat analysis identified no differences in CVD events, CVD mortality, or all-cause mortality between the 3 groups, but a J-shaped relationship was noted between achieved DBP and CVD. In the AASK trial (African American Study of Kidney Disease and Hypertension), participants with CKD were randomly assigned to a MAP target of 102 to 107 or ≤92 mm Hg.26 Although the intention-to-treat analyses by randomized groups did not show an effect of intensive BP lowering, the achieved BP analyses suggested that lower achieved MAP was associated with better kidney outcomes. Thus, analyses based on achieved BP can lead to markedly different inferences than intention-to-treat analyses, in part as a result of confounding and reverse causality.

Tissue perfusion depends on MAP. Because interventions that target lower SBP also reduce DBP, they will decrease the MAP and hence tissue perfusion. In theory, this might be particularly important in those with wider PP (with already lower DBP, and the drop in SBP might have an even greater effect on MAP). However, we also did not find evidence of heterogeneity of the effects of intensive SBP lowering by baseline MAP and PP quintiles.

A strength of the present analysis was our ability to examine the role of baseline DBP on treatment effect with the use of a randomized comparison. Other strengths included the availability of a relatively large sample size and a diverse population with relatively low pretreatment levels of DBP and a high risk for CVD. In addition, SPRINT was a rigorously conducted trial with careful measurement of BP and outcomes data. Weaknesses include the post hoc nature of the analyses and lack of intermediate biomarkers of tissue damage such as cardiac troponin. As with any subgroup analyses of randomized controlled trial data, power might be limited to definitively exclude potential harm of intensive SBP lowering on CVD outcomes in the lowest DBP quintile. Because the 95% confidence interval in the lowest DBP quintile for the primary CVD outcome ranges from 0.57 to 1.07, the potential effects of the intervention range from 43% reduction in primary CVD outcome up to a small 7% increase in risk of primary CVD outcome in this quintile.

Conclusions

Intensive SBP lowering in SPRINT participants led to substantial reductions in DBP and MAP. Although participants with lower DBP at baseline experienced higher rates of major cardiovascular events, SBP lowering appears beneficial across the spectrum of baseline DBP, even among those in the lowest quintile of DBP at baseline. Low levels of DBP, at least within the ranges examined here, should not be an impediment to intensive treatment of hypertension.

Acknowledgments

All components of the SPRINT study protocol were designed and implemented by the investigators. The investigative team collected, analyzed, and interpreted the data. All aspects of manuscript writing and revision were carried out by the coauthors. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health, the US Department of Veterans Affairs, or the US government. For a full list of contributors to SPRINT, please see the supplementary material and acknowledgment list at https://www.sprinttrial.org/public/dspScience.cfm. We also acknowledge the support from the following Clinical and Translational Science Awards funded by National Center for Advancing Translational Sciences: Case Western Reserve University, UL1TR000439; Ohio State University, UL1RR025755; University of Pennsylvania, UL1RR024134 and UL1TR000003; Boston University, UL1RR025771; Stanford University, UL1TR000093; Tufts, UL1RR025752, UL1TR000073, and UL1TR001064; University of Illinois, UL1TR000050; University of Pittsburgh, UL1TR000005; UT Southwestern, 9U54TR000017-06; University of Utah, UL1TR000105-05; Vanderbilt University, UL1 TR000445; George Washington University, UL1TR000075; University of California, Davis, UL1 TR000002; University of Florida, UL1 TR000064; University of Michigan, UL1TR000433; and Tulane University, P30GM103337 Center for Biomedical Research Excellence Award National Institute of General Medical Sciences. This work is also supported by grants from the National Institute of Diabetes and Digestive and Kidney Diseases (R01 DK091437 and R21 DK106574) and the University of Utah Study Design and Biostatistics Center (funded in part from the Public Health Services research grants UL1-RR025764 and C06-RR11234 from the National Center for Research Resources).

Footnotes

The online-only Data Supplement is available with this article at http://circ.ahajournals.org/lookup/suppl/doi:10.1161/CIRCULATIONAHA.117.030848/-/DC1.

Circulation is available at http://circ.ahajournals.org.

Correspondence to: Srinivasan Beddhu, MD, 85 N Medical Drive E, Room 201, Salt Lake City, UT 84112. E-mail

References

  • 1. Rapsomaniki E, Timmis A, George J, Pujades-Rodriguez M, Shah AD, Denaxas S, White IR, Caulfield MJ, Deanfield JE, Smeeth L, Williams B, Hingorani A, Hemingway H. Blood pressure and incidence of twelve cardiovascular diseases: lifetime risks, healthy life-years lost, and age-specific associations in 1·25 million people.Lancet. 2014; 383:1899–1911. doi: 10.1016/S0140-6736(14)60685-1.CrossrefMedlineGoogle Scholar
  • 2. Lewington S, Clarke R, Qizilbash N, Peto R, Collins R; Prospective Studies Collaboration. Age-specific relevance of usual blood pressure to vascular mortality: a meta-analysis of individual data for one million adults in 61 prospective studies.Lancet. 2002; 360:1903–1913.CrossrefMedlineGoogle Scholar
  • 3. Klag MJ, Whelton PK, Randall BL, Neaton JD, Brancati FL, Ford CE, Shulman NB, Stamler J. Blood pressure and end-stage renal disease in men.N Engl J Med. 1996; 334:13–18. doi: 10.1056/NEJM199601043340103.CrossrefMedlineGoogle Scholar
  • 4. Hsu CY, McCulloch CE, Darbinian J, Go AS, Iribarren C. Elevated blood pressure and risk of end-stage renal disease in subjects without baseline kidney disease.Arch Intern Med. 2005; 165:923–928. doi: 10.1001/archinte.165.8.923.CrossrefMedlineGoogle Scholar
  • 5. Murakami Y, Hozawa A, Okamura T, Ueshima H; Evidence for Cardiovascular Prevention From Observational Cohorts in Japan Research Group (EPOCH-JAPAN). Relation of blood pressure and all-cause mortality in 180,000 Japanese participants: pooled analysis of 13 cohort studies.Hypertension. 2008; 51:1483–1491. doi: 10.1161/HYPERTENSIONAHA.107.102459.LinkGoogle Scholar
  • 6. Law MR, Morris JK, Wald NJ. Use of blood pressure lowering drugs in the prevention of cardiovascular disease: meta-analysis of 147 randomised trials in the context of expectations from prospective epidemiological studies.BMJ. 2009; 338:b1665.CrossrefMedlineGoogle Scholar
  • 7. Xie X, Atkins E, Lv J, Bennett A, Neal B, Ninomiya T, Woodward M, MacMahon S, Turnbull F, Hillis GS, Chalmers J, Mant J, Salam A, Rahimi K, Perkovic V, Rodgers A. Effects of intensive blood pressure lowering on cardiovascular and renal outcomes: updated systematic review and meta-analysis.Lancet. 2016; 387:435–443. doi: 10.1016/S0140-6736(15)00805-3.CrossrefMedlineGoogle Scholar
  • 8. Wright JT, Williamson JD, Whelton PK, Snyder JK, Sink KM, Rocco MV, Reboussin DM, Rahman M, Oparil S, Lewis CE, Kimmel PL, Johnson KC, Goff DC, Fine LJ, Cutler JA, Cushman WC, Cheung AK, Ambrosius WT; SPRINT Research Group. A randomized trial of intensive versus standard blood-pressure control.N Engl J Med. 2015; 373:2103–2116. doi: 10.1056/NEJMoa1511939.CrossrefMedlineGoogle Scholar
  • 9. Williamson JD, Supiano MA, Applegate WB, Berlowitz DR, Campbell RC, Chertow GM, Fine LJ, Haley WE, Hawfield AT, Ix JH, Kitzman DW, Kostis JB, Krousel-Wood MA, Launer LJ, Oparil S, Rodriguez CJ, Roumie CL, Shorr RI, Sink KM, Wadley VG, Whelton PK, Whittle J, Woolard NF, Wright JT, Pajewski NM; SPRINT Research Group. Intensive vs standard blood pressure control and cardiovascular disease outcomes in adults aged ≥75 years: a randomized clinical trial.JAMA. 2016; 315:2673–2682. doi: 10.1001/jama.2016.7050.CrossrefMedlineGoogle Scholar
  • 10. Cruickshank JM, Thorp JM, Zacharias FJ. Benefits and potential harm of lowering high blood pressure.Lancet. 1987; 1:581–584.CrossrefMedlineGoogle Scholar
  • 11. Farnett L, Mulrow CD, Linn WD, Lucey CR, Tuley MR. The J-curve phenomenon and the treatment of hypertension: is there a point beyond which pressure reduction is dangerous?JAMA. 1991; 265:489–495.CrossrefMedlineGoogle Scholar
  • 12. Ambrosius WT, Sink KM, Foy CG, Berlowitz DR, Cheung AK, Cushman WC, Fine LJ, Goff DC, Johnson KC, Killeen AA, Lewis CE, Oparil S, Reboussin DM, Rocco MV, Snyder JK, Williamson JD, Wright JT, Whelton PK; SPRINT Study Research Group. The design and rationale of a multicenter clinical trial comparing two strategies for control of systolic blood pressure: the Systolic Blood Pressure Intervention Trial (SPRINT).Clin Trials. 2014; 11:532–546. doi: 10.1177/1740774514537404.CrossrefMedlineGoogle Scholar
  • 13. NHLBI. Systolic Blood Pressure Intervention Protocol version 4.0.https://www.sprinttrial.org/public/Protocol_Current.pdf. Accessed August 22, 2017.Google Scholar
  • 14. D’Agostino RB, Vasan RS, Pencina MJ, Wolf PA, Cobain M, Massaro JM, Kannel WB. General cardiovascular risk profile for use in primary care: the Framingham Heart Study.Circulation. 2008; 117:743–753. doi: 10.1161/CIRCULATIONAHA.107.699579.LinkGoogle Scholar
  • 15. Levey AS, Bosch JP, Lewis JB, Greene T, Rogers N, Roth D. A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation: Modification of Diet in Renal Disease Study Group.Ann Intern Med. 1999; 130:461–470.CrossrefMedlineGoogle Scholar
  • 16. National Heart, Lung and Blood Institute. Systolic Blood Pressure Intervention Trial Primary Outcome Paper (SPRINT-POP) data.https://biolincc.nhlbi.nih.gov/studies/sprint_pop/. Accessed March 3, 2017.Google Scholar
  • 17. Messerli FH, Mancia G, Conti CR, Hewkin AC, Kupfer S, Champion A, Kolloch R, Benetos A, Pepine CJ. Dogma disputed: can aggressively lowering blood pressure in hypertensive patients with coronary artery disease be dangerous?Ann Intern Med. 2006; 144:884–893.CrossrefMedlineGoogle Scholar
  • 18. Vidal-Petiot E, Ford I, Greenlaw N, Ferrari R, Fox KM, Tardif JC, Tendera M, Tavazzi L, Bhatt DL, Steg PG; CLARIFY Investigators. Cardiovascular event rates and mortality according to achieved systolic and diastolic blood pressure in patients with stable coronary artery disease: an international cohort study.Lancet. 2016; 388:2142–2152. doi: 10.1016/S0140-6736(16)31326-5.CrossrefMedlineGoogle Scholar
  • 19. Sim JJ, Shi J, Kovesdy CP, Kalantar-Zadeh K, Jacobsen SJ. Impact of achieved blood pressures on mortality risk and end-stage renal disease among a large, diverse hypertension population.J Am Coll Cardiol. 2014; 64:588–597. doi: 10.1016/j.jacc.2014.04.065.CrossrefMedlineGoogle Scholar
  • 20. Bangalore S, Messerli FH, Wun CC, Zuckerman AL, DeMicco D, Kostis JB, LaRosa JC; Treating to New Targets Steering Committee and Investigators. J-curve revisited: an analysis of blood pressure and cardiovascular events in the Treating to New Targets (TNT) Trial.Eur Heart J. 2010; 31:2897–2908. doi: 10.1093/eurheartj/ehq328.CrossrefMedlineGoogle Scholar
  • 21. Kang YY, Wang JG. The J-curve phenomenon in hypertension.Pulse (Basel). 2016; 4:49–60. doi: 10.1159/000446922.CrossrefMedlineGoogle Scholar
  • 22. Somes GW, Pahor M, Shorr RI, Cushman WC, Applegate WB. The role of diastolic blood pressure when treating isolated systolic hypertension.Arch Intern Med. 1999; 159:2004–2009.CrossrefMedlineGoogle Scholar
  • 23. Fagard RH, Staessen JA, Thijs L, Celis H, Bulpitt CJ, de Leeuw PW, Leonetti G, Tuomilehto J, Yodfat Y. On-treatment diastolic blood pressure and prognosis in systolic hypertension.Arch Intern Med. 2007; 167:1884–1891. doi: 10.1001/archinte.167.17.1884.CrossrefMedlineGoogle Scholar
  • 24. McEvoy JW, Chen Y, Rawlings A, Hoogeveen RC, Ballantyne CM, Blumenthal RS, Coresh J, Selvin E. Diastolic blood pressure, subclinical myocardial damage, and cardiac events: implications for blood pressure control.J Am Coll Cardiol. 2016; 68:1713–1722. doi: 10.1016/j.jacc.2016.07.754.CrossrefMedlineGoogle Scholar
  • 25. Hansson L, Zanchetti A, Carruthers SG, Dahlöf B, Elmfeldt D, Julius S, Ménard J, Rahn KH, Wedel H, Westerling S. Effects of intensive blood-pressure lowering and low-dose aspirin in patients with hypertension: principal results of the Hypertension Optimal Treatment (HOT) randomised trial: HOT Study Group.Lancet. 1998; 351:1755–1762.CrossrefMedlineGoogle Scholar
  • 26. Davis EM, Appel LJ, Wang X, Greene T, Astor BC, Rahman M, Toto R, Lipkowitz MS, Pogue VA, Wright JT; African American Study of Kidney Disease and Hypertension Research Collaborative Group. Limitations of analyses based on achieved blood pressure: lessons from the African American Study of Kidney Disease and Hypertension Trial.Hypertension. 2011; 57:1061–1068. doi: 10.1161/HYPERTENSIONAHA.111.169367.LinkGoogle Scholar

eLetters(0)

eLetters should relate to an article recently published in the journal and are not a forum for providing unpublished data. Comments are reviewed for appropriate use of tone and language. Comments are not peer-reviewed. Acceptable comments are posted to the journal website only. Comments are not published in an issue and are not indexed in PubMed. Comments should be no longer than 500 words and will only be posted online. References are limited to 10. Authors of the article cited in the comment will be invited to reply, as appropriate.

Comments and feedback on AHA/ASA Scientific Statements and Guidelines should be directed to the AHA/ASA Manuscript Oversight Committee via its Correspondence page.