Introduction

Hypertension is an independent risk factor for coronary events, stroke, heart failure, and ESKD.1 Impaired natriuresis is one of the fundamental perturbations which leads to development of essential hypertension and hypertension associated with CKD.2,3 Contemporary guidelines recommend thiazide or thiazide-like diuretics as one of the first-line agents for management of hypertension.4,5 It has been suggested that the beneficial effects on major cardiovascular disease end points and all-cause mortality observed in the intensive systolic BP-lowering group in the Systolic Blood Pressure Intervention Trial (SPRINT) were due to differential use of diuretics in the two groups.6 In a recent publication that conducted an observational time-varying analysis of postrandomization antihypertensive use in SPRINT, it was reported that more than a year of thiazide use was associated with lower risk of cardiovascular end points after adjusting for the randomized BP group and BP reduction.7 From those analyses, the authors concluded that greater use of thiazide diuretics in the intensive systolic BP group might have contributed to the beneficial effects of the BP intervention.

This clinically relevant question of whether postrandomization diuretics use at least partly account for the effects of intensive systolic BP lowering in SPRINT can be statistically explored using two direct approaches: first, whether adjusting for time-varying postrandomization diuretics use attenuated the beneficial effects of intensive systolic BP lowering in regression models and second by conducting formal mediation analyses to examine whether postrandomization diuretics use mediated the effects of the BP intervention on outcomes. We conducted these analyses to test whether postrandomization diuretics use at least partly explained or mediated the beneficial effects of intensive systolic BP lowering.

Methods

SPRINT was a randomized, controlled, open-label trial of 9361 participants designed to compare the effects of intensive (systolic BP target <120 mm Hg) versus standard (systolic BP target <140 mm Hg) BP control on a primary composite cardiovascular end point in participants aged 50 years or older with a baseline systolic BP of 130–180 mm Hg (on none or up to four antihypertensive medications depending on the range of systolic BP) and a higher risk of cardiovascular disease.8 The details of the SPRINT protocol including inclusion and exclusion criteria, interventions, measurements, and follow-up have been previously published.9

SPRINT Systolic BP Interventions and Follow-Up

An automated measurement device (Model 907XL, Omron Healthcare) was used to record BP at clinic visits after the participant had been seated for 5 minutes of quiet rest. The recorded BP was based on an average of three BP readings taken 1 minute apart. Participants were seen monthly for the first 3 months and every 3 months thereafter. Additional monthly visits were scheduled as needed. Lifestyle modification was encouraged, and antihypertensive medications were adjusted on the basis of clinical judgment to target a systolic BP <120 mm Hg in the intensive group and 135–139 mm Hg in the standard group.

Participant recruitment began on November 8, 2010, and a decision to terminate the SPRINT systolic BP interventions was taken on August 20, 2015, because the hazard ratio (HR) for the primary composite end point exceeded the monitoring boundary in favor of intensive systolic BP lowering at two consecutive time points. This analysis includes follow-up during the trial phase, that is, up to August 20, 2015.

Postrandomization Diuretics Use

Diuretics were classified in SPRINT as follows: thiazide-type, which included both thiazide (hydrochlorothiazide, chlorothiazide, and metolazone) and thiazide-like diuretics (chlorthalidone and indapamide); loop diuretics (furosemide, torsemide, and bumetanide); and potassium (K)-sparing diuretics (spironolactone, triamterene, and amiloride). Chlorthalidone, spironolactone, furosemide, hydrochlorothiazide/triamterene, and amiloride were on the SPRINT formulary. Chlorthalidone was encouraged as the diuretic of choice. The protocol allowed the use of off-formulary diuretics, but these were neither provided nor paid for by the trial organization. For this report, we categorized diuretics use as follows: no diuretics use, thiazide-type diuretics (alone or in combination with loop diuretics and/or K-sparing diuretics), and non–thiazide-type diuretics (loop and/or K-sparing [loop/K] diuretics without thiazides). Loop and K-sparing diuretics were grouped together because few participants were on these medications.

SPRINT End Points

The SPRINT primary end point was a composite of first occurrence of myocardial infarction, acute coronary syndrome not resulting in myocardial infarction, stroke, acute decompensated heart failure, or death from cardiovascular causes. Death from any cause was a predefined secondary outcome. A committee blinded to the study group assignments adjudicated these end points as specified in the SPRINT protocol.9

Statistical Methods Time-Varying Cox Regression Models

First, we conducted intention-to-treat analyses relating the systolic BP intervention to the primary composite end point and all-cause mortality via proportional hazards (Cox) regression. Next, to test whether adjusting for postrandomization diuretics use attenuated the effects of the BP intervention on the primary composite end point and all-cause mortality, we compared the effect of intensive versus standard systolic BP control before and after adjusting for time-varying diuretics use, assigning participants to one of the three categories described above over successive 3-month intervals throughout follow-up. We considered participants to be on a diuretic if they were prescribed any diuretics in that category during that 3-month period (with no diuretics as the reference level) and related the baseline systolic BP control intervention and the diuretics prescription to the risks of end points in the subsequent 3 months. In other words, baseline diuretics prescription was related to events during the first 3 months, the first 3 months diuretics prescription was related to events during the following 3 months, and so on. In both models, we also adjusted for baseline age, sex, self-reported race, ethnicity, smoking, body mass index (BMI), comorbidities (coronary artery disease, heart failure, stroke, peripheral vascular disease, and cancer), baseline BP, CKD stage, urine albumin-creatinine ratio, baseline diuretics use, other antihypertensive medications, and number of antihypertensive medications, stratified by clinical sites.

Mediation Analyses

To formally assess the role of diuretics in the effects of intensive systolic BP control on these outcomes, we conducted mediation analysis under the modern causal inference framework. We assigned participants into a diuretic category on the basis of their status by month 6, when most of the participants had the adjustment of their antihypertensive regimen completed to achieve assigned goals. We extend the widely used Cox model-based approach for binary mediator with survival outcome 10 to our case where the mediator, diuretics use, is categorical with three groups (no diuretics, thiazide-type, and non–thiazide-type). The total effect of intensive systolic BP control goal on a given end point was disaggregated into direct (i.e., the effect of intensive systolic BP control on the end point that occurred through pathways other than affecting diuretics use) and indirect or mediation effects (i.e., the effect of intensive systolic BP control on the end point that occurred through diuretics use). We used the bootstrap method to estimate the confidence interval (CI) for each of the effects. We conducted similar mediation analyses for other antihypertensive medication classes.

Effect Modification

In addition, we compared the regression coefficients of diuretics use with the primary composite end point and all-cause mortality in separate Cox models using time-varying diuretics use within the standard and intensive systolic BP-lowering groups to test whether these associations differed by randomized group. Owing to smaller sample sizes (50% of total), these models were not adjusted for stratification by clinic site.

We performed all analyses using Stata, version 15.1 and R Studio 2022.07.0.

Results

The current analysis included all 9361 SPRINT participants. Supplemental Figure 1 shows CONSORT flowchart of eligible participants. The mean age of the study population was 68±9 years; 36% were women, 30% were non-Hispanic Black, and 10% were Hispanic (Table 1). At baseline, there was similar diuretics use in the standard (48%) and intensive (47%) systolic BP-lowering groups. Thiazide type (40%) was the most common diuretics followed by loop (6%) and K sparing (4%). Participants on diuretics at baseline compared with those not on diuretics were more likely to be female, race other than non-Hispanic White, had a higher BMI, more frequent heart failure and CKD, and were more likely to be treated with a higher number of antihypertensive medications and had more frequent use of angiotensin-converting enzyme inhibitors or angiotensin receptor blockers (Table 1). Participants on loop or K-sparing diuretics had a higher BMI, more frequent comorbidities such as coronary artery disease, heart failure, advanced CKD, and higher number of antihypertensive medications as compared with participants on thiazide-type diuretics.

Table 1 - Baseline characteristics of participants by baseline diuretics use in three categories Variable Total Not on Diuretics Thiazide-Type Diuretics Loop/K-Sparing Diuretics N=9361 n=4950 n=3789 n=622 Age, yr 68±9.4 68±9.5 67±9.2 71±9.8 Female, No. (%) 3332 (36) 1520 (31) 1540 (41) 272 (44) Hispanic, No. (%) 984 (11) 607 (12) 333 (9) 44 (7) Non-Hispanic Black, No. (%) 2802 (30) 1198 (24) 1388 (37) 216 (35) Non-Hispanic White, No. (%) 5399 (58) 3031 (61) 2011 (53) 357 (57) Other, No. (%) 176 (2) 114 (2) 57 (2) 5 (1) BMI, kg/m2 29.9±5.8 29.1±5.5 30.4±5.8 32.0±6.9 Systolic BP, mm Hg 140±16 141±16 138±15 138±15 Diastolic BP, mm Hg 78±12 79±12 78±12 74±12 Randomization, No. (%)  Standard group 4683 (50) 2439 (49) 1938 (51) 306 (49)  Intensive group 4678 (50) 2511 (51) 1851 (49) 316 (51) Smoking history, No. (%)  Current 1240 (13) 690 (14) 495 (13) 55 (9)  Former 3973 (43) 2070 (42) 1616 (43) 287 (46)  Never 4122 (44) 2175 (44) 1667 (44) 280 (45) Comorbidities, No. (%)  Coronary artery disease 1206 (13) 721 (15) 357 (9) 128 (21)  Stroke 48 (1) 17 (0) 22 (1) 9 (1)  Peripheral vascular disease 503 (5) 256 (5) 191 (5) 56 (9)  Congestive heart failure 326 (4) 140 (3) 92 (2) 94 (15)  Cancer 1131 (12) 613 (12) 425 (11) 93 (15) CKD stage, No. (%)  eGFR >60 ml/min 6662 (71) 3688 (75) 2700 (71) 274 (44)  eGFR 30 to <60 ml/min 2490 (27) 1165 (24) 1035 (27) 290 (47)  eGFR <30 ml/min 155 (2) 66 (1) 32 (1) 57 (9) Albuminuria, No. (%)  <30 mg/g 7182 (77) 3791 (77) 2700 (71) 274 (44)  30–300 mg/g 1482 (16) 807 (16) 1035 (27) 290 (47)  >300 mg/g 248 (3) 133 (3) 32 (1) 57 (9) Medications, No. (%)  ACEi/ARB 5469 (58) 2697 (55) 2387 (63) 385 (62)  Calcium channel blockers 3319 (36) 1775 (36) 1247 (33) 297 (48)   β-blocker 3413 (37) 1749 (35) 1295 (34) 369 (59)  Statin 4063 (43) 2117 (43) 1605 (42) 341 (55) No. of antihypertensives, No. (%)  Zero 851 (9) 851 (17) 0 (0) 0 (0)  One 2497 (27) 2072 (42) 389 (10) 36 (6)  Two 3170 (34) 1486 (30) 1541 (41) 143 (23)  Three 2105 (23) 479 (10) 1348 (36) 278 (45)  Four or more 738 (8) 62 (1) 511 (14) 165 (27)

ACEi, angiotensin converting enzyme inhibitor; ARB, angiotensin receptor blocker; BMI, body mass index.

Follow-Up Diuretics Use

The median duration of follow-up was 3.3 (interquartile range, 2.8–3.8) years. The average proportion of standard group participants on diuretics during follow-up (47%) was not different from baseline (48%) (Figure 1). However, the proportion of intensive group participants on diuretics increased from 47% at baseline to 76% by month 6, with an average proportion during follow-up of 74%. At month 6, the proportion on thiazide-type diuretics was 41% in the standard group and 67% in the intensive group (Figure 1). Overlap of postrandomization diuretics use by intensive and standard groups is summarized in Supplemental Table 1.

Figure 1:

Number of participants taking diuretics in three categories (thiazide type, loop/potassium sparing, and no diuretics use) in the two SPRINT systolic BP interventions groups (standard [Std] and intensive [Int]) during the period of study). SPRINT, Systolic Blood Pressure Intervention Trial.

Primary Composite Cardiovascular End Point by Time-Varying Diuretics Use and Mediation Analysis in the Entire Cohort

There were 243 events over 17,129 person-years of follow-up (1.65% per year) in the intensive group and 319 over 16,933 person-years of follow-up (2.19% per year) in the standard group. There were 232 events in participants who used thiazide-type diuretics anytime during the trial with a cumulative diuretics exposure of 14,833 person-years (1.56% per year) and 94 events in participants who used loop/K-sparing diuretics over 2498 person-years (3.76% per year). A total of 236 events occurred in participants who did not use diuretics during the trial with a cumulative 11,927 person-years of follow-up (1.98% per year) (Figure 2A).

Figure 2:

Effect of systolic BP interventions on primary cardiovascular composite end points and all-cause mortality by three categories of diuretics use. Shown are the primary composite end point (a composite end point comprising the first occurrence of a of nonfatal myocardial infarction, acute coronary syndrome not resulting in myocardial infarction, stroke, acute decompensated heart failure, or death from cardiovascular causes) in SPRINT participants by their diuretics use. (A) Forest plots with HRs for the effect of two SPRINT systolic BP interventions on the primary composite end point in the entire cohort using the intention-to-treat analysis (n=9361) (first row) and in a time-varying model for diuretics use (second row). The time-varying model was adjusted for longitudinal diuretics use, baseline diuretic use, age, sex, race/ethnicity, smoking, BMI, comorbidities, baseline BP, CKD staging, UACR, other antihypertensive medications, number of antihypertensive medications and intervention group, and stratified by clinical sites. The bottom three rows show the HR for end points in the participants treated with no diuretics, thiazide-type diuretics, and loop/K-sparing diuretics, respectively. (B) The results of mediation analysis in forest plots with HRs by direct effects (mediated by pathways other than two types of diuretics use), indirect effect (mediated by two types of diuretics use), and total effect. (C and D) The corresponding forest plots for the prespecified secondary end point of all-cause mortality. Event rates are in percent per year. BMI, body mass index; CI, confidence interval; HR, hazard ratio; UACR, urine albumin-creatinine ratio.

In the intention-to-treat analysis, intensive systolic BP lowering resulted in a 25% reduction (HR, 0.75; 95% CI, 0.64 to 0.89) in risk for the primary composite end point (Figure 2A) as reported in the SPRINT primary manuscript.8 The magnitude of the beneficial effect of intensive systolic BP lowering on the primary composite end point (HR, 0.74; 95% CI, 0.62 to 0.89) was similar in a multivariable Cox model after adjustment for time-varying diuretics use (Figure 2A). Across the entire population, time-varying diuretics use per se was not associated with the primary composite end point (compared with no diuretics use, HR for thiazide type, 0.89; 95% CI, 0.73 to 1.10 and for loop/K sparing, 1.29; 95% CI, 0.97 to 1.73).

For the analysis that examined whether diuretics use in the first 6 months postrandomization mediated the effects of the intervention on the cardiovascular composite beyond 6 months, there were 197 events over 12,407 years of follow-up (1.59% per year) in the intensive group and 280 over 12,266 years of follow-up (2.28% per year) in the standard group. In the mediation analysis, the HRs for total effect, direct effect (not mediated through diuretics use), and indirect effect (mediated through diuretics) of the intervention on the outcomes were 0.66 (95% CI, 0.54 to 0.79), 0.67 (95% CI, 0.54 to 0.81), and 0.98 (95% CI, 0.88 to 1.10), respectively (Figure 2B). The corresponding HRs for total, direct, and indirect effects for angiotensin-converting enzyme inhibitor/angiotensin receptor blocker, β blocker, and calcium channel blocker on the cardiovascular composite end points are shown in Supplemental Figure 2.

All-Cause Mortality by Time-Varying Diuretics Use and Mediation Analysis in the Entire Cohort

There were 155 deaths over 15,086 years of follow-up (1.03% per year) in the intensive group and 210 over 15,061 years of follow-up (1.39% per year) in the standard group. There were 141 deaths over 15,113 person years of follow-up (0.93% per year) in participants who used thiazide-type diuretics anytime during the trial, 72 deaths over 2704 person-years of follow-up (2.66% per year) in participants used loop/K-sparing diuretics, and 152 events over 12,319 person-years of follow-up (1.23% per year) in participants who did not use diuretics.

Intensive systolic BP lowering resulted in a lower risk of all-cause mortality in the intention-to-treat model (HR, 0.73; 95% CI, 0.60 to 0.90) and in a model incorporating time-varying diuretics use (HR, 0.75; 95% CI, 0.60 to 0.93) (Figure 2C). In the time-varying diuretics use model, compared with no diuretics use, HR for thiazide type was 0.88; 95% CI, 0.68 to 1.14 and for loop/K sparing, 1.47; 95% CI, 1.04 to 2.07.

For the mediation analysis that examined whether diuretics use in the first 6 months postrandomization mediated the effects of the intervention on end point after 6 months, there were 144 deaths over 12,778 years of follow-up (1.12% per year) in the intensive group and 190 over 12,749 years of follow-up (1.49% per year) in the standard group. In this analysis, HRs for total effect, direct effect (not mediated through diuretics use), and indirect effect (mediated through diuretics) of the intervention on the cardiovascular composite were 0.75 (95% CI, 0.59 to 0.95), 0.71 (95% CI, 0.55 to 0.94), and 1.05 (95% CI, 0.93 to 1.20), respectively (Figure 2D).

Associations of Diuretics Use with End Points within Standard and Intensive BP Groups

Within the standard BP group, compared with no diuretics use, neither thiazide-type nor loop/K-sparing diuretics use was associated with the primary composite end point (Figure 3A) or all-cause mortality (Figure 3B). However, within the intensive BP group, compared with no diuretics use, thiazide-type use associated with lower risk of primary composite end point (Figure 3A) and trend toward lower risk of all-cause mortality (Figure 3B). By contrast, loop/K-sparing diuretics use was not associated with primary composite end point and all-cause mortality (Figure 3, A and B).

Figure 3:

Effect modification of two systolic BP interventions and three diuretics categories on primary cardiovascular composite end points and all-cause mortality. (A) The forest plots with HRs of association of use of different diuretics categories with primary cardiovascular composite end point in standard (n=4683) and intensive (n=4678) treatment groups. (B) The corresponding forest plots for the prespecified secondary end point of all-cause mortality. Event rates are in percent per year.

Discussion

In this post hoc analysis of SPRINT, the beneficial effects of intensive systolic BP lowering on cardiovascular end points and all-cause mortality were independent of postrandomization time-varying diuretics use. Furthermore, in mediation analyses, there was no evidence that the effects of intensive systolic BP lowering on cardiovascular end points and all-cause mortality were mediated through postrandomization diuretics use. Thus, these data suggest that postrandomization diuretics use did not explain the beneficial effects of intensive systolic BP treatment on the cardiovascular end points and all-cause mortality. However, in subgroup analyses, it appears that compared with no diuretics use, thiazide diuretics use was associated with lower risk of the primary composite (cardiovascular) end point and trend toward lower risk of all-cause mortality among participants randomized to intensive systolic BP group but not to the standard systolic BP group.

In the Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial (ALLHAT), thiazide-type diuretics were superior to several other agents in preventing major forms of cardiovascular disease.11 On the basis of ALLHAT and other randomized clinical trial results,12 the Seventh Report of the Joint National Committee recommended thiazide-type diuretics, alone or in combination with other antihypertensive drug classes, as the initial drug therapy for hypertension.13 The subsequent Avoiding Cardiovascular Events through Combination Therapy in Patients Living with Systolic Hypertension trial reported a 20% advantage in cardiovascular risk reduction when BP was lowered using the single-pill combination of benazepril-amlodipine as compared with benazepril-hydrochlorothiazide.14 The better outcomes in the ALLHAT diuretics group have been attributed to utilization of the longer acting and more potent (relative to hydrochlorothiazide) agent chlorthalidone. Chlorthalidone has also been demonstrated to reduce BP effectively in advanced CKD.15 In SPRINT, chlorthalidone was the predominant diuretic used in the intensive systolic BP-lowering group to achieve the BP goal. However, in recently published VA Cooperative Studies Program randomized pragmatic trial, patients who received chlorthalidone did not have a lower occurrence of major cardiovascular end points compared with hydrochlorothiazide, although 95% of patients were on lower doses (25 mg of hydrochlorothiazide or 12.5 mg of chlorthalidone) than those typically used in major hypertension outcome trials, including SPRINT in which 60% of intensive systolic BP-lowering group participants on chlorthalidone were on the higher (25 mg) dose.16,17

More frequent use of thiazide-type diuretics in the intensive systolic BP-lowering group has been raised as a possible explanation for the positive results in the SPRINT.6 The results of the current analysis do not support this contention because adjusting for time-varying diuretics use did not attenuate the beneficial effects of the BP intervention in regression models, and there was no evidence that the beneficial effects of BP intervention on outcomes were mediated through postrandomization diuretics use. In a previous observational analysis of SPRINT, it was reported that time-varying thiazide diuretic use of more than 1 year compared with less than a year of thiazide use was associated with lower risk of the primary cardiovascular end points in the entire cohort.7 In this analysis, we further separated participants not on thiazides as those on no diuretics or those on loop/K-sparing diuretics as use of loop diuretics might be indicated in participants with heart failure or CKD (or more severe disease) which in turn might be associated with worse outcomes. When analyzing the entire SPRINT cohort, neither thiazide-type nor loop/K-sparing diuretics use was significantly associated with major cardiovascular events or all-cause mortality. However, among participants randomized to the intensive systolic BP-lowering group, thiazide-type diuretics use was associated with lower risk of cardiovascular end points and a trend toward lower all-cause mortality relative to participants not treated with diuretics suggesting that there might be a benefit to thiazide-type diuretics use when used in conjunction with intensive systolic BP lowering. By contrast, intensive systolic BP lowering yielded beneficial cardiovascular effects in those not on diuretics and those on thiazide-type diuretics. In the small number of participants who received loop/K-sparing diuretics, intensive systolic BP lowering had no effect on cardiovascular end points with a trend toward higher all-cause mortality which likely indicates abovementioned indication bias. Caution is warranted in interpreting nonsignificant results of a postrandomization analysis of a small subgroup with a possible indication for loop/K-sparing diuretics. Our results corroborate to a subanalysis of SPRINT which included only intensive group and demonstrated lower risk of cardiovascular diseases with thiazide use.18

A trial with SPRINT inclusion criteria but excluding CKD, targeting intensive systolic BP control, and randomizing participates to thiazide-type diuretics versus no diuretic on top of other antihypertensive medications as needed, might tease out the beneficial effect of intensive BP control from thiazide-type diuretics. However, a small pilot trial to assess the feasibility of achieving intensive BP control without diuretics will be needed first before any such trial can be planned.

Strengths of our analysis includes the use of data from a large, randomized study that compared the effects of intensive versus standard systolic BP control, with collection of an extensive array of clinical and laboratory data at baseline and during follow-up and with adjudicated end points. The study cohort was racially diverse, and a substantial fraction of the cohort was older than 75 years, allowing generalization to these important subgroups. Despite the limitations of a postrandomization analysis, our approach determined whether the effects of the intervention on major cardiovascular events and/or all-cause mortality were attenuated by postrandomization diuretics use in regression analyses or explained by postrandomization diuretics use in mediation analyses. Thus, our study attempted to answer the question of whether the effects of the BP intervention on outcomes were independent of postrandomization diuretics use.

Our study also has several limitations. First, we cannot definitively infer that use of a thiazide-type diuretic in patients treated with an intensive systolic BP-lowering strategy would yield more favorable outcomes relative to other antihypertensive agents. Future randomized controlled trials would be required to answer that question. Second, diabetes mellitus, a disease state strongly associated with impaired natriuresis was an exclusion criterion in SPRINT.19,20 Although more than 40% of the SPRINT participants were considered to have prediabetes, these participants were shown to experience a similar benefit as participants with normoglycemia.21 Third, this analysis did not examine adverse events associated with thiazide-type diuretics. Finally, only 30% of the SPRINT participants had albuminuria >30 mg/g and only 6% had eGFR <30 ml/min per 1.73 m2 at baseline; thus, we cannot make definitive conclusions for patients with more advanced or proteinuric CKD.

In conclusion, irrespective of time-varying diuretics use, intensive control of systolic BP lowered the risk of major adverse cardiovascular end points and all-cause death in adults older than 50 years with hypertension at high risk of cardiovascular disease. These results also suggest but do not provide conclusive evidence that thiazide-type diuretics may be of benefit when adopting a strategy of more intensive systolic BP lowering, as was employed in SPRINT.

Disclosures

S. Bansal reports consultancy for Calliditas NA Enterprises Inc., Daxor Corporation, and PD Excellence Academy, a branch of Baxter Healthcare Corporation; research funding from 3ive labs, AstraZeneca, Bayer, Boehringer Ingelheim, NIH, and Novo Nordisk; royalties from UpToDate; honoraria from Home Dialysis University; advisory or leadership role on the Editorial Boards of CJASN and Kidney360 and Section Editor for Clinical Nephrology Journal; and speakers bureau for Home Dialysis University. S. Beddhu reports research funding from Bayer, Boehringer Ingelheim, NIH, Novo Nordisk, and VA; royalties from UpToDate; and an advisory or leadership role for NIH Study Section. G.M. Chertow reports consultancy for Akebia, Ardelyx, AstraZeneca, Calico, Gilead, Miromatrix, Reata, Sanifit, Unicycive, and Vertex; ownership interest in Ardelyx, CloudCath, Durect, DxNow, Eliaz Therapeutics, Outset, Renibus, and Unicycive; research funding from CSL Behring, NIAID, and NIDDK; advisory or leadership role on Board of Directors of Satellite Healthcare, a nonprofit dialysis provider, and as Co-Editor of Brenner & Rector's The Kidney (Elsevier); role as Chair or Co-Chair of Trial Steering Committees with Akebia, AstraZeneca, CSL Behring, Sanifit, and Vertex; role as an advisor to Applaud, Ardelyx, CloudCath, Durect, Eliaz Therapeutics, Miromatrix, Outset, Renibus, and Unicycive; and DSMB service for Bayer, Gilead, Mineralys, NIDDK, and ReCor. A.K. Cheung reports consultancy for 3D Communications, Alucent, Boehringer-Ingelheim, and CSL Behring; ownership interest in Merck; patents or royalties from UpToDate; and advisory or leadership role for KDIGO. W.C. Cushman reports research funding from George Medicines, NIH, and ReCor Medical. All remaining authors have nothing to disclose.

Funding

S. Beddhu: NHLBI Division of Intramural Research (R21HL145494), Office of Rural Health (VA ORH-14398), National Institute of Diabetes, Digestive and Kidney Diseases (R01DK128640 and R01DK118219) and National Institute of Aging (R01AG074592).

Acknowledgments

We thank the SPRINT (Systolic Blood Pressure Intervention Trial) study participants and trial investigators. This article was prepared using SPRINT research materials obtained from the National Heart, Lung, and Blood Institute (NHLBI) Biologic Specimen and Data Repository Information Coordinating Center and does not necessarily reflect the opinions or views of SPRINT or the NHLBI.

Author Contributions

Conceptualization: Shweta Bansal, Srinivasan Beddhu, Alfred K. Cheung, William C. Cushman.

Data curation: Shweta Bansal, Srinivasan Beddhu.

Formal analysis: Shweta Bansal, Srinivasan Beddhu, Robert Boucher, Jincheng Shen, Guo Wei.

Investigation: Shweta Bansal, Srinivasan Beddhu, Robert Boucher, Glenn M. Chertow, Jincheng Shen, Guo Wei, Paul K. Whelton.

Methodology: Shweta Bansal, Srinivasan Beddhu, Alfred K. Cheung, William C. Cushman, Paul K. Whelton.

Project administration: Shweta Bansal.

Supervision: Shweta Bansal, Srinivasan Beddhu.

Validation: Glenn M. Chertow.

Writing – original draft: Shweta Bansal.

Writing – review & editing: Shweta Bansal, Srinivasan Beddhu, Robert Boucher, Glenn M. Chertow, Alfred K. Cheung, William C. Cushman, Jincheng Shen, Guo Wei, Paul K. Whelton.

Data Sharing Statement

Previously published data were used for this study. https://biolincc.nhlbi.nih.gov/home/.

Supplemental Material

This article contains the following supplemental material online at https://links.lww.com/CJN/B850.

Supplemental Table 1. Diuretics use type by standard and intensive systolic BP groups, per post baseline visit.

Supplemental Figure 1. CONSORT flowchart.

Supplemental Figure 2. Mediation analyses for the other antihypertensive medication classes on primary cardiovascular composite end points.

References 1. 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(9349):1903–1913. doi:10.1016/s0140-6736(02)11911-8 2. Kimura G, Brenner BM. A method for distinguishing salt-sensitive from non-salt-sensitive forms of human and experimental hypertension. Curr Opin Nephrol Hypertens. 1993;2(3):341–349. 3. Guyton AC. Roles of the kidneys and fluid volumes in arterial pressure regulation and hypertension. Chin J Physiol. 1989;32(2):49–57. 4. Whelton PK, Carey RM, Aronow WS, et al. 2017 ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA guideline for the prevention, detection, evaluation, and management of high blood pressure in adults: executive summary: a report of the American College of Cardiology/American Heart Association task force on clinical practice guidelines. Circulation. 2018;138(17):e426–e483. doi:10.1161/CIR.0000000000000597 5. James PA, Oparil S, Carter BL, et al. 2014 evidence-based guideline for the management of high blood pressure in adults: report from the panel members appointed to the Eighth Joint National Committee (JNC 8). JAMA. 2014;311(5):507–520. doi:10.1001/jama.2013.284427 6. Kjeldsen SE, Mariampillai JE, Nilsson PM. Optimal blood pressure target in diabetic and nondiabetic hypertensive patients. Circ Res. 2018;123(5):528–530. doi:10.1161/CIRCRESAHA.118.312763 7. DeCarolis DD, Gravely A, Olney CM, Ishani A. Impact of antihypertensive drug class on outcomes in SPRINT. Hypertension. 2022;79(5):1112–1121. doi:10.1161/HYPERTENSIONAHA.121.18369 8. Wright JT Jr., Williamson JD, Whelton PK, et al.; SPRINT Research Group. A randomized trial of intensive versus standard blood-pressure control. N Engl J Med. 2015;373(22):2103–2116. d