REVIEW
A Tale of 3 Trials: ACCORD, SPRINT, and SPS3. What
Happened?
Lawrence R. Krakoff1
Within the last several years, the National Institutes of Health has sup-
ported three randomized clinical trials to determine whether lower
than usually recommended goals for treatment of hypertension would
have greater benefit for prevention of cardiovascular disease and stroke.
These were the ACCCORD, SPRINT, and Secondary Prevention of Small
Subcortical Strokes (SPS3) Trials. Together they enrolled 17,114 par-
ticipants. Results for all three have been reported. The trials differ from
each other in their inclusion criteria, target blood pressures for the lower
goal (intensive treatment), but are similar in many respects. The results
with regard to their primary outcome were different: not significant
In epidemiologic surveys of untreated adult populations,
the level of arterial pressure is linearly related to risk of fatal
stroke and cardiac disease from 120 mm Hg systolic pressure
upward.1 Clinical trials have, over the past several decades,
demonstrated the value of antihypertensive drug treatment
for reducing event rates for both fatal and nonfatal car-
diovascular disease. In general, these trials recruited those
with definite Stage I  or II hypertension, often with base-
line systolic pressures ≥160/100 mm Hg and goal pressures
<140/90 mm Hg. However, trials comparing lower on-treat-
ment goals have been reported since 1995 as summarized in
a recent comprehensive meta-analysis. In combining older
and newer trials, the overall average baseline blood pres-
sures were 157/91 mm Hg and on-treatment pressures were
141/80 mm Hg for the higher or usual goals and 132/77 mm
Hg for the lower or intensive treatment goals.2
The results of 3 large, well-designed clinical trials, sup-
ported by the National Institutes of Health, have become
available to provide substantial evidence that bears on the
issue of proper goals for antihypertensive drug therapy:
these are the ACCORD,3 SPRINT,4 and the Secondary
Prevention Trial for Small Subcortical Stroke (SPS3) trials.
The usual goal blood pressures for ACCORD and SPRINT
trials were <140/90 mm Hg and the lower on-treatment
goals were <120/80 mm Hg. For SPS3, the usual goal was a
systolic pressure range of 139–130 mm Hg and for intensive
treatment, <130 mm Hg.
SPRINT, ACCORD, and SPS3 are similar in many
respects with regard to trial methodology. All used currently
accepted antihypertensive medications as they are used in
Correspondence: Lawrence R. Krakoff (Lawrence.krakoff@mountsinai.org).
Initially submitted January 20, 2016; date of first revision February 3,
2016; accepted for publication June 3, 2016; online publication June 29,
2016.
1020  American Journal of Hypertension  29(9)  September 2016
for ACCORD and SPS3, but definitely significant for SPRINT. Subgroup
analysis revealed differences and similarities. When viewed together and
with recent large observational studies, they support a conclusion that
Downloaded from https://academic.oup.com/ajh/article-abstract/29/9/1020/2622270 by guest on 12 July 2020
a systolic pressure in the range of 125–135 mm Hg range is likely to be
optimal on treatment for most hypertensive patients.
Keywords: blood pressure; clinical trials; diabetes; hypertension; inten-
sive treatment; low goals; outcomes; stroke.
doi:10.1093/ajh/hpw065
clinical practice. None had a placebo arm. ACCORD and
SPRINT were conducted in the United States. SPS3 was
conducted predominantly in the United States (65%), with
lesser fractions in South American (23%) and Spain (12%).
It is likely that the results of these 3 trials will have a major
impact on guidelines for management to be developed over
the next several years. This review is intended to provide a
useful concise comparison of these trials and recent related
publications for a balanced perspective of their relevance.
DESCRIPTION OF THE TRIALS
The major baseline characteristics for participants in
ACCORD, SPRINT, and SPS3 are given in Table 1. SPRINT
enrolled hypertensive participants, most of who were
already treated, and included those with high-normal pres-
sures (130–139 mm Hg systolic pressure), but excluded those
with diabetes or prior stroke. ACCORD, in contrast, only
included diabetics with hypertension. SPS3 enrolled hyper-
tensives with a history of stroke within 18 months of entry
with or without diabetes. Age ranges, sex distribution, per-
cent Black (self-designated African-American), and other
traits were similar, except that the ACCORD population had
higher body mass index and epidermal growth factor recep-
tor compared to the others and SPS3 had a higher fraction
of current smokers compared to ACCORD and SPRINT.
All 3 trials randomized participants into 2 groups for goals
of antihypertensive treatment: a usual and similar goal of
<140 mm Hg systolic pressure and a low goal of <120 mm
Hg for ACCORD and SPRINT. SPS3’s usual goal pressure
1Department of Medicine/Cardiology, Icahn School of Medicine at Mount
Sinai, New York, New York, USA. 
© American Journal of Hypertension, Ltd 2016. All rights reserved.
For Permissions, please email: journals.permissions@oup.com
 A Tale of 3 Trials

was 139–130 mm Hg systolic pressure and the low goal was
<130 mm Hg.
Table 2 displays the baseline pressures and the 1-year on-
treatment pressures. These are generally similar to the final
pressures, but the trials differ in their time course, so that the
1-year pressures provide a more consistent basis for compar-
ison. Baseline systolic pressures were 2–4 mm Hg higher for
SPS3 compared to ACCORD and SPRINT. The 1-year pres-
sures for usual treatment were slightly lower for ACCORD,
compared to the other 2 trials, and the 1-year pressures for
the lower goals were similarly near 120 mm Hg for ACCORD
and SPRINT, but 5–6 mm Hg higher for SPS3. However, a
highly significant difference between the low goal and usual
groups was found: 14.3 mm Hg lower for ACCORD (95%
confidence interval: −13.7 to −14.7), 14.8 mm Hg lower
for SPRINT and 11 mm Hg for SPS3. Thus, the Intensive
treatment for all 3 studies was highly effective for achiev-
ing a substantial reduction in systolic pressure for low goal
cohorts compared to the usual goal groups.
Table 3 displays the primary event rates for the 3 trials.
The absolute annual rates for usual goals and low goals are
shown and the differences between them. The hazard ratios
with confidence intervals and P values are also given. The
SPRINT trial achieved a highly significant reduction in its

Downloaded from https://academic.oup.com/ajh/article-abstract/29/9/1020/2622270 by guest on 12 July 2020

Table 1  Baseline characteristics for all participants for ACCORD, SPRINT, and SPS3

ACCORD SPRINT SPS3

Inclusion Diabetes Hypertension without diabetes or stroke Poststroke

Numbera 4,733 9,361 3,020

Age, years 62.2 ± 6.9 67.9 ± 9.4 63 ± 10.7
% Men 52.3 64.4 63.0
% Black 24.1 31.5 15.0
% Hispanic 7 10 16
BP medications NA 1.83 ± 1.04 1.7
BMI, kg/m2 32.1 ± 5.6 29.9 ± 5.8 29.1
eGFR 91.6 ± 28.8 71.8 ± 20.6 80 ± 18.5
Smokers current 13.2 13.3 20.5
Smokers past 41.9 42.6 NA
Fasting cholesterol, mg/dl 190 ± 40 190 ± 40 NA
Statin use 51% 44% NA
Aspirin use 56% 51% NA
Serum creatinine, mg/dl 0.9 ± 0.2 1.06 ± 0.34 NA
CVD history, %b 33.7 20.1 10.5

All trials enrolled hypertensive participants. Results expressed as % or mean ± SD. Abbreviations: BMI, body mass index; BP, blood pressure;
CVD, cardiovascular disease; eGFR, epidermal growth factor receptor; NA, not available; SPS3, Secondary Prevention of Small Subcortical
Strokes.
aTotal for all 3 trials = 17,114.
bCoronary artery disease + other CVD.

Table 2  Baseline and on-treatment blood pressures

ACCORD ACCORD SPRINT SPRINT SPS3 SPS3

Trial group Usual Low Usual Low Usual Low

Number 1,536 1,551 4,683 4,678 1,309 1,301

Baseline systolic 139 ± 16 139 ± 16 140 (139–141) 140 (139–141) 144 ± 19 142 ± 19

1-year systolic 133.5 (133.1–133.8) 119.3 (118.9–119.7) 136 (135–137) 121 (120–122) 138 (137–139) 127 (126–128)

Baseline diastolic 75.7 ± 10.2 75.8 ± 10.1 78.0 ± 12.0 78.2 ± 11.9 79 ± 11 78 ± 10

1-year diastolic 70.5 (70.2–70.8) 64.4 (64.1–64.7) 76.3a 68.7a 75.3 ± 7.9 69.4 ± 7.5

Results presented as mean with either (95% confidence intervals) or mean ± SD. Abbreviation: SPS3, Secondary Prevention of Small
Subcortical Strokes.
aEstimated from ref. 4, Supplementary Appendix.

American Journal of Hypertension  29(9)  September 2016  1021

Krakoff

primary outcome rate for the low goal treatment compared
to usual care and was discontinued after 3.7 years because
the trend between usual goal and low goal was so clear. At
this interval, the absolute difference in the combined pri-
mary outcome between the 2 groups was 1.6% supporting
a number needed to treat of 63.5 A  separate and planned
analysis of those ≥75  years old reported a highly signifi-
cant reduction of primary events and mortality (4.6%) for
the low goal group, compared to usual care over 3.1 years
of follow-up.6 The number needed to treat for this benefit
was 22. For both ACCORD and SPS3, the low goal treat-
ment failed to achieve a statistically significant reduction
in primary outcome rates. With regard to ACCORD, spe-
cifically, the actual event rate was substantially lower than
fibrillation and diminution of left atrial enlargement by
electrocardiogram.8
COMMENT
In all 3 trials, intensive antihypertensive treatment reduced
pressure substantially so that the issue of a possible J-curve,
i.e., increased outcome rates for the lowest pressures achieved,
is a possibility. This issue has been explored only for SPS3
with the calculation of an optimal on-treatment pressure of
124/64 mm Hg with higher outcome (stroke) rates above and
below this nadir.9 It would be helpful to analyze ACCORD
and SPRINT to determine an optimal on-treatment for
patients similar to those enrolled in these trials.

Downloaded from https://academic.oup.com/ajh/article-abstract/29/9/1020/2622270 by guest on 12 July 2020

predicted for the usual goal cohort reducing the power of
the prediction for benefit. In the past, stroke rates have
been considered the most reliable basis for the effect of
antihypertensive treatment. It is somewhat of a surprise
that in subgroup analysis, SPRINT failed to significantly
reduce stroke hazard ratio 0.89 (95% confidence interval:
−0.63 to +1.25, P  =  0.5) in primary prevention and that
SPS3 failed to significantly reduce stroke rate in secondary
prevention. The nonsignificant trends for better outcomes
with low goal treatment for ACCORD and SPS3 need not
be ignored in making individual decisions for patients who
tolerate more intensive antihypertensive treatment. Recent
analyses of the ACCORD trial indicate that the low goal
cohort of that trial had favorable outcome rates, compared
to the usual group for (i) regression of electrocardiogram
left ventricular enlargement7 and (ii) prevention of atrial
Do the results of these 3 trials inform decisions about anti-
hypertensive treatment of individual patients? In part this
issue depends upon whether or not the patient is similar to
a subgroup within the trials. For ACCORD, no differences
in trends were apparent for nearly all subgroup comparisons.
However, women had more frequent outcomes compared to
men (P < 0.01) and Whites had a trend (P < 0.09) to better
outcomes compared to non-Whites. For SPRINT, the lower
goal treatment for prevention of total mortality was supe-
rior for: those over 75 years old at entry, men compared to
women, and non-African-Americans compared to African-
Americians (ref. 4, Supplementary Appendix). SPRINT
included a substantial fraction of frail participants both
above and below age 7510 and frailty did not seem to alter
the beneficial outcome.6 Increased risk of serious falls has
been linked to intensive antihypertensive treatment of frail

Table 3  Primary outcome and results for ACCORD, SPRINT, and SPS3

ACCORD SPRINT SPS3

Composite: nonfatal MI, nonfatal Composite: AMI, ACS, All stroke: ischemic
Primary outcome stroke, CVD death stroke, AHF, CVD death and hemorrhagic

Absolute rate for usual treatment (% per year) 2.09 2.19 2.77
Absolute rate for intensive treatment (% per year) 1.87 1.65 2.25
Difference for absolute rate: intensive—usual (% per year) −0.22 −0.54 −0.52
Hazard ratio for low goal/usual goal 0.88 0.75 0.81
95% CI for hazard ratio 0.73–1.06 0.64–0.89 0.64–1.03
P value 0.20 <0.001 0.08

Abbreviations: ACS, acute coronary syndrome; AHF, acute heart failure; AMI, acute myocardial infarction; CI, confidence interval; CVD, car-
diovascular disease; MI, myocardial infarction; SPS3, Secondary Prevention of Small Subcortical Strokes.

Table 4  Recent observational studies that have calculated optimal on-treatment blood pressures

Author Number of subjects Type Population Optimal on-treatment pressures (mm Hg)

Sundstrom et al.14 34,009 Observational Established diabetes 135–139/74–76

Vamos et al.15 126,092 Observational New diabetics No CVD 139–140/80–84
CVD 139–140/80–84
Cederholm et al.16 53,583 Observational Established diabetes 130–135/75–79
Xie et al.17 44,989 Meta-analysis Mixed 133/76 for intensive treatment

Abbreviation: CVD, cardiovascular disease.

1022  American Journal of Hypertensio  29(9)  September 2016


elderly hypertensives confined to nursing homes.11 However,
older ambulatory patients with minimal or absent evidence
of frailty might benefit from careful intensive treatment of
their hypertension.12 Thus, for those without diabetes, opti-
mal benefit for intensive treatment would be expected for
men ≥75 years old, without previous chronic kidney disease
or cardiovascular disease, a relatively healthy older group.
The prediction that benefit for intensive treatment would be
equally distributed in all risk groups, based on meta-analy-
sis,13 is not supported by SPRINT and ACCORD. Subgroup
analysis has not yet been reported for SPS3.
The optimal on-treatment goal for hypertension in those
with diabetes may well be in the range of 130–140 mm Hg
systolic pressure as described in a large recent epidemiology
surveys14–16 and a meta-analysis comparing intensive blood
pressure lowering with less intensive treatment.17 Table  4
summarizes these reports with regard to the optimal on-
treatment blood pressure that might guide decisions. The
optimal on-treatment blood pressure for those with coro-
nary heart disease is also uncertain, but may be in the range
of 130–146/80–90 mm Hg, especially for older patients18 and
those with extensive vascular disease.19
Lowering risk of cardiovascular mortality and morbidity
will be, for the foreseeable future, dependent on antihyper-
tensive drug therapy. Should the goal of this therapy be to
reduce blood pressure to that of nonhypertensive groups at
the lowest risk level, i.e., <120 mm Hg or to accept on-treat-
ment pressures at higher, but effective levels? Should these
goals be tailored to easily definable subgroups (i.e., diabetes,
coronary artery disease, elderly, chronic renal disease, post-
stroke) when the best evidence supports that strategy? The
current evidence indicates that <140 mm Hg, but >120 mm
Hg for systolic pressure is getting close to the optimal range,
in the absence of signs of overtreatment.
DISCLOSURE
The authors declared no conflict of interest.
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; 360:1903–1913.
2. Thomopoulos C, Parati G, Zanchetti A. Effects of blood pressure lower-
ing on outcome incidence in hypertension: 7. Effects of more vs. less
intensive blood pressure lowering and different achieved blood pres-
sure levels—updated overview and meta-analyses of randomized trials.
J Hypertens 2016; 34:613–622.
3. The ACCORD Study Group. Effects of intensive blood-pressure control
in type 2 diabetes mellitus. N Engl J Med 2010; 362:1575–1585.
4. The SPRINT Research Group. A randomized trial of intensive
versus standard blood-pressure control. N Engl J Med 2015;
373:2103–2116.
American Journal of Hypertension  29(9)  September 2016  1023
A Tale of 3 Trials
5. Cook RJ, Sackett DL. The number needed to treat: a clinically useful
measure of treatment effect. BMJ 1995; 310:452–454.
6. 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 Jr, 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; e-pub ahead of print 19 May 2016.
7. Soliman EZ, Byington RP, Bigger JT, Evans G, Okin PM, Goff DC Jr,
Chen H. Effect of intensive blood pressure lowering on left ventricular
hypertrophy in patients with diabetes mellitus: action to control car-
diovascular risk in diabetes blood pressure trial. Hypertension 2015;
66:1123–1129.
8. Chen LY, Bigger JT, Hickey KT, Chen H, Lopez-Jimenez C, Banerji
MA, Evans G, Fleg JL, Papademetriou V, Thomas A, Woo V, Seaquist
Downloaded from https://academic.oup.com/ajh/article-abstract/29/9/1020/2622270 by guest on 12 July 2020
ER, Soliman EZ. Effect of intensive blood pressure lowering on inci-
dent atrial fibrillation and P-wave indices in the ACCORD blood
pressure trial. Am J Hypertens 2015; e-pub ahead of print 16 October
2015.
9. Odden MC, McClure LA, Sawaya BP, White CL, Peralta CA, Field TS,
Hart RG, Benavente OR, Pergola PE. Achieved blood pressure and out-
comes in the secondary prevention of small subcortical strokes trial.
Hypertension 2016; 67:63–69.
10. Pajewski NM, Williamson JD, Applegate WB, Berlowitz DR, Bolin
LP, Chertow GM, Krousel-Wood MA, Lopez-Barrera N, Powell
JR, Roumie CL, Still C, Sink KM, Tang R, Wright CB, Supiano MA;
SPRINT Study Research Group. Characterizing frailty status in the
systolic blood pressure intervention trial. J Gerontol A Biol Sci Med Sci
2016; 71:649–655.
11. Tinetti ME, Han L, McAvay GJ. Risks and benefits of antihyper-
tensive medications in older adults–reply. JAMA Intern Med 2014;
174:1873–1874.
12. Odden MC, Peralta CA, Haan MN, Covinsky KE. Rethinking the
association of high blood pressure with mortality in elderly adults: the
impact of frailty. Arch Intern Med 2012; 172:1162–1168.
13. Blood Pressure Lowering Treatment Trialists Collaboration, Sundstrom
J, Arima H, Woodward M, Jackson R, Karmali K, Lloyd-Jones D,
Baigent C, Emberson J, Rahimi K, MacMahon S, Patel A, Perkovic V,
Turnbull F, Neal B. Blood pressure-lowering treatment based on cardio-
vascular risk: a meta-analysis of individual patient data. Lancet 2014;
384:591–598.
14. Sundstrom J, Sheikhi R, Ostgren CJ, Svennblad B, Bodegard J, Nilsson
PM, Johansson G. Blood pressure levels and risk of cardiovascular
events and mortality in type-2 diabetes: cohort study of 34009 primary
care patients. J Hypertens 2013; 31:1603–1610.
15. Vamos EP, Harris M, Millett C, Pape UJ, Khunti K, Curcin V, Molokhia
M, Majeed A. Association of systolic and diastolic blood pressure and
all cause mortality in people with newly diagnosed type 2 diabetes: ret-
rospective cohort study. Br Med J 2012; 345:e5567.
16. Cederholm J, Gudbjornsdottir S, Eliasson B, Zethelius B, Eeg-Olofsson
K, Nilsson PM. Blood pressure and risk of cardiovascular diseases in
type 2 diabetes: further findings from the Swedish National Diabetes
Register (NDR-BP II). J Hypertens 2012; 30:2020–2030.
17. 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.
18. Denardo SJ, Gong Y, Nichols WW, Messerli FH, Bavry AA, Cooper-
Dehoff RM, Handberg EM, Champion A, Pepine CJ. Blood pressure
and outcomes in very old hypertensive coronary artery disease patients:
an INVEST substudy. Am J Med 2010; 123:719–726.
19. Dorresteijn JA, van der Graaf Y, Spiering W, Grobbee DE, Bots ML,
Visseren FL; Secondary Manifestations of Arterial Disease Study Group.
Relation between blood pressure and vascular events and mortality in
patients with manifest vascular disease: J-curve revisited. Hypertension
2012; 59:14–21.