Nobody doubts that lowering blood pressure in people in whom it is elevated, reduces mortality and cardiovascular events. Yet the Systolic Blood Pressure Intervention Trial (SPRINT) study results  were a surprise to many, where people with high cardiovascular risk benefitted from treatment that achieved stricter SBP targets than are recommended in some guidelines. For a brief period, the results were taken at face value but other reactions soon followed including questions on how the findings applied to the routine clinical setting .
SPRINT was a well designed and well executed randomized controlled trial. The results are consistent with data from accumulated sources . There is no question that the results are applicable to the particular population selected for the study, treated and monitored according to the study protocol. The results were internally coherent and there was little if any dishomogeneity within the results. The question is whether there was anything particular about the patient selection, protocol or study methodology that limits the generalizability of the findings to other clinical scenarios.
Much debate has focused on the way blood pressure was measured in the SPRINT study as the protocol required unobserved automated blood pressure recordings. It has been suggested that this produces blood pressure readings that are 5–15 mmHg lower than standard clinical measurements recorded in the presence of a health professional , that is, strict control of blood pressure may not have been so strict at all.
The opposing argument to direct application of the SPRINT findings to the broader population with hypertension has largely been drawn from post hoc analysis of studies performed for other purposes.
Okin et al.  have taken this approach in the present issue. On the basis of a post hoc analysis of the Losartan Intervention for Endpoint Reduction (LIFE) study it is proposed that SPRINT achieved successful outcomes for cardiovascular and total mortality because the pre-entry blood pressure in the group as a whole was relatively low. They found that people in the LIFE study below the 25th percentile of pretreatment SBP (≤164 mmHg) who achieved on-treatment blood pressure within the middle tertile (≤52 mmHg) or below also had lower overall mortality than those whose blood pressure reduction remained above 152 mmHg. However, people with higher pretreatment SBP did not benefit from achieving a lower on-treatment blood pressure than 152 mmHg compared with those who did not. Putting simply a stricter target was only beneficial in people with lower blood pressure to begin with.
Is it valid though to analyse LIFE data to help explain SPRINT? Are there other possible explanations for this finding? The results do appear to contrast with a more recent study, the Heart Outcomes Prevention Evaluation-3 in a different patient population again where only those at the higher end of the spread of blood pressure in the population seemed to benefit from blood pressure-lowering drugs .
Perhaps a better place to look to understand the results of SPRINT is within the SPRINT dataset. A post hoc analysis of SPRINT showed that although low baseline DBP was associated with increased (not decreased) risk of cardiovascular events, this was not seen with baseline SBP. Furthermore, there was no evidence that the benefit of the intensive lowering of SBP differed by the baseline level of blood pressure .
The LIFE protocol was published 20 years ago and the results 5 years later . LIFE participants all had left ventricular hypertrophy determined by electrocardiographic criteria. SPRINT participants were people without diabetes at high risk because of age, concomitant disease and/or high Framingham risk score. Only 7.4% had left ventricular hypertrophy defined by Cornell voltage criteria as used in the LIFE study .
Pretreatment blood pressures were measured after 2 weeks off-treatment in the LIFE study whereas more than 90% of people in the SPRINT study had their pretreatment measurements performed on-treatment. Not surprisingly the initial blood pressure levels were much lower in SPRINT than in LIFE. We do not know what pretreatment blood pressure was in the majority of SPRINT participants and whether they were comparable to those in LIFE. However, blood pressure was measured differently. Semi-automated recorders were used in both studies but the SPRINT protocol famously required health professionals to leave the room at the time of measurement.
The antihypertensive medications used in the studies were different. LIFE patients received losartan or atenolol. The latter no longer features as first-line antihypertensive therapy in many international guidelines. As a more contemporary study designed to look at blood pressure targets irrespective of treatments, SPRINT patients received different therapy both to lower blood pressure and for associated conditions. Half were taking aspirin and 43% were on statins.
Unfortunately, other patient characteristics were not evenly balanced between tertiles of baseline blood pressure in the post hoc LIFE study analysis. The group that seemed to benefit from a lower on-treatment blood pressure with baseline SBP 164 mmHg or less were younger, more likely to be men and/or black and had better lipid, metabolic and renal profiles. They also had less left ventricular hypertrophy using electrocardiographic scoring. All of these things are known to be associated with outcomes, either positively or negatively. The authors have gone to great lengths to adjust for these confounding variables using propensity scores and other statistical adjustment but the case would be stronger if the groups were more closely matched.
Although some SPRINT participants had left ventricular hypertrophy, ECG criteria was an entry requirement for LIFE. High voltages on an electrocardiogram can be because of concentric left ventricular hypertrophy. It has been shown in experimental studies that this can compromise sub-endocardial perfusion at lower blood pressures, particularly, if there is coronary artery disease, effectively shifting the DBP-perfusion autoregulation curve to the right . Alternatively, high voltages can be because of cardiac dilatation and incipient or actual cardiac dysfunction. This is a high-risk group in which low blood pressure is often seen. Thus, the entry criteria for the LIFE study may explain some of the mortality seen whenever blood pressure was lowered from high before treatment to low on-treatment according to the criteria used in this analysis.
Whilst the interaction between baseline blood pressure and on-treatment blood pressure was significant for total mortality, this did not hold for cardiovascular end points. It is difficult to see how differences in blood pressure would be manifest in noncardiovascular mortality. The question arises as to whether preexisting disease was more common in the patients who achieved lower blood pressure on treatment.
In conclusion, the post hoc analysis of the LIFE study by Okin et al. has generated the hypothesis that SPRINT findings of better outcomes with a lower SBP treatment target may be a result of the particular patient group selected for the study and their relatively low pretreatment blood pressure albeit mostly on antihypertensive therapy. This hypothesis draws a long bow and can only be tested by repeating SPRINT in different populations. In the meantime, LIFE stands alone and SPRINT stands alone, given their many differences in patient population, blood pressure measurement techniques and medications. Much has been learned from both, and 20 years after there still seems to be life after LIFE! No doubt, SPRINT has much more to give us in future too.
Supported by a Program Grant from the National Health & Medical Research Council of Australia.
Conflicts of interest
There are no conflicts of interest.
1. The SPRINT Research Group. A randomized trial of intensive versus standard blood-pressure control. N Engl J Med
2. Kjeldsen SE, Hedner T, Narkiewicz K. HOPE-3, SPRINT, VALUE and a meta-analysis of trials in patients with diabetes support treatment of hypertension to a target below 140 mmHg. Blood Press
3. Verdecchia P, Angeli F, Gentile G, Reboldi G. More versus less intensive blood pressure-lowering strategy: cumulative evidence and trial sequential analysis. Hypertension
4. Mancia G, Kjeldsen SE. Adopting Systolic Pressure Intervention Trial (SPRINT)-like office blood pressure measurements in clinical practice. J Hypertens
5. Okin PM, Kjeldsen SE, Devereux RB. The relationship of all-cause mortality to average on-treatment systolic blood pressure is significantly related to baseline systolic blood pressure: implications for interpretation of the Systolic Blood Pressure Intervention Trial study. J Hypertens
6. Lonn EM, Bosch J, Lopez-Jaramillo P, Zhu J, Liu L, Pais P, et al. Blood-pressure lowering in intermediate-risk persons without cardiovascular disease. N Engl J Med
7. Beddhu S, Chertow GM, Cheung AK, Cushman WC, Rahman M, Greene T, et al. Influence of baseline diastolic blood pressure on effects of intensive compared to standard blood pressure control. Circulation
2017; [Epub ahead of print].
8. Dahlof B, Devereux RB, Kjeldsen SE, Julius S, Beevers G, de Faire U, et al. LIFE Study Group. Cardiovascular morbidity and mortality in the Losartan Intervention For Endpoint reduction in hypertension study (LIFE): a randomised trial against atenolol. Lancet
9. Soliman EZ, Ambrosius WT, Cushman WC, Zhang ZM, Bates JT, Neyra JA, et al. Effect of Intensive blood pressure lowering on left ventricular hypertrophy in patients with hypertension: SPRINT (Systolic Blood Pressure Intervention Trial). Circulation
10. Smolich JJ, Weissberg PL, Broughton A, Korner PI. Aortic pressure reduction redistributes transmural blood flow in dog left ventricle. Am J Physiol
1988; 254 (2 Pt 2):H361–H368.