Achieving very low BP in patients with diabetes associated with higher CV risk
Cardiovascular outcomes and achieved blood pressure in patients with and without diabetes at high cardiovascular risk
Introduction and methods
Different associations have been described for systolic blood pressure (SBP) and different outcomes [1.2]. Conversely, patients with stable coronary artery disease (CAD)  or those at high CV risk  show an increased risk of CV death when on-treatment SBP or diastolic BP (DBP) drop below 120 mmHg and 70 mmHg, respectively. Diabetes or metabolic syndrome and hypertension frequently co-exist, which increases the risk for microvascular and macrovascular complications [5,6].
If guidelines recommend stricter BP levels, and more people achieve lower BP, this may put more patients at risk. Especially in patients with diabetes, BP targets remain a matter of debate. The ONgoing Telmisartan Alone and in Combination with Ramipril Global Endpoint Trial (ONTARGET) 
and the Telmisartan Randomised AssessmeNt Study in ACE intolerant subjects with cardiovascular Disease (TRANSCEND)  studies randomized patients with a history of MI, stroke, peripheral artery disease (PAD), or those with high CV risk to receive ramipril, telmisartan or both. 31546 patients were randomized, of whom 11730 had diabetes, and 19806 had not. The trials shared the same committees, inclusion and exclusion criteria (except regarding ACEi intolerance). After (median of) 56 months of treatment, no difference in outcomes was seen between treatment groups, which allows a direct comparison of groups based on diabetes status.
This secondary analysis of ONTARGET and TRANSCEND data aimed to assess the risk in patients with or without diabetes over the whole spectrum of achieved SBP in a broad range of patients with a history of CAD or PAD or transient ischemic attack or stroke or diabetes mellitus complicated by end-organ damage. Existing BP medication was adjusted according to clinical practice. Study medication was given on top of standard treatment. Pooling data allowed performing an adequately powered post hoc analysis of patients with (n=11.487) or without (n=19.450) diabetes according to mean achieved in-trial SBP and DBP. Attended BP was measured with an automated validated device after resting for 3 minutes in a sitting position, in the presence of the study nurse or investigator. The primary outcome was a composite of CV death, MI, stroke and hospital admission for HF.
- The distributions of SBP achieved on treatment were very similar for diabetes and no diabetes, but in those with diabetes, it was shifted to the right (mean in diabetes: 137.9 ± 13.6 mmHg, no diabetes: 133.5 ± 13.8 mmHg, P<0.0001). The distributions of DBP in the two groups were superimposable.
- Cumulative incidence of the primary composite endpoint, as well as of its components, was higher in those with, compared to those without diabetes.
- When compared to the reference group of individuals without diabetes and achieved SBP 120-140 mmHg and DBP 70-80 mmHg, those with in-trial SBP ≥140 mmHg or DBP ≥80 mmHg were at increased risk of the primary outcome, both when with diabetes (SBP 140-160 mmHg: HR: 1.67, 95%CI: 1.48-1.89, ≥160 mmHg: HR: 2.31, 95%CI: 1.93-2.76, DBP 80-90 mmHg: HR: 1.61, 95%CI: 1.40-1.85, and DBP ≥90 mmHg: HR: 2.32, 95%CI: 1.92-2.82) and without diabetes (SBP 140-160 mmHg: HR: 1.14, 95%CI: 1.03-1.26, ≥160 mmHg: HR: 1.66, 95%CI: 1.36-2.02, DBP 80-90 mmHg: HR: 1.16, 95%CI: 1.05-1.28, and DBP ≥90 mmHg: HR: 1.61, 95%CI: 1.352-1.93).
- Patients with diabetes and in-trial SBP <120 mmHg (HR: 1.53, 95%CI; 1.27-1.85) and all patients with DBP <70 mmHg (diabetes: HR: 1.77, 95%CI: 1.52-2.06 and no diabetes: HR: 1.30, 95%CI 1.16-1.46) showed increased risk.
- No interactions were observed between diabetes and SBP or DBP, indicating that the adverse effects of diabetes and high or low BP were additive.
- Similar observations were done for the components of the composite outcome.
- Restricted spline analyses with mean achieved SBP and DBP analysed as continuous variables confirmed that hazards for the primary outcome and its components were higher in those with, compared to those without diabetes. Most curves were J- or U-shapes with lowest risk at SBP 120-140 mmHg and DBP 70-80 mmHg.
This analysis of ONTARGET/TRANSCEND trial data showed that the association between mean achieved SBP and DBP and clinical outcomes is non-linear, in patients at high CV risk, of whom many were hypertensive and taking antihypertensive drugs. The lowest risk was seen at SBP 120-140 mmHg and DBP 70-80 mmHg. Higher outcome rates were seen in patients with diabetes, across the whole spectrum of achieved BP. Thus, lower BP levels are associated with increased risk in patients with diabetes and diabetes with end-organ damage who are already at high risk. These data support consideration of lower BP boundaries for patients with high CV risk, in particular those with diabetes.