hypertension goal in diabetics

a recent systematic review and meta-analyses looked at a common problem: optimal blood pressure levels in diabetics (see See dm htn goal bmj2016 in dropbox, or BMJ 2016;352:i717). this issue is clouded by the variability of results of several large studies and the variability of guideline recommendations by different societies and their changes over the past few years  (eg: Am Diabetic Assoc now  suggests goal of <140/90, though <130/80 "may be appropriate" if young, has albuminuria, or multiple cardiovasc risk factors; JNC8 suggests a goal of <140/90). these recommendations were based on published data; the current one systematic review includes previously unpublished data (eg, getting more granular data on the subset of patients with diabetes in larger trials, by contacting drug companies, authors, etc). this review also assessed both baseline blood pressure and achieved blood pressure on meds, and included only reasonably large studies (>100 patients with diabetes) treated at least 12 months, and included studies comparing meds vs placebo, 2 meds vs 1, or different BP targets, and assessed total and cardiovascular mortality and an array of comorbidities. details:

--mean duration of follow-up was 3.7 years, majority had type 2 diabetes, and there was a wide range of comorbidities

--49 trials, 73738 patients (25 trials had 26,625 who were diabetic subgroups of larger trials; 24 trials had 47,133 exclusively diabetic patients). unpublished data included 8916 patients from 12 studies. their table 1 has a brief review of the individual studies, noting that 36 trials had patients with baseline cardiovasc disease,  5 had none and 8 without baseline data)

--results:

    --added medications on average reduced the baseline blood pressure by 10.2mmHg    

    --baseline blood pressure: if SBP (systolic BP)>150mmHg

        --all cause mortality decreased 11 % [RR 0.89 (.080-0.99)]

        --cardiovascular mortality decreased 25% [RR 0.75 (.057-0.99)]

        --MI decreased 26% [RR 0.74 (.063-0.87)]

        --end-stage renal disease decreased 18 % [RR 0.82 (.071-0.94)]

        --stroke decreased 23% [RR 0.77 (0.65-0.91)]

    --baseline blood pressure: if SBP 140-150mmHg

        --all cause mortality decreased 13% [RR 0.87 (.078-0.98)]

        --heart failure decreased 20% [RR 0.80 (.066-0.79)]

        --MI decreased 16% [RR 0.84 (.076-0.93)]

    --baseline blood pressure: if SBP <140mmHg

        --cardiovascular mortality increased 15% [RR 1.15 (1.10-1.32)], with tendency to increased all-cause mortality [RR 1.05 (0.95-1.16)]

        --overall: for each 10mmHg decrease in SBP in those with lower baseline SBP: 15% increased cardiovascular mortality [RR 1.15 (1.03-1.29)] and 13% increased MI [RR 1.13 (1.03-1.22)]

    --similar results for the attained blood pressure (approx 10mmHg less than the baseline SBP)

        --ie, increased mortality if attained SBP <130, risk reduction if >130mmHg

        --but, there was still a decreased risk of stroke with attained SBP<130mmHg of 35% [RR 0.65 (0.42-0.99)]​ --though this had the largest confidence intervals, reflecting low numbers of events

    --risk of bias: felt to be low, and removing the higher risk studies, there was a shift to even more harm with more aggressive treatment; similarly, removing the ALTITUDE study which had double renin-angiotensin inhibition (aliskiren plus another renin-angiotensin ihnibitor), now felt to be deleterious, similarly led to no change in the calculated risk.

    ​--also, no reduction in albuminuria (early surrogate end-point for renal failure) if baseline SBP <140 mmHg

    ​--outcome data looking at diastolic blood pressure largely mirrored those of systolic, with decreased clinical outcomes if baseline DBP >90mmHg; still improvement but less so with DBP 80-90mmHg; but no difference (vs worsening with SBP) if DBP<80mmHg; similarly with attained DBP of >80 mmHg, 75-80 mmHg, or <75 mmHg (though their metaregression analysis did find that cardiovascular mortality increased 28 percentage points for each 10mmHg decrease in baseline DBP, with lines crossing from benefit to harm at 78mmHg)

so, to me this studies raises several important issues:

--blood pressure goals of patients with diabetes and on medications should probably be in the 140/80 range, which is different from those found in the SPRINT study in nondiabetics (see blog http://gmodestmedblogs.blogspot.com/2015/11/tighter-blood-pressure-control-sprint.html  for a critique of SPRINT). this raises the very real possibility, that contrary to prior guidelines, the goal blood pressure in diabetics should perhaps be higher than non-diabetics (prior guidelines had set a lower diabetic BP goal. the newest JNC8 recommends the same goal for diabetics as nondiabetics). a potential biological rationale for this might be: diabetics have more extensive atherosclerotic disease, so lowering the systolic blood pressure that we measure in the arms may actually lead to marked reductions in flow in the vessels in the vital organs and increased risk of ischemia (data do show those with coronary artery stenoses have lower fractional flow reserve with decreasing diastolic pressures and resultant myocardial hypoperfusion; and, arterial stiffening, frequent in diabetics, makes myocardial perfusion more dependent on  systolic pressures). this all supports a U- or J-shaped curve, with an optimal blood pressure above or below which is associated with worse outcomes.

--this study and data on goals of BP in diabetics apply to patients already on antihypertensive meds. the situation for those not on meds is unclear (ie, when do we start meds?, what is the real role of lifestyle changes?). it is reasonable to assume that those not on meds may well have different outcomes than those on meds, since they are likely to have had hypertension for shorter periods of time (so the vascular remodeling issues are likely to be less), and the meds themselves have significant adverse outcomes. for example, there was a really really old study which looked at hypertensive patients and the sodium-lithium countertransport across red cells, finding that in patients with similar lowering the blood pressure by meds (really really old meds) vs lifestyle, only lifestyle led to reversal of Na-Li countertransport​ (it was then felt that Na-Li countertransport was a good marker of longterm renal dysfunction. which it may or may not be. but the point is that weight loss, diet and exercise may have different physiologic effects than meds, even independent of the recorded adverse effects). so, getting the human body back to its ancient normal state (eating lots of wholesome foods instead of calorie-rich, nutrient-depleted prepared foods, getting lots of exercise, not being obese) might allow the body to avoid complications of hypertension, for example, better than adding on drugs....

--so, all things being equal, i do think that the goal of 140/75-80 seems to be the most reasonable one for diabetic patients at this time. but i still really push for lifestyle changes, in an attempt to avoid meds and perhaps improve clinical outcomes.