Higher than average blood pressure is likely associated with higher mortality, but is there any study that showed that lowering it with medication is beneficial?
Yes, mortality benefits for blood pressure medicine have been demonstrated in trials.
Let me just preface this by saying this was much harder to find than I was expecting. Questioning the benefit of blood pressure reduction is medical heresy, so you'd think you wouldn't have trouble finding the data out there.
First of all, let's define the question. We're asking whether treatment of hypertension leads to lower mortality in the setting of a randomized clinical trial. Reading between the lines, I think what we really want to know is whether treatment of isolated hypertension (i.e. in patients that have no strokes, heart failure or other diseases associated with hypertension) leads to lower mortality in a randomized clinical trial. Treating hypertension after a stroke or heart attack is hands down beneficial (see HOPE, PART2, IDNT, NICOLE or PREVENT trials [1-5]). You can't answer the question with this data though, because maybe the drug is really just treating the heart attack or stroke.
To get data specifically on treating hypertension itself, not in the setting of other medical problems, you have to go back to the 1960s. The VA COOP Study Group on Antihypertensive Agents [6,7] trial specifically looked at treating people who just came into the clinic with high blood pressure. Mortality was 5% lower in the treatment group, or, for every 20 people treated for 3.3 years (the average time people were enrolled in the study), 1 person will have their life saved. Honestly, this is a pretty good outcome as far as drugs go, taking aspirin to prevent heart attacks doesn't work nearly as well, for example. The authors collected these numbers on mortality but they didn't test for whether the numbers were likely to have just popped up by chance or not (statistical significance). I crunched them myself with Fisher's exact test and the results were unlikely to have just come up by chance (p value = 0.015)
Some caveats. This study was old (the word negro is used), but it was really well done. They had the patients go through a 2 month run in phase where they had to take pills that turned their urine orange just so they could see whether they took their pills regularly before letting them in the trial. All the patients and the doctors were blinded. They used sealed envelope randomization. Patients were enrolled from eight different sites. Of course, it was done at the VA in the 60s, so every single patient was a man. Also, the patients weren't exactly free of other diseases. For some reason, the authors didn't just say how many patients had strokes or heart attacks in the past. They devised this "severity score" to assess how many health problems people had at the start of the trial. The score went from 0-4 and on average the patients were less than 1. I would say most weren't very sick.
Some other trials tried to test blood pressure medicines versus placebo but fell short. The Australian Therapeutic Trial in Mild Hypertension  had many fewer events than the VA study so weren't quite able to show statistical significance. The benefits to treatment that they measured in this study were way smaller. Treatment reduced death by .15%. So for every 666 people taking the drug for 1 year, 1 person's life would be saved. They were only able to show this was statistically significant when they looked at the numbers for people actually taking the drug. You want to look at everyone that entered the study in the first place though (intention to treat), because you can always invent scenarios where you get biased results if you don't do this.
There was one other study that looked at this question. The Oslo study  also failed to show that treatment actually saved lives when treating patients with just hypertension.
Keep in mind that all of these studies were able to show benefit to treating (fewer strokes, less kidney failure) but mortality was really only lower in the VA trial. My gut tells me that this was because aged American veterans were less healthy to start with then relatively healthy Norwegians and Australians (the population from the other studies). It was less of a needle-in-a-haystack challenge in the VA trial.
Sometime in the 70s or 80s, it seems that doctors all decided that treating hypertension was the way to go no matter how healthy the patient was otherwise so we don't have any more studies.
- HOPE (Heart Outcomes Prevention Evaluation) Study Investigators. Effects of an angiotensin-converting-enzyme inhibitor, ramipril, on cardiovascular events in high-risk patients. N Engl J Med 2000; 342: 145–53.
- MacMahon S, Sharpe N, Gamble G, et al. Randomised, placebocontrolled trial of the angiotensin converting enzyme inhibitor, ramipril, in patients with coronary or other occlusive vascular disease. J Am Coll Cardiol 2000; 36: 438–43.
- Lewis E, Hunsicker L, Clarke W, et al. Renoprotective effect of the angiotensin-receptor antagonist irbesartan in patients with nephropathy due to type 2 diabetes. N Engl J Med 2001; 345: 851–60.
- Dens J, Desmet W, Coussement P, et al. Usefulness of nisoldipine for prevention of restenosis after percutaneous transluminal coronary angioplasty (results of the NICOLE study). Am J Cardiol 2001; 87: 28–33.
- Pitt B, Byington R, Furberg C, et al. Effect of amlodipine on the progression of atherosclerosis and the occurrence of clinical events. Circulation 2000; 102: 1503–10.
- Effects of Treatment on Morbidity in Hypertension: Results in Patients With Diastolic Blood Pressures Averaging 115 Through 129 mm Hg. JAMA. 1967;202(11):1028-1034. doi:10.1001/jama.1967.03130240070013.
- Effects Morbidity of Treatment on in Hypertension: II. Results in Patients With Diastolic Blood Pressure Averaging 90 Through 114 mm Hg. JAMA. 1970;213(7):1143-1152. doi:10.1001/jama.1970.03170330025003.
- THE AUSTRALIAN THERAPEUTIC TRIAL IN MILD HYPERTENSION: Report by the Management Committee, The Lancet, Volume 315, Issue 8181, 14 June 1980, Pages 1261-1267, ISSN 0140-6736, http://dx.doi.org/10.1016/S0140-6736(80)91730-4. (http://www.sciencedirect.com/science/article/pii/S0140673680917304)
- Anders Helgeland, Treatment of mild hypertension: A five year controlled drug trial: The Oslo study, The American Journal of Medicine, Volume 69, Issue 5, November 1980, Pages 725-732, ISSN 0002-9343, doi: 10.1016/0002-9343(80)90438-6.
This is one of the few areas of primary prevention* where the data are clear.
The (intelligently) skeptical tone of your question suggests to me that you would be (appropriately) wary of drawing conclusions based on observational data or surrogate endpoints. Fortunately, you have asked a question about an area where rigorous data are available showing reductions in the risk of cardiovascular disease and mortality on the basis of randomized, controlled trials.
In large-scale randomized trials of people with primary hypertension, antihypertensive therapy produces a nearly 50 percent relative risk reduction in the incidence of heart failure, a 30 to 40 percent relative risk reduction in stroke, and a 20 to 25 percent relative risk reduction in myocardial infarction.1,2,3,4
The benefits show a consistent "dose-response" relationship. That is, larger improvements in blood pressure control are associated with greater decreases in risk. This is an important point, since it adds credibility to the association. This is demonstrated in graphs like this one:
Image from Reference 1, below: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2386598/
On the x-axis you see the degree of blood pressure lowering achieved with medications. On the y-axis is “relative risk” (RR). By definition, a null intervention yields RR=1. A relative risk of 0.5 represents a 50% decreased risk, etc.
The graph is showing meta-analysis data, i.e. data compiled from many clinical trials in order to increase statistical power. The basic idea is that each circle is a clinical trial and bigger circles represent “stronger” data (i.e. with lower variance). The regression line shows that there is a linear relationship between the degree of blood pressure lowering and the relative risk reduction (here for a composite endpoint of stroke, myocardial infarction, and heart failure.) This analysis included 31 randomized, placebo-controlled trials, with 190,606 participants. These are strong data.
In addition to the dramatic reductions in adverse cardiovascular outcomes, blood pressure control has also been shown to reduce mortality. One meta-analysis used data from 42 randomized, controlled studies including nearly 200,000 subjects (Psaty). They found a reduction in cardiovascular disease mortality (RR, 0.81; 95% CI, 0.73-0.92); and total mortality (RR, 0.90; 95% CI, 0.84-0.96). The fact that these relative risk (RR) confidence intervals do not cross 1 demonstrates statistical significance.
Although the RR value of 0.90 is considerably less impressive than the reductions in more specific outcomes (stroke, heart failure, etc), this is expected due to the myriad of other factors affecting mortality. A statistically significant relative risk of 0.90 for mortality is actually quite dramatic. One would be challenged to find any other intervention for primary prevention that, in randomized trials, can be shown to decrease overall mortality with this degree of certainty.
There are many interventions in modern medicine that are of questionable long-term benefit to healthy patients (i.e. primary prevention).** These include cholesterol lowering medications, aspirin, various forms of cancer screening, etc. In most cases, the disease-specific improvements in outcomes are subtle and debated, and randomized data showing a reduction in overall mortality are lacking or inconsistent. The use of blood pressure lowering medications in patients with hypertension falls into a different category. These medications are effective.
*That is, interventions aimed at preventing disease in healthy people. This is in contrast to secondary prevention, treating people after they have already had an adverse outcome. In general, secondary prevention is a much "easier" arena in which to demonstrate benefit because the risk of adverse outcomes is so much higher.
**Here, I’m considering hypertension, hyperlipidemia to fall within the range of “healthy” because these abnormalities are only problematic if they cause a cardiovascular event of some sort.
Blood Pressure Lowering Treatment Trialists' Collaboration, Turnbull F, Neal B, Ninomiya T, Algert C, Arima H, Barzi F, Bulpitt C, Chalmers J, Fagard R, Gleason A, Heritier S, Li N, Perkovic V, Woodward M, MacMahon S. Effects of different regimens to lower blood pressure on major cardiovascular events in older and younger adults: meta-analysis of randomised trials. BMJ. 2008 May 17;336(7653):1121-3.
Law MR, Morris, Wald NJ. Use of blood pressure lowering drugs in the prevention of cardiovascular disease: meta-analysis of 147 randomised trials in the context of expectations from prospective epidemiological studies. BMJ 2009; 338.
Kostis JB, Davis BR, Cutler J, Grimm RH Jr, Berge KG, Cohen JD, Lacy CR, Perry HM Jr, Blaufox MD, Wassertheil-Smoller S, Black HR, Schron E, Berkson DM, Curb JD, Smith WM, McDonald R, Applegate WB. Prevention of heart failure by antihypertensive drug treatment in older persons with isolated systolic hypertension. SHEP Cooperative Research Group. JAMA. 1997 Jul 16;278(3):212-6.
Gueyffier F, Boutitie F, Boissel JP, Pocock S, Coope J, Cutler J, Ekbom T, Fagard R, Friedman L, Perry M, Prineas R, Schron E. Effect of antihypertensive drug treatment on cardiovascular outcomes in women and men. A meta-analysis of individual patient data from randomized, controlled trials. The INDANA Investigators. Ann Intern Med. 1997 May 15;126(10):761-7.
Psaty BM, Lumley T, Furberg CD, Schellenbaum G, Pahor M, Alderman MH, Weiss NS. Health outcomes associated with various antihypertensive therapies used as first-line agents: a network meta-analysis. JAMA. 2003 May 21;289(19):2534-44.