by Malcom Kendrick, MD
It was not until the 1970s that anyone actually set out to answer the question. Does lowering blood pressure in people with moderately raised blood pressure do any good? Recruiting for this study started in 1973. Seven hundred thousand people were contacted, and half a million people accepted an invitation to participate.
As is the way with such things, for various reasons, the enormous initial number was whittled down to just under eighteen thousand people with diastolic blood pressures between 90 - 109, and systolic pressure below 200. This, the UK MRC study, reported its results in 1985. I remember it well, as I was at a cardiovascular conference at the time.
And what were the results? Well, if you get down to the most important result of all, which is overall mortality, there were 248 deaths in the treated group and 253 in the placebo group (2).
Or to put this another way. Nine thousand people were treated for raised blood pressure for five years. At the end of the study five more people were alive in the treatment arm than the placebo arm. A result so deeply unimpressive that it fitted comfortably within the possibility of it being purely a chance finding. Or, to put it another way, this study failed completely to reach the holy grail of medical studies - statistical significance.
It seemed from this very large, long-term study, that lowering mild/moderately raised blood pressure was of no benefit. Certainly not when you set it against billions of dollars it costs, and years of potential side-effects.
I do remember thinking at the time. Well, that should cause a massive re-think in the whole area. But it did not. Not even slightly. This result was basically swept aside and ignored.
"Men occasionally stumble over the truth, but most of them pick themselves up and hurry off as if nothing ever happened." Sir Winston Churchill.
In fact, once the dust settled, what actually happened was that the goalposts were pulled out of the ground, and moved. Whilst it was clear that there was no impact on heart disease, or increased life expectancy. There was a small, but statistically significant, effect on stroke [one stroke delayed for around nine hundred years of treatment - whoopee].
Other studies had shown the same thing, which seems reasonable. If you think that a stroke is often due to an artery in the brain bursting (cerebral hemorrhage), then the idea that lowering the blood pressure may reduce the risk of arterial bursting seems plausible.
Stroke is a disease that fits within the grouping of cardiovascular disease - which sounds as though it has something to do with the heart, but does not - necessarily. So, instead of looking at overall mortality, or heart disease, high blood pressure drugs are now measured on their impact on ‘cardiovascular disease'. And when such studies are reported, they are normally, also, reported as composite end-point outcomes. Which means adding together every cardiovascular outcome together. This is a well-known method of creating statistical significance when none actually exists.
Over the years, a veil has been drawn over the fact that reducing blood pressure has no effect on heart disease, or life expectancy. When it comes to this area, the discussion has moved on to cardiovascular mortality (which basically means stroke), and the trials all report on ‘composite end-points' often including the progression of kidney disease - which may, or may not be, important.
In a parallel move, everyone began to move away from the actual data and began to use a model, called the log-linear model. This model (which is a mathematical way of smoothing data), has meant that people now stand up and say ‘the relation of systolic blood pressure to risk of death is continuous, graded, and strong, and there is no evidence of a threshold.' (3).
This, in turn, has allowed the development of guidelines, which now prove that the further you can drive down blood pressure, the better. And you can do all of this without the need for data. Everyone uses the log-linear model instead, it is the holy grail of research in this area.
Just one slight problem. Twenty years ago Ancel Keys - the man who created the diet-heart hypothesis - concluded that the linear model was wrong. Nine years ago, a group of medical statisticians [hold on, here are two parties with whom I usually disagree violently] anyway, a group of medical statisticians re-analyzed the Framingham data, which underpinned the log-linear model and they concluded that: ‘Shockingly, we have found that the Framingham data in no way supported the current paradigm to which they gave birth. In fact.... The paradigm MUST be false (4).'
They went on (in the European Heart Journal), to make the following statement:
‘No randomized trial has ever demonstrated any reduction of risk either overall, or cardiovascular death by reducing systolic blood pressure to below 140mmHg.'
The effect of their analysis was, as you may expect, a deafening silence. This, despite the fact that these statisticians had just stated that everything that everybody believed about lowering blood pressure was wrong.
Now, nine years later, we have another analysis showing that: ‘A new review has found that lowering blood pressure below the "standard" target of 140/90 mm Hg is not beneficial in terms of reducing mortality or morbidity1.' July 2009
During this nine year period the thresholds for ‘treating' blood pressure have got lower and lower. For diabetics, hypertension is now defined as 115 systolic. This definition was, or course, developed from combined end-point cardiovascular data, and the log-linear model. The one that has been proved to be false.
I will now make two predictions. The level of blood pressure deemed high, and in need of treatment, will continue to fall. In nine years, or thereabouts, another study will appear proving that there is no evidence that lowering blood pressure in moderate hypertension has any benefits. My final prediction is that the evidence will continue to be ignored. Because, so far, all evidence has been ignored, so I see no reason why this will change.
The simple fact is that when blood pressure is raised, it is raised for a reason. The reason is an underlying ‘disease'. Lowering the blood pressure will certainly get rid of an annoyingly high measurement, but it cannot (unless by complete coincidence), have any impact on the underlying disease...... the one that actually causes the damage. So, unless it is startlingly high, what good can lowering blood pressure actually do?
Does Treating High Blood Pressure Do Any Good? Part 1 of 2
2: MRC trial of treatment of mild hypertension: principal results. Medical Research Council Working Party: BMJ 1985;291(6488):97
3.Stamler J. Blood pressure and high blood pressure, aspects of risk. Hypertension 1991; 18(Suppl I): I-95 - I - 107
4: Port S, Garfinkel A, Boyle N: There is a non-linear relationship between mortality and blood pressure. EHJ (2000) 21 p 1635-1638
Dr. Malcolm Kendrick (MbChB MRCGP) M.D.
Dr. Kendrick has worked in family practice for almost twenty years.
He has specialized in heart disease and set up the on-line educational website for the European Society of Cardiology.