Does BP Treatment Reduce Stroke Risk? (1)
I wasn’t planning to write about hypertension anytime soon, but a conversation I had with a specialist colleague a few weeks ago encouraged me to produce this article sooner. It is also the perfect time for you to read this, especially if you’ve already read my previous post on NNT (Numbers Needed to Treat). Let’s put that knowledge to use while it is still fresh in your mind.
As you might expect, working in the CBD while practicing medicine a little differently can be intriguing to some. The colleague I had the honor of sitting down with is a consultant-grade hospital physician. I surprised him with one bold statement: to date, we have no evidence that treating non-severe hypertension prevents stroke.
Before you throw your blood pressure medications away, read on. I am dissecting evidence-based medicine, which is largely an academic exercise. In contrast, the decisions you reach in discussion with your doctor are practical—the "real deal." Even though I am about to show you the studies that led us to where we are today, all or none of it may apply to your personal health. With that, let’s turn back the clock to 1964.
A cardiologist in Washington, called Edward Freis, looked at 452 severely hypertensive patients with recent stroke or TIA across 17 hospitals and showed that only 10 of them needed to be treated for hypertension for 2 years, before 1 person is saved from a stroke event. Soon after he conducted 2 other studies to prove that this treatment opportunity is not exclusive to stroke sufferers.
The VA Cooperative Study was a landmark—the first study of its kind to trigger the general "war on hypertension." Prior to 1967, high blood pressure was largely accepted as a normal part of aging. This study was conducted on 143 male inpatients who had multiple comorbidities and were evidently very sick; their diastolic BP was at least 115 mmHg. Guess what the 2 year NNT was for preventing stroke? An impressive 3.
Realising that his first cohort might not reflect a healthier population, Dr Freis ran a second study three years later. This time he enrolled 380 people. Although these patients were hospitalised and had multiple comorbidities, they weren't as acutely ill as the first group, with an average diastolic pressure around 105 mmHg. The 5 year stroke NNT was still remarkable at 13. Two years later, in 1972, blood pressure treatment became a staple in community-based preventative medicine. There was significant speculation that Dr Freis would be the next recipient of the Nobel Prize.
But US doctors were yet to test their hypothesis in the community. If sick, hospitalised patients benefit this much from antihypertensives, could we save even more people by treating the less-ill general population? In 1979, the HDFP study followed nearly 11,000 people in the community with hypertension (defined then as diastolic >95 mmHg, which is still quite high by today’s standards). The NNT to prevent one case of stroke in 5 years? 100. Following HDFP, the Australians conducted ANBP in 1980 and the Brits ran MRC in 1985, resulting in stroke NNTs of 125 and 143 respectively.
The British study was particularly relevant to me, as most British medical students are taught the "Inverse Care Law"—an idea promoted by Dr Julian Tudor Hart that helped reform the NHS. This influential Welsh GP was famous for his critique of our approach to treating blood pressure. He famously cited the one-year NNT for the MRC study as an "absurd" 850 (noting that the NNT improves after five years).
By contrast, the Number Needed to Harm (NNH) is significantly lower: for every 33 people treated for one year, one will experience side effects severe enough to discontinue the medication. Dr Hart warned that, aside from the one "lucky" patient among the 850, success is guaranteed only for the profiting pharmaceutical industry. Ironically, he advocated for high-intensity, personalised medicine—something most modern doctors no longer have the time to provide. This approach is arguably incompatible with a "Pharma-controlled" NHS. Realising this, Dr Hart co-founded the Keep Our NHS Public campaign in 2005.
In 1991, the STOP-H study in Sweden attempted to find a more convincing result by studying older participants (aged 70–84) with severe hypertension (systolic 180–230 and diastolic 105–120 mmHg). By including only high-risk subjects, they did achieve a better NNT at 2 years, but the number was still 67.
With those studies, pharmaceutical medicine essentially closed the case on hypertension for nearly two decades. Guidelines encouraged people to be treated for this "disease," and doctors told patients incessantly that they risked a stroke if their numbers stayed elevated. I suspect most of those doctors never cited the NNT. This was Dr Julian Hart’s frustration too.
Subsequent studies before 2008 focused on different outcomes:
MDRD (1994): Targeted renal damage (proteinuria).
RENAAL (2001): Focused on preventing end-stage renal failure in diabetic patients.
IDNT (2001): Compared two different agents to see which was better for the kidneys.
ALLHAT (2002): Compared medications for heart disease prevention.
AASK (2002): Examined which agents better protected the kidneys in a specific racial group.
PREVEND (2005): Focused on hypertension with renal damage (microalbuminuria).
These trials were gold mines of knowledge, but they either compared medications against each other or focused on specific cases like renal damage, end-stage renal failure etc. They weren't particularly relevant to the typical "Jack and Jill" walking around with high blood pressure but no other major health issues.
In 2008, the ACCORD study arrived from the US and Canada. It included only diabetic patients with high cardiovascular risk and compared intensive treatment (systolic <120 mmHg) to standard treatment (<140 mmHg). The medical world was shocked when the study was stopped early because more people were dying in the intensive treatment group; it was no longer ethical to continue.
I remember encountering this study shortly after graduating in 2011, but the emphasis was never on the serious issue of over-medicating. Instead, pharmaceutical interests found a twist—they claimed they had shown a reduction in stroke. In their view, aggressive treatment was worthwhile because it "saved" people from strokes. In reality, you can't save people if they are dying from over-treatment. In other words, the number of stroke diagnoses went down in the intensive treatment group as they died before they could get a stroke. Regardless, the 5-year NNT for stroke in ACCORD that they were hailing? 94.
Everything changed in 2015 with SPRINT. But curiously, I had been a doctor for four years prior to SPRINT, and the dogma had always been to treat all community hypertension until it was below 130 or 140 mmHg. Without SPRINT, did we really have a leg to stand on? The VA Cooperative studies cannot be used to paint a picture of everyone living outside a hospital. Most studies between 1992 and 2007 were for very specific conditions.
That leaves us with HDFP, ANBP, MRC and STOP-H—the last of which sampled only high-risk groups with severe hypertension. The ‘best’ of the remaining three, HDFP, tells us we need to treat 100 people for 5 years to prevent a single stroke. Yet, that is what we used until 2015 to justify treating non-severe, non-complicated hypertension—the most common type there is.
Edward Freis never received the Nobel Prize. Perhaps the Nobel Committee in Stockholm saw what you are reading here in this article today.