Questioning Statin Dogma in the Boiler Room

In my previous post, I wrote about a sailor and the overreliance on proxies that ultimately contributed to the ship’s demise — and perhaps our own. Of course, much more is happening on a ship than what’s visible on the surface; countless parts and people work synergistically to keep it afloat. In undergrad, I learned about the fluid dynamics of water in its solid, liquid, and gas forms. When water transitions from liquid to gas, its molecules move apart, creating distance and occupying more space. Ships have boiler rooms that harness this principle. They heat water well above its boiling point, and naturally, this superheated water wants to turn to gas (steam) and expand. The boiler system manages this by keeping the superheated water under immense pressure. Despite all the energy trying to separate the molecules and expand them into steam, this pressure forces them to remain densely packed. This controlled condition allows for regulated steam production, which is then directed to power various parts of the ship. I wonder if my professor would be proud of me recalling an example from their class, or maybe embarrassed at my attempt to describe boiler mechanics.

In the clinic recently, a question was posed to me: “But you still believe in them, right?” I paused, unsure how to respond. The question sparked two immediate thoughts: first, its leading nature, positioning positive belief as the default, the status quo. Second, the jarring use of the word ‘belief’ in a conversation that had, until that point, revolved around statistical findings. I was discussing statins and the LDL hypothesis with a colleague. With the USPSTF guidelines having been recently covered in our teaching, the topic was ripe for discussion. These guidelines ostensibly form a foundation of evidence-based medicine, providing recommendations on interventions and qualifying the strength of evidence for them.

To be fair, I already harbored a baseline weariness regarding LDL guidelines. I’m aware of the financial incentives of pharmaceutical companies for widespread chronic statin use. This awareness stems from a piece of solid advice I’ve received throughout life: always ask ‘Who benefits?’ and remember that if something seems too good to be true, it probably is. Baseline bias aside, however, my colleague and I reviewed the pool of studies used by the USPSTF for its statin recommendations. A prominent example, which I believe typifies much of the research in this area, is the JUPITER trial. In it, LDL was reduced from a median of ~110 mg/dL to ~55 mg/dL; treatment with rosuvastatin effectively halved the LDL. A 50% reduction in a target metric would, one might presume, lead to equally large reductions in cardiovascular outcomes. The JUPITER authors reported that “rosuvastatin significantly reduced the incidence of major cardiovascular events.”

However, looking at the absolute risk reductions (ARR) from their published study for various outcomes, a different picture emerges:

• Cardiovascular mortality: ARR 0.09% (Number Needed to Treat [NNT]=1113)

• Non-fatal Myocardial Infarction (MI): ARR 0.45% (NNT=223)

• Any MI: ARR 0.42% (NNT=241)

• Non-fatal stroke: ARR 0.31% (NNT=318)

• Any stroke: ARR 0.35% (NNT=288)

• Arterial revascularization: ARR 0.71% (NNT=142)

• Hospitalization for unstable angina: ARR 0.12% (NNT=810)

  
  

These NNTs mean that to prevent just one case of death from cardiovascular causes over the study’s duration, 1,113 people needed to take the medication. Many trials reviewed by the USPSTF show similarly small absolute risk reductions. Yet, given the ubiquity of this medication class and its strong guideline endorsement, conversations questioning widespread statin use — despite these small ARRs and large NNTs — are often treated as heretical.

It makes a certain kind of systemic sense, though. Statin therapy and LDL monitoring are drilled into students in lecture halls. This knowledge is then assessed in high-stakes cardiovascular block exams and reinforced during clinical rotations. Imagine a stressed, anxiety-filled class of high achievers preparing for an upcoming test. Hundreds of students, in this institution alone, are taught the LDL hypothesis, its validity underscored by “evidence-based guidelines.” Multiply this by the thousands of healthcare education programs worldwide, and an enormous status quo is forged: statins are beneficial, and lowering LDL unequivocally improves cardiovascular outcomes. A voice suggesting otherwise is a mere drop in a vast ocean. It’s akin to the ‘doublethink’ described in Orwell’s 1984: despite evidence that prompts critical questions about the medication’s true efficacy and the wisdom of chasing the LDL metric, an unwavering belief in the status quo persists.

In my “Proxy Problems” post, I never detailed the proximate cause of the ship’s demise. I illustrated the misguided attempt to declutter an actively sinking ship but didn’t specify why it was sinking. Back in the boiler room was the engineer, who saw the pressure gauge climbing into the red, exceeding capacity. A critical pipe began to creak, then crack. She acted quickly, sealing the heavy boiler room doors, hoping to contain the escalating problem. Outside those sealed doors, a deceptive calm returned. No more ominous sounds, no visible signs of distress. Inside, however, pressure continued to build. Suddenly, the pipe split open. The superheated water, once contained by immense pressure, was suddenly unleashed. All the pent-up energy in those superheated molecules, desperate to expand, seized the opportunity. And expand they did. Explosively. This instantaneous, violent expansion created an explosion the size of a football field, splitting the ship’s hull in two.

By now, the comparison I’m drawing is likely clear. Is the metaphor hyperbolic? Perhaps. Yet, it points to an equally unsettling truth in our approach to health. Using a metric like LDL to assure ourselves of ‘control’ over cardiovascular risk can be like that engineer shutting the door on the escalating boiler room crisis. An LDL of 55 mg/dL is celebrated by the patient, the clinician, and the clinic alike. It’s a metric we all use to applaud our efforts, offering a sense of security, control, and comfort. Yet, many large trials show us that the absolute risk difference for individuals is often minimal. This frequently means treating hundreds of people with a medication for only one person to achieve a specific, positive outcome, while the others experience the costs and potential side effects without that particular benefit.

My point is not that statins are inherently ‘bad,’ nor that I am ‘anti-big pharma.’ It’s that relying heavily on LDL as the primary gauge of cardiovascular risk can be a profound misdirection, a proxy that masks underlying problems. These root issues may continue unabated while the patient, the clinician, and the clinic all receive positive reinforcement for achieving a target number — a number that, as the data suggests, may offer less security than we collectively believe.