Sources
All information has been taken from an upcoming video series on NutritionFacts.org. All citations and source material will be available once the videos are published.
Evidence for Elevated LDL Causing CHD
While there has been plenty of research on the elevated LDL cholesterol being the primary cause for CHD, one of the most interesting pieces of evidence is to look at a subset of the population with a particular mutation of a gene called PCSK9. About 1 in 50 African-Americans have this mutation, and as a result they have about a 40% lower LDL cholesterol on average than African-Americans who do not have the mutation. The group with the mutation studied had plenty of other non-LDL related CHD risk factors:
Most were overweight or obese
50% had hypertension
33% were smokers
20% had diabetes
Yet, in spite of these risk factors, the group with the PSK9 mutation enjoyed an 88% reduction in the risk of developing heart disease as compared to their peers.
Who Should Be Taking a Statin?
If you have a history of CVD, MI, or stroke, you should be on a statin without question. If you don't yet have CVD, you should calculate your personal 10 year risk, either by using online resources or working with a medical professional. Here are some online resources to calculate your risk:
American College of Cardiology: bit.ly/ACCrisk (Greger recommended)
Framingham Risk Profiler: bit.ly/FRArisk
Reynolds Risk Score: bit.ly/REYrisk
Risk Enhancing Factors to consider for Borderline and Intermediate scores:
Family history of CVD
High level of LDL cholesterol
Presence of metabolic syndrome
Chronic kidney disease
Chronic inflammatory conditions
History of premature menopause
History of preeclampsia
High risk race/ethnicity
Persistently elevated triglycerides
Elevated C-reactive protein
Elevated Lipoprotein(a)
Elevated apoB (>130 mg/dL)
Low ankle-brachial index (<0.9)
Relative vs Absolute Risk and Numbers Needed to Treat
Prescribing statins is very different from treating someone who is injured or acutely sick. The doctor starts to resemble an insurance salesman in that, for an ongoing cost deferred benefits hope to be realized. In this scenario, understanding risk and probability is critical for proper decision making.
Suppose a hypothetical disease has a 2% chance of killing you. A drug is available that cuts your risk from dying from that disease by 50%. This fifty percent number represents your relative risk reduction. However, considering the disease only has a 2% chance of killing you, cutting that chance in half means your absolute risk was lowered by 1% (from 2% to 1%).
Another way to look at these numbers is looking at how many people would need to take the drug to prevent an extra death from the disease. In this case, if 100 people took the drug, 1 person would avoid dying from the disease. This is known as Numbers Needed to Treat (NNT).
The ability to shift between relative and absolute risk can be used both to manipulate perceived benefits and downplay side effects. Let’s take a real world example of statins.
A New England Journal of Medicine study found that over a 5 year period those randomly chosen to take a statin (versus those randomly chosen to be on a placebo) had a reduction of relative risk of an MI of 24%, but the absolute risk reduction is just 3%. One of the study’s findings was to look at the incidents of breast cancer.
To see how mixing the use of statistics can be deceptive, a company marketing this statin could say something like,
“Pravastatin has been shown to reduce the risk of heart attacks by 24 percent, while only increasing the risk of breast cancer by 3 percent.”
On the flip side, the anti-Pravastatin spin would say something like,
“Pravastatin increases the relative risk of breast cancer by 1,267%, while only decreasing the absolute risk of a heart attack by 3%.”
Note: Subsequent studies found no evidence of a link between statins and breast cancer, so this example is used only to highlight how statistics can be communicated in a manipulative way.
This is why Numbers Needed to Treat (NNT) is so useful. It eliminates the confusion and misperception of relative risk vs absolute risk. In the next section, we will see how NNT helps to cut through the spin and confusion of different kinds of risk.
How Effective Are Statins?
With regards to the effectiveness of statins in preventing an MI, the balance of evidence seems to be centering on a 25% reduction in relative risk. The table below is based upon this 25% reduction in relative risk (these numbers are accurate). Notice, however, how much clearer the use of NNT is in determining how effective the use of a statin is in reducing the risk of an MI:
As you can see, when the risk for an MI increases, the number of patients (out of 100) needing to take a statin to prevent the occurence of an MI decreases.
There seems to be a contradiction between this relative risk reduction of 25% and the risk reduction of those having the PSK9 genetic mutation (mentioned earlier). Even though both groups (those taking the statin and those with the mutation) achieve about the same 40 point LDL drop, those with the mutation have an 88% risk reduction. Why the difference? Those with the mutation have enjoyed this reduction their entire lives, whereas those on statins haven’t taken the drug nearly that long. It simply comes down to the duration of having a lower LDL.
Preventive Pill Paradox
When surveyed, most people said the benefit of statins needed to be a 30% reduction in MI absolute risk in order for them to agree to take it. However, the actual absolute reduction of MI for daily use varies from about 1% to 7.5% depending on your 10 year risk score. This presents physicians with a difficult situation: How to balance the patients’ right to know the truth against the expected reduction in compliance when patients act upon that knowledge? What about those that could have avoided an MI with compliance? This is known as the “preventive-pill paradox.” Informing patients about absolute risk reductions would increase the very outcomes that the medicine was designed to prevent.
Downsides of Statins
Statins Muscle Pain Side Effect
The most common reported side effect from a statin is muscle pain. About 20% of statin users report this, but researchers wanted to know if statins were truly to blame. The ASCOT-LLA trial was designed to find out. It was a double-blind study that consisted of 10,000 participants randomized to a statin or placebo for years at a time (because of the concern that the muscle pain takes time to manifest or ameliorate). The findings showed no excess number of reports of muscle pain for those assigned to statins over those assigned to placebos. At the end of the trial, only when patients learned which group they were in, did those on statins retrospectively report an increase in muscle pain. This seems to be a case of “nocebo” effects. Instead of 1 in 5 suffering from muscle pain, the ASCOT-LLA found the actual rate is more like 1 in a 1,000.
The ASCOT-LLA study was terminated early because of an increased rate of death among those in the placebo group, and this precisely highlights the risk of going off statins when they are indicated for proper use. For every 83 people that discontinue statins, there is 1 excess death within a 4 year period. The authors of the ASCOT-LLA study write:
“Seldom in the history of modern therapeutics have the substantial proven benefits of a treatment been compromised to such an extent by serious misrepresentations of the evidence for its safety.”
A meta-analysis of studies looking into muscle pain and statins found something similar: 1 in 2,000 instances of muscle pain being directly caused by statin use.
How can there be such a difference in what is seen in clinical practice (1 in 5 reporting muscle pain) versus what is shown in clinical studies (between 1 in 1000 and 1 and 2000)? One clue might be to look at the funding of the clinical trials. ASCOT-LLA was partially funded by Pfizer. Likewise, the meta-analysis was funded by drug manufacturers. So is this a case of mass delusion of statin users, or mass suppression of side effects by researchers? We simply don’t know, because the science hasn’t been done on this topic.
Diabetic Risks of Statins
While industry funded studies call into question the frequency of associated muscle pain, one side effect that even these studies agree on is an increased risk of causing type 2 diabetes. Even though it took 23 years to prove this cause and effect, the underlying medical mechanism isn’t fully known. The leading theory is that statins simultaneously inhibit the pancreas’ secretion of insulin, and increases overall insulin resistance.
Some studies have shown that even short term statin use doubles the odds for developing diabetes and diabetic complications. This increased risk persisted for 5 years even after statin use stopped. The chart below shows the increase rate of diabetes diagnosis for those taking a statin:
This risk of diabetes doesn’t seem to be well known, even among providers, and even less among patients. Research by the American Heart Association (Clough, Martin, Navar, et al) show that only 28% of providers believed that stains can cause diabetes and only 17% report discussing this risk with their patients.
Below downsides to statin use:
Daily use for long periods of time
Expense (although almost all statins are now generic)
Common side effects: nausea, diarrhea, constipation
Muscle Aching/stiffness (maybe 5 in 100?)
Liver Damage (2 in 100)
Developing type 2 diabetes (1 in 500 )
Muscle and Kidney damage (1 in 20,000)
Mayo Clinic Statin Decision Calculator
This is a very use tool that will show you a graphic representation of your current risk and how a statin would change that risk:
How Much Longer Can You Expect to Live on a Statin?
Several studies have tried to answer this question. Estimates have ranged from 3-4 days to 10 years. These types of studies are completely dependent on the length of the study and the remaining life expectancy of the subjects. A 2016 analysis of prior studies tried to account for these variables and extrapolate life extension assuming the patient took the drug for significant periods of time and had decades of life expectancy left. Here’s what they found:
For every 39 points (1 mmol/L) of LDL reduction, an additional 2.7 - 5.6 years of additional life.
Diet vs Statins
The impressive effect of dropping LDL cholesterol on life expectancy isn’t tied to the method you use to achieve it. Suppose you are not in the high risk category for a 10 year MI, and you decide to reduce your risk through diet alone. How long would it take to achieve the same 39 point LDL drop with a strict WFPB diet? Two weeks of a diet packed with fruits, vegetables, and nuts. A 2001 study (Jenkins, Kendall, Vidgen, Mehling, et al) was done to see the changes in LDL cholesterol when subjects were fed a very high fiber diet of fruits, vegetables, and nuts.
Surprise, surprise. Using a lifestyle intervention is far and away the best treatment for lifestyle diseases. Drugs treat symptoms, not the cause--only lifestyle modification does that.
We have already covered the risks and side effects of statins, but what of those of a plant based diet? Using a WFPB diet to lower LDL has the following side effects:
Weight loss
Reduces hypertension
Reduces systemic inflammation
Improves insulin sensitivity
Reduces oxidative stress
Improves endothelial function
Reduces the risk of thrombosis
Improves or reverses several autoimmune diseases
Remember how there was a difference in CHD risk between the PSK9 gene mutation group vs the statin user group, and this came down to how long you were living with a low LDL? This is another reason why adopting a healthy lifestyle, particularly one emphasizing a WFPB diet, and doing this earlier in life can show much better risk reduction than simply taking statins when the need is determined. How much better? Healthy lifestyle choices may produce an 81-94% reduction in risk, whereas statin therapy alone achieves a reduction of only about 25%.
What Should Be Your Target LDL?
How late you start the journey to reduce LDL and any prior history of CHD determine how low your target LDL should be to minimize any future risk.
Benefits From Lowering Beyond Targets and Can Cholesterol Be Too Low?
Having an LDL cholesterol target of less than 30 (mg/dL) as a result of experiencing a prior MI, is a difficult goal to achieve. Getting LDL this low used to be only possible with both drug therapy and a very strict diet (like the one Dr. Caldwell Esselstyn gives his cardiac patients). Recently, however, numbers like this can be achieved solely with a new class of drug called PSK9 inhibitors (and, yes, these came about as drug makers studied those with the PSK9 genetic mutation). This has allowed the question of whether cholesterol can be too low to be studied.
A study was published in JAMA Cardiology in August 2018 (jamacardio.2018.2258) that studied this very question. Its goal was to study both whether additional benefits were realized as LDL drops lower and lower, and how safe it is to have a very low LDL. The study had over 15,000 patients. One of the findings from the study was that there is a consistent risk reduction in major vascular events as LDL-C decreases starting at 63 (mg/dL) and going as low as 21 mg/dL). In other words, as your LDL goes lower and lower, your chance of an MI goes lower and lower.
Furthermore, during the course of the study, those participants having extremely low LDL had no observed adverse effects. As a result of seeing continued benefits as LDL decreases, and no observed side effects of very low LDL, the authors recommend that lowering the current LDL-C guidelines even further would reduce cardiovascular risk.
Other evidence that extremely low LDL is neither unnatural nor harmful can be found in looking at two population groups: newborns and those with the PSK9 genetic mutation. In newborns, the average LDL varies between 22 and 45. LDL levels of around 30 are common among those with the PSK9 mutation, and as a whole, that group is known to have an exceptionally long life expectancy compared to those without the genetic mutation.
Where does the notion that cholesterol lowering too far can be dangerous come from? Science has known for some time just how vital cholesterol is for life. One of its most important functions is controlling the permeability of every cell in your body in that it helps reduce just any molecule being able to penetrate your body’s cells. Cholesterol is also essential in the production of many hormones. The fear is that as cholesterol levels drop, these essential functions could start to compromise with as yet unknown manifestations. Currently, this fear is unsupported by evidence and does not account for the body’s ability to achieve homeostasis by regulating cellular cholesterol content.
In fact, in the studies done on those taking PSK9 inhibitors, and even among those with LDL under 15, no evidence of impairment of adrenal, ovarian, or testicular hormone synthesis has been found (these are the hormones most commonly associated with cholesterol). Impairments of these functions would be expected should the supply of necessary cholesterol run out. Again, it appears as though the body is compensating to achieve homeostasis. Why isn’t there a consensus on this topic yet? It comes down to the duration of observing any possible effects. The longest such study is about 6 years currently.
While there isn’t yet scientific consensus on adverse effects of very low cholesterol, there is little debate on its beneficial effects in regards to CVD, the #1 killer of Americans. While it appears as though the extremely low levels of LDL itself isn’t dangerous, it is a totally different question about the means of achieving those levels. The safety of PSK9 inhibitors has been studied only for a few years, and as you may recall from earlier, the risk of developing Type-2 diabetes as a result of taking statins wasn’t established for the first 23 years they were on the market. Long term risks of PSK9 inhibitors are as yet unknown. Lastly, PSK9 inhibitors cost about $14K a year.
Dietary Questions Regarding Cholesterol
What if I follow a WFPB diet, but cholesterol remains above target numbers?
First examine your risk and situation. Are you trying to prevent developing CVD or stop from having a second MI, for example? It is very likely that you will need the help of drugs to get to LDL levels of below 50. Even if this is the case, you should focus on the quality of your diet. First ensure you are truly WFPB. For example, cutting out as many oils as possible, especially palm, palm kernel, and coconut oils (oils aren’t WFPB). Coconut oil raises LDL more effectively than lard. Obviously, cut out animal based foods as those add cholesterol to your diet as well. Once you have checked these boxes, you aren’t actively adding cholesterol to your diet, so now it is time to focus on removing cholesterol from your body.
So, assuming you are truly WFPB, you might have genetics working against you. There is a genetic bell curve, where PSK9 mutations are on one end, and you might be closer to the other. This simply means you have to work harder by focusing on a subset of WFPB foods specifically designed to pull cholesterol from your body. Check out the portfolio diet. The link below is not WFPB but you can modify it to be. These are foods that actively pull out cholesterol from the body: https://www.stmichaelshospital.com/media/hospital_news/2018/0706.php
If you really want to geek out on watching a cause and effect based on changes in your diet, you can get a blood cholesterol meter. It works just like the common blood glucometer for measuring blood sugars. You could try adding or removing foods from your diet, and as long as you eat the same way for two weeks, you should be able to get a meaningful, at home measurement of any changes.
At least until recently, blood cholesterol meters would only measure total cholesterol, although some now will give you Total, HDL, and Triglycerides. A total number can only give you an approximation of what LDL would be. Generally, 150 total cholesterol is about equal to 70 LDL. If you get a meter that gives you total, HDL, and triglycerides, you can calculate LDL by the following formula:
LDL = (Total - HDL - Triglycerides) / 5
How Much Do Genes vs Lifestyle Play Into CVD Risk?
Excluding risks associated with lipoprotein (a), lifestyle modifications erase 90% of the risk, while 10% is strictly genetic. This doesn’t mean that if you have poor genetics, you have no control. Genetics can load the gun, but it is almost always lifestyle that pulls the trigger. The higher your genetic risk, the harder you will need to work to maintain a clean WFPB diet.
Why, Even with WFPB, Do My Triglycerides (TGs) Remain Elevated?
Typically, TGs are elevated by excess sugar intake. Your liver only has a certain rate at which it can process sugars (especially fructose and sucrose). If this rate is exceeded (for example drinking OJ instead of eating an orange), one of the mechanisms your liver has to deal with this is to convert the sugar to fat and export it into the bloodstream as TGs. On a proper WFPB diet, this shouldn’t happen because the fiber in naturally sweet foods slows down the rate of sugar absorption to a level the liver can handle.
So assuming you are truly WFPB and not consuming excess sugar, some people are sensitive to powdered grains (even whole grains). See the chapter in Dr. Greger’s How Not To Diet about “Deflouring Your Diet.” Even though a powdered whole grain is WFPB, it undergoes a mechanical process that destroys the structure of the whole grain which can affect your body’s absorption of these foods. Try cutting out floured grains and eat only intact grains (oat and barley groats, millet, etc).