Lipid Changes on a Very-Low-Carb Ketogenic Diet: My Own Experience

Moscarpone…

I'd like to preface this blog post by apologizing for its length, including links to several long articles. Also, for anyone who doesn't know me, I'm a vocal and enthusiastic supporter of low-carbohydrate diets, but I always strive to be balanced in my writing. I'm very nonconfrontational and don't like "getting into it" with people who disagree with me. However, I expect I'll receive plenty of negative feedback from this article because of the controversial topic.

Cholesterol Results From June 2013 through November 2013

My cholesterol levels have always been higher than average. LDL has ranged from 120s-150s as far back as I can remember, long before I began following a moderately carbohydrate-restricted diet back in 2011. In June of last year, I reported my NMR (Nuclear Magnetic Resonance) LipoProfile results after almost a year of consuming a very-low-carb ketogenic diet (VLCKD) containing less than 50 grams net carb per day. I was very happy with these values and frankly a little surprised that I achieved them while eating delicious, satiating foods. 

---

Lipid Profile from November 2013

In November of last year, I had a standard lipid profile done as part of lab work for my annual physical: 

Total Cholesterol: 300
LDL-C: 160
HDL-C: 128
TG: 56

My numbers had increased, but I wasn't terribly concerned about the LDL-C, since on a few occasions it had been nearly that high in the past. Seeing a total cholesterol of 300 was a bit troubling, but I knew it was partially due to having extremely high HDL (Apparently high levels of some types of HDL can also be problematic, although I didn't realize this at the time). Looking back, although I wasn't tracking my intake online regularly back then, I'm pretty sure I was eating the same or perhaps a little more fat than when I had the NMR done five months earlier.

Nutritional Ketosis Experiment

At the beginning of January, I decided to experiment with lowering my carb intake further in order to achieve nutritional ketosis. I didn't want or need to lose weight, but after speaking with a few people who'd reported improved mental focus and energy on minimal carbs and ketone levels between 1.5-3.0, I was intrigued.  For the record, I felt great prior to this experiment: no symptoms of adrenal fatigue, excellent blood sugar control, lots of energy, good sleep, etc. But was there a possibility I could feel even better in deep ketosis? I'm a curious type, so I decided to try it for a few months. I had a ketone meter but didn't test very often because the strips are ridiculously expensive. But when I did check prior to this experiment (first thing in the morning, the only time I've ever tested), my ketones ranged between 0.4-1.0 mm. 

I began tracking my intake on My Fitness Pal, as many of my clients were doing. I lowered my net carbs to roughly 20 grams per day, although total carbs were often still around 50 grams because I ate a lot of avocados, unsweetened cocoa powder, and high-fiber vegetables like cauliflower. However, my consumption of berries dropped from 1-1.5 cups per day to 5 or 6 every morning at breakfast. I tried to keep protein around 70-80 grams daily (I'm 5'8" and 125 lbs, so this isn't all that low), and I ate more fat in order to maintain rather than lose weight. I never drank bulletproof coffee or added lots of butter or coconut oil to my food. But I did eat a fair amount of cheese, cream cheese, ricotta, and moscarpone, and I began using heavy cream instead of half-and-half in my coffee and tea. I still ate vegetables at every meal, although smaller amounts.

I tested blood ketones a couple of times a week in the morning, and results ranged from 1.2-1.8. After 3 months of eating this way, in all honesty, I didn't feel any different. I still felt great, slept great, etc., but I can't say I had more energy or experienced any cognitive benefits.  My weight stayed the same, and my blood sugar control remained good. However, my lipids had definitely changed, and not for the better.

Cholesterol Results from April 2014

I had an NMR drawn at the end of April, and this time I'll admit to being more than a little upset when I saw the results:

I ordered this NMR through a different lab, so there are a few additional labs (mainly VLDL related) that weren't included in the one from June of 2013.

I was really surprised by how much my cholesterol had gone up since the prior test. My first thought was that perhaps my thyroid levels were off. (I have hypothyroidism that was diagnosed shortly before I went low carb, but my levels have been stable for the past few years on desiccated thyroid). However, I didn't feel at all hypothyroid and wasn't scheduled to have my thyroid labs re-checked until summer.

"Why Are You Concerned When You Have Such High HDL-C, Low Triglycerides, and Large, Fluffy LDL-C?"

While I've always been comfortable with higher than ideal cholesterol levels, having an LDL-C over 200 is a different story. The highest value I'd ever seen prior to last December was 158, I believe, about eight years or so ago when I was still following a low-fat, high-carbohydrate diet. But as far back as I can remember, my LDL-C was in the mid 120s to  150s regardless of what I ate, and my total cholesterol was never more than 260. My first NMR was the one last year, so I don't know what my LDL-P values were prior to 2013, but I'm assuming they were above the optimal range, although likely not over 1600. You can see that my LDL and total cholesterol each went up about 100 points and my LDL-P increased by 700 points in a 10-month period. My triglycerides even went up somewhat, although 60 is still pretty low. Although it's my understanding that LDL-C in an NMR is measured directly rather than by using the Friedewald equation (maybe a lipid expert can confirm this), when I plugged my numbers into an online calculator that estimates LDL-C, I got exactly the same number as in the NMR report, 221, for the Friedewald equation and 182 for the Iranian formula (The Iranian formula is believed to be more accurate when triglycerides are over 400 or less than 100).

You may be wondering what LDL-P is, since it's not reported in a standard lipoprotein profile and most doctors don't order it. Dr. Axel Sigurdsson does a great job explaining everything you ever wanted to know about it in his post about LDL-P, but I'll try to give a quick summary. LDL-P is a measurement of the number of LDL (low-density lipoprotein) particles in your blood which carry cholesterol, triglycerides, and another type of fat called phospholipids. According to lipidologists (experts in the field of cholesterol and other lipids), LDL-P is the strongest predictor of risk for cardiovascular disease (CVD) and future cardiac events. Total cholesterol greater than 300 and LDL-C greater than 190 are also associated with significant CVD risk. High levels of LDL-C are prone to oxidation, and oxidized LDL has been linked to the development of arterial plaque and coronary artery disease (CAD). Sometimes people have normal LDL-C and high LDL-P or vice versa (the term for this is discordance), but most people with very high LDL-C have high LDL-P as well. These findings are from recent studies, not decades-old research reported by Ancel Keys.

I want to make it clear that this type of dramatic elevation in LDL-C and LDL-P doesn't occur in most people who adopt a very-low-carb, high-fat diet. I've seen estimates that somewhere between one quarter and one third of low-carbers experience this. I've met and read about several who have. Most people who eat VLCKDs see their cholesterol rise only slightly, not at all, or even decrease, remaining within or near the normal range. I've met plenty of folks like this as well. I've also spoken with people who tell me their LDL cholesterol has always been over 200 and didn't really change after switching to a VLCKD. This is in sharp contrast to what happened to me: going from relatively stable LDL-C between 120s-150s to 221 within a very short period of time.

Of course, many things can affect a person's cholesterol levels, including stress, illness, and injury. Aside from familial hyperlipidemia (FH), there are other genetic disorders of lipid metabolism. Some people's livers produce large amounts of cholesterol (hyper secretors), while others absorb a lot of cholesterol from food (hyper absorbers), and some have both of these issues. My past lipid profiles didn't suggest FH, and I haven't been tested to see whether I have increased hepatic cholesterol production or increased intestinal absorption. I assume I'm probably a hyper secretor, since my levels were higher than average even during my 10 years as a low-fat vegetarian who ate a lot of egg whites but very few yolks or other cholesterol-containing foods.

I do have a family history of heart disease on both sides. My maternal grandfather suffered four heart attacks (the last one fatal), and my maternal grandmother also had coronary artery disease (CAD). My dad's brother has had two heart attacks, and his mother had CVD and died of a stroke. My mom has been on statin therapy since she was diagnosed with CAD ten years ago. (I'm not going to debate the risks vs. benefits of statin therapy in this post, but I'm not a big fan except in certain instances.) You may be wondering what kind of diet my relatives followed. Given that they all grew up and spent their entire lives in Switzerland (with the exception of my mom, who immigrated to the US at age 19), they obviously weren't following the Standard American Diet, but they weren't low-carbers either. My grandfather smoked and had diabetes, and my mom smoked for many years, but my other relatives didn't, and all were moderately active. I've never had a calcium scan or a carotid-intima thickness test(CIMT) to check for atherosclerosis but am looking into having these done. Even if they show no disease at this point, my goal is obviously preventing CAD, heart attack, and stroke in the future. 

My NMR results indicate I have the large, pattern A type of LDL with a low number of the more atherogenic small LDL particles (small LDL-P). This is definitely a good thing. However, although I've heard large, fluffy LDL characterized as "harmless" and even "protective," I'm having trouble finding convincing evidence supporting this assertion, especially in the setting of cholesterol levels as markedly elevated as mine. In fact, the authors of the Multi-Ethnic Study of Atherosclerosis (MESA) study summed up their findings as follows:

"Contrary to current opinion, both small and large LDL were significantly associated with subclinical atherosclerosis independent of each other, traditional lipids, and established risk factors, with no association between LDL size and atherosclerosis after accounting for the concentrations of the two subclasses."

Subclinical atherosclerosis is the period when changes are happening in the arteries but the hallmarks of atherosclerosis (i.e., plaque and fatty streaks) haven't developed to the point where the disease can be diagnosed. 

It's been pointed out that no studies have been conducted on people following VLCKDs who have very high LDL-C and LDL-P levels, and that's certainly fair to say. However, according to many MDs with expertise and/or personal experience in this area, we really don't know whether CVD risk is lower in low-carbers with cholesterol elevations of this magnitude. 


What Do The Experts Say About Very High LDL-C and LDL-P?

I studied lipid metabolism in college as part of the coursework required to become a registered dietitian, but I'll be the first to admit that I have no expertise in that area. I think it's important to listen to the experts in this field since they best understand all of its complexities, including the genetic variations that influence cholesterol levels and the development of CAD. Keep in mind that the physicians listed below are all advocates of carbohydrate restriction to some degree.

Dr. James Underberg is a lipidologist and hypertension specialist in New York City who told me that he has seen similar dramatic increases in total and LDL cholesterol in some of his patients following a carbohydrate-restricted diet. One of the interventions he recommends in these cases is replacing a portion of dietary saturated fat with monounsaturated and polyunsaturated fat sources.

Although technically not a lipid expert, Dr. Rakesh "Rocky" Patel is very familiar with current lipid research as a family doctor in Arizona with hyperlipidemia who treats many people with diabetes and metabolic syndrome. He recommends the CarbNite (cyclical low-carb) method for most of his patients and also follows this approach himself. Back in the fall of 2012, he wrote a fantastic blog post entitled Does LDL-P Matter? in which he described improvement in his carotid intima thickness despite a significant increase in LDL-C and LDL-P after switching to a carbohydrate-restricted diet. When I received my NMR results from April, I asked him if we've learned any more about very high lipids in the context of a VLCKD since he wrote that piece. He responded:

"Not really. It really is an understudied issue. Unfortunately, all the trials in the literature involve the Standard American Diet. Really, I think that before we engage in any discussion regarding cholesterol, one has to establish if atherosclerosis is present in any form. So using testing like CT calcium scoring, carotid intimal thickness testing (CIMT), and genomic scoring (Corus CAD, Cardiodx) becomes imperative and certainly provides context to the lipids." 

Dr. Axel Sigurdsson is a cardiologist who practices at a large university hospital as well as a private heart clinic in Iceland. In my opinion, his Doc's Opinion blog provides some of the most balanced, easily understood information about lipids and cardiovascular disease online. I described my experience to him and asked for his thoughts. His response:

"I've seen this lipid response (a very high jump in LDL-C and LDL-P) a number of times in individuals who adopt a low carb/ketogenic diet with relatively high amounts of saturated fat. It seems that a certain percentage of people react in this way. In fact, the lipid response to this type of diet may be genetically determined. Of course, we know that high LDL-C and LDL-P are associated with increased risk of CHD (coronary heart disease). However, nobody really knows what it means in this metabolic situation (nutritional ketosis) and to what degree it is associated with increased risk. Some claim it's not, but I think the evidence is lacking for such a conclusion. On the other hand, we also know that many people with high LDL-C and high LDL-P never have CHD. Of course, you may be one of those people. However, it is difficult to ignore altogether the possibility that high LDL-C and LDL-P may increase the risk of atherosclerotic problems."

Lipidologist Dr. Thomas Dayspring wrote an excellent article about a woman who had an experience similar to mine on a low-carb, high-fat diet, although her case involved weight loss as well. The article is available from his Lecture Pad series, and I highly recommend reading it in its entirety. (You'll have to register to view it, but registration is free). Although it may not always seem like it, he's actually quite supportive of carbohydrate restriction, particularly for people with metabolic syndrome. I didn't discuss my case with him, but here are two quotes from that article:

"We now recognize that the cholesterol usually gains arterial entry as a passenger inside of an apoB-containing lipoprotein (the vast majority of which are LDLs) and the primary factor driving LDL entry into the artery is particle number (LDL-P), not particle cholesterol content (LDL-C)."

"Could the low-carb crowd be outliers and in them we can ignore LDL-C and LDL-P? The advocates of those diets say there is no study showing harm of elevated LDL-P and LDL-C in patients who have eliminated or drastically reduced their insulin resistance and inflammatory markers by low carbing. That is true, but what they want to ignore is that there is no data anywhere that shows they are an exception. Their belief is that by reducing all other atherosclerotic risk factors and normalizing their arterial wall and endothelial biology, that apoB-containing lipoproteins like LDL cannot enter the arterial wall. Although LDL-C and LDL-P in plasma are high, none of the cholesterol content of the apoB particles gains entry into the arterial wall. Is that plausible??? Sure! But is that also erroneous or wishful thinking? Sure? Does one want to bet their CV health or life on a plausible theory? Some do and some do not. Seems to me the first step is to do what this woman did: adjust the nutritional regimen."

He also states that when ketone bodies are present in excess, they can enter the cholesterol synthesis pathway, thereby increasing serum cholesterol levels.  

While I agree with Dr. Dayspring on several issues, I disagree with his position (stated in another great article, Understanding the Entire Lipid Profile) that cholesterol-lowering medication is indicated for everyone with LDL-C greater than 190. I think nutritional intervention should be tried first, as it seems to be effective for at least a portion of people willing to do it.

Some of you may have seen spikes in cholesterol similar to mine after being on a low-carbohydrate, high-fat diet for a short period of time or possibly after a few years. You may not be that concerned, and I can understand that given the many positive effects LCHF can have on health, including certain cardiac risk factors. I also think there are still a lot of unanswered questions regarding the risk of elevated cholesterol in the setting of low insulin levels and optimal blood glucose control. But based on the evidence we do have, along with my strong family history of heart disease, I just wasn't comfortable with my numbers. And although I haven't seen this happen in any of my clients yet, I'd definitely recommend some sort of dietary intervention for them if it occurs in the future.

Dietary Changes and NMR Results from June 2014

Over the past two months I made a few small but significant changes to my diet in an effort to lower my cholesterol levels:

1. I cut back on saturated fat, particularly dairy fat and coconut oil, which contain the types of saturated fatty acids with the greatest potential to raise cholesterol.

2. I increased protein back to my previous intake of about 100 grams per day.

3. I doubled my net carb intake from 20 grams to 35-45 grams per day.

4. I began having chia seeds almost every day.

5. I ate sardines 4-5 times a week.

I still eat plenty of saturated fat, including some dairy fat. I drink coffee and tea with half-and-half (only 1 gram of carb in 2 Tbsp), always order Insalate Caprese made with fresh mozzarella at Italian restaurants, and continue to eat eggs cooked in a little butter for breakfast every other day. I still have burrata, ricotta, and moscarpone occasionally and continue eating red meat about 3 times a week. My total fat intake now ranges from roughly 80-100 grams per day, which is about 50-65% of my total caloric intake. That's still a LCHF diet! And in my case, it's also a mildly ketogenic one, since when I've checked my ketones in the morning (again, I only do this sporadically), they've been 0.4-0.8. Personally, I don't see the need to be in ketosis for my own health; to control my blood glucose, I eat a low-carb diet which just happens to be ketogenic. My weight hasn't changed (which was my goal), energy levels are good, sleep is excellent, etc.

I just received my new NMR results from labs drawn earlier this week:

My LDL-P and LDL-C are still higher than I'd like, but they've dropped considerably in a short period of time. I'm especially impressed by the 44-point drop in my LDL-C. My HDL decreased a bit as well but is still quite high. Considering this occurred in less than two months, I'm pretty happy with these results and hope they continue to improve until they return to the "Above Optimal" to "Borderline" ranges, which I consider normal for me.

As I said at the beginning, I'm a strong proponent of a low-carbohydrate lifestyle. I don't think that's ever going to change. But I feel it's important to look beyond the benefits and address the changes in lipids some people experience that could potentially have adverse effects. This was an n=1 experiment, of course. Remember, most people won't experience extremely high cholesterol levels on a VLCKD. But for me and others who do, I don't believe in shrugging it off and dismissing the results of studies because their subjects weren't following a carb-restricted diet. As a dietitian, I just can't say, "Go ahead and eat as much butter, cream, and bacon as you want. It doesn't matter how high your LDL-C and LDL-P are as long as you're eating low carb and your other markers are low," even if that's what many want to hear. Because we just don't know at this point. Maybe one day there will be evidence demonstrating that VLCKDs are cardioprotective even in the setting of significant hyperlipidemia. I truly hope that's the case. But in the meantime, I'm going to eat a low-carb diet that keeps my lipids in a range I feel more comfortable with.

***UPDATE: Recent NMR results, cardiovascular disease risk and what I eat

References

1. Otvos JD, et al. Clinical Implications of Discordance Between LDL Cholesterol and LDL Particle Number. J Clin Lipidol. 2011 Mar-Apr;5(2):105-13 
2. El Harchaoui K, et al. Value of low-density lipoprotein particle number and size as predictors of coronary artery disease in apparently healthy men and women: the EPIC-Norfolk Prospective Population Study. J Am Coll Cardiol. 2007 Feb 6;49(5):547-53
3. Cromwell WC, et al. LDL Particle Number and Risk of Future Cardiovascular Disease in the Framingham Offspring Study - Implications for LDL Management J Clin Lipidol. 2007 Dec;1(6):583-92
4. Mora S, et al. LDL particle subclasses, LDL particle size, and carotid atherosclerosis in the Multi-Ethnic Study of Atherosclerosis (MESA). Atherosclerosis 2007 May;192(1):
211-7
5. Waterworth DM, et al. Genetic variants influencing circulating lipid levels and risk of coronary artery disease. Arterioscler Thromb Vasc Biol. 2010 Nov;30(11):2264-76
6. Moriel P, et al. Lipid peroxidation and antioxidants in hyperlipidemia and hypertension. Biol Res. 2000;33(2):105-12
7. Ohlsson L. Dairy products and plasma cholesterol levels. Food Nutr Res. 2010 Aug 19;54
8. Mensink RP, et al. Dietary saturated and trans fatty acids and lipoprotein metabolism. Ann Med. 1994 Dec;26(6):461-4