By Matt Lalonde
To the uninitiated, making sense of the vast and ever-expanding nutrition literature may appear to be an incredibly daunting, if not impossible, task. Where does one start and what is the best way to identify legitimate research? Having a scientific background certainly helps but I’d argue that, with sufficient determination, any individual could sift through papers in the field of nutrition and make sense of the mayhem. I thought I would offer a few tips by answering George’s question:
Hi Robb –
I’ve subscribed to an email newsletter put out by Dr. Gabe Mirkin for years. Dr. Mirkin is an MD and endurance athlete (a competitive 71 year-old bicyclist). He writes a lot about sports nutrition, and I’ve found that his advice generally runs about 180 degrees from yours. Below is a snippet from today’s email where he cites a new study that shows that a high-carb low-fat diet provides better muscle growth than a low-carb high-fat diet. I’d love to hear your take.
Dr. Gabe Mirkin’s Fitness and Health E-Zine? January 24, 2010
Obesity and High-Fat Diet May Both Prevent Muscle Growth
A study from the University of California, Davis shows? that a high-fat diet prevents exercising mice from enlarging their? muscles (Journal of Physiology, December 2010). The mice ?received either a low fat, high carbohydrate diet or a high fat,? low carbohydrate diet for 14 weeks. Each group was divided into? those who performed progressive resistance exercises with their? plantaris muscles or those that did not do this exercise.? Those who exercised on the low fat, high-carbohydrate diet had? substantially larger muscles than those who exercised on the? high-fat diet. Chemical analysis of their muscles showed that the ?high fat diet group had lower levels of polysomes (Akt and S6K1)? necessary for making protein.? If this study can be applied to humans, it will mean that ?not only does a high-saturated-fat diet make you fatter, it also? keeps you from enlarging your muscles. We know that both full fat? cells and eating large amounts of saturated fats (the dominant fat? in meat) turns on your immunity to cause inflammation that can ?prevent the body from making protein necessary for enlarging? muscles. (Journal of Nutrition, January 2009). A high saturated-fat? diet also blocks insulin receptors and thus prevents your body from? responding to insulin, which is necessary for muscles to heal from ?intense workouts. Insulin drives amino acids, the protein building? blocks, into muscles to help them heal faster. Anything that blocks? muscles’ ability to respond to insulin will decrease amino acid? entry into muscles and thus delay healing so you can’t recover as? fast for your next workout. Further journal references and? recommendations can be found at
The study in question is “Chronic High Fat Feeding Attenuates Load-Induced Hypertrophy in Mice” M. Sitnick, S. C. Bodine, J. C. Rutledge J. Physiol 2009, 587(23), 5753–5765. The first thing to do when evaluating research is to download the full paper and carefully read in its entirety. This may sound obvious but most people make the mistake of focusing exclusively on the abstract, or worse yet, on the conclusion. Details that may make or break a paper can be found anywhere in the body of the text. I particularly like to scrutinize the methods section (a.k.a. procedures, experimental details, or supporting information). I always ask myself the following questions when attempting to assess the legitimacy of nutrition research: a) Was this a controlled trial or an observational study (epidemiological, cohort, and prospective studies are similar in nature to observational studies)? b) Was the research performed on human subjects? c) What are the sources of funding for this research and do the authors have any conflicts of interest or specific agendas that could bias the study’s outcome? Getting answers to these questions will save you a lot of time.
I tend to ignore anything that is an observational, epidemiological, cohort, or prospective study because they can only establish correlation between monitored variables, not causation. I’m not going to elaborate any further on the subject. Instead, I’ll refer readers to an excellent post by Dr. Eades that highlights the shortcomings of observational studies.
Controlled trials are typically more reliable than observational studies. These trials attempt to establish cause and effect by performing experiments where specific variables are isolated and manipulated while others are kept constant. The caveat here is that, in a living system, changing only one variable while keeping the others constant is nearly impossible. As such, there is a very real possibility that the observed experimental outcome is the result of manipulating an unknown variable that is dependent on the variables that were controlled.
The study by Sitnick and coworkers is in fact a controlled trial. The researchers acquired 5-week-old male mice and put them on a low-fat diet for 2 weeks (10% fat). After two weeks, the animals were randomly assigned either to a high-fat diet (HFD, 45% fat) or a low-fat diet (LFD, 10% fat). After a period of 14 weeks on the diets, the mice were randomly assigned either to a sedentary group or a functional overload group until the end of the study. Contrary to Dr. Mirkin’s impression, the mice were not separated into sedentary and exercise groups. Mice can’t perform back squats and deadlifts, so the researchers induced mechanical overloaded in the animals’ plantaris muscles by surgically ablating the soleus and gastrocnemius muscles. I’m not kidding, here is the procedure as described by the authors:
“Briefly, mice were anaesthetized with 2–4% isoflurane, and using aseptic surgical procedures, an incision was made to the lower hind limb exposing the ankle extensor muscle complex. The soleus and one-third of the lower medial and lateral gastrocnemius were carefully removed with particular attention to preserving the neural and vascular supply to the plantaris muscle. The incision site was irrigated with sterile saline and closed using subcuticular sutures. This procedure was performed on both hind limbs. At 3, 7, 14 and 30 days post surgery, animals were anaesthetized with 2–4% isoflurane and the plantaris muscles were excised, weighed and flash frozen in liquid nitrogen.”
There is plenty of material to delve into here but let’s start with the diets. Click on these links to see the composition of the high-fat and low-fat diets. The caloric breakdown for the high-fat diet is 20% protein, 35% carbohydrate, and 45% fat. The carbohydrate content of the HFD is too high for it to be considered a ketogenic diet (where 10% of calories from carbohydrate is typically the norm). This is just a detail; let’s continue with our analysis.
The protein source in both the HFD and LFD is casein (200g), which is supplemented by a minute amount of L-cystine (3g). Incidentally, casein is the most problematic protein fraction in milk with regards to the leaky gut syndrome and autoimmune diseases.
Half of the carbohydrate content of both diets is table sugar (a source of glucose and fructose) whereas the rest is derived from cornstarch and maltodextrin 10 (sources of glucose). Vitamins and a few other goodies are thrown in the mix to make sure the mice don’t die prematurely while on this wholly inadequate diet.
Things get interesting when we look at the fat sources. In the HFD, 177.5 grams of fat come from lard and 25 grams from soybean oil. On the other hand, the LFD contains 25 grams of fat from soybean oil and 20 grams from lard.
Now the fun part. Here is what happened to the mice on the HFD:
“At the end of the experiment, the control HFD mice were 31% heavier than the LFD mice (Fig. 1 and Table 1). Furthermore, following 14 weeks on the high fat diet, mice were hyperinsulinaemic, having a 7.6-fold increase in fasting insulin levels relative to LFD mice (Table 1).”
“In both LFD and HFD mice, functional overload led to a significant increase in plantaris mass at all time points studied (Fig. 2); however, significant differences were noted in the absolute growth response between mice in the two diet groups at 14 and 30 days. At 7 days the growth response of the LFD mice (mean LFD – mean HFD/mean LFD°—100) was8%greater (P >0.05) relative to the HFD mice and increased to a significant difference of 10 and16% (P <0.05) at 14 and 30 days, respectively. At 30 days of FO, the relative difference between control and overloaded muscles (mean FO/mean control) was 2.1- and 1.7-fold in the LFD and HFD mice, respectively. These data reveal that diet-induced obesity resulted in a significant impairment in the response of skeletal muscle to increased mechanical loading.”
So the mice on the HFD developed symptoms of metabolic syndrome and their plantaris muscles showed less hypertrophy compared to those of the mice on the LFD. And this is in spite of the fact that the HFD mice were heavier. Now that must be telling us that fat is really bad right? Wrong!
Stephan Guyenet recently pointed out that the symptoms of metabolic syndrome observed in rodents fed a high-fat diet are actually due to an omega-6 : omega-3 imbalance caused by the low omega-3 and/or high omega 6 content of the fat sources employed in the studies. This is a perfect example of how controlled trials can reach erroneous conclusions by inadvertently manipulating variables that were not taken into consideration. Unbeknownst to the researchers, modifications in the amount of saturated fat consumed by the rodents changed the omega-6 : omega-3 balance of the diet, which increased inflammation (1).
There are a few other things to consider here. First, mice consume mostly plants (leaves, roots, seeds, stems) and insects in the wild and most likely did not evolve while consuming large quantities of fat, let alone lard. As such, it seems appropriate to assume that mice are not adapted to a high-fat diet. Second, regardless of the diet, how can we possibly assume that an experiment involving mice who’s muscles were artificially loaded can tell us anything about the ideal dietary requirements for lean muscle mass gains in human beings? Seriously!? This may seem obvious to most folks but it is important to underscore the fact that the conclusions of scientific research are only relevant within the specific context under which the experiments were conducted. In this case, the findings of the study only apply to mice. For this reason, I tend to ignore studies that were not conducted on human subjects. I acknowledge the fact that there are many gene sequences that are conserved between humans and mice (as well as other mammals and other species). I don’t consider studies that are performed on mice (or other species) to be completely useless. Like observational studies, they are useful for testing and generating new hypotheses. Nevertheless, I don’t consider the results of these studies to be applicable to human beings until the hypotheses have been specifically tested in human beings (2). When scientists extrapolate the findings of a study to another species, as in “this may apply to humans”, they are either speculating or generating a new hypothesis that will need to be validated with further research (on human subjects). Dr. Mirkin did highlight this fact in his newsletter when he wrote, “If this study can be applied to humans”. Unfortunately, the remainder of Dr. Mirkin’s newsletter is written in a fashion that gives the reader the impression that the outcome of the study does apply to humans. Dr. Mirkin proceeds to reference more flawed research to support his belief that consumption of saturated fat is detrimental to human health. The tone of Dr. Mirkin’s newsletter, as well as the fact that his advice runs counter to Robb’s (according to George), indicate that he is biased toward low-fat diets. He even comments that “high-saturated-fat diet make you fatter”. I would recommend that Dr. Mirkin read Gary Taubes’ Good Calories Bad Calories for an alternative interpretation of the research on the subject. It appears as though the diet-heart hypothesis is on its death bed and that the weight of the evidence is making prominent researchers reconsider their lipophobic stances (see Whole Health Source Blog here and here).
In the end, my take on this study is very different than Dr. Mirkin’s. I don’t think the outcome of the Sitnick study is at all relevant to human beings. Aside from the fact that mice were used as subjects, there exists the possibility that the high-fat diet induced obesity and metabolic syndrome through an omega-6 : omega-3 imbalance. As such, saturated fat itself may not be detrimental to the health of rodents. Even if we were to assume that the saturated fat itself was responsible for the obesity and symptoms of metabolic syndrome observed in the mice, would that give us any insight into the effect of saturated fat on the metabolism of human beings (a species that evolved while consuming some saturated fat)? Absolutely not!
It is interesting to look at the conclusion of the study:
“The present study represents the first investigation to show that diet-induced obesity has a negative impact on the ability of skeletal muscle to adapt to growth signals such as mechanical loading. A decrease in the ability of muscle to respond to growth signals could adversely affect glucose homeostasis, as well as prevent recovery from injuries and accelerate the effects of ageing. Additional studies are required to investigate the mechanisms by which obesity interferes with the response of skeletal muscle to growth stimuli.”
Note the use of the words “diet-induced obesity” instead of “high-fat diet-induced obesity” or “saturated fat-induced obesity”. This correctly states that the obesity was induced by the diet but leaves open the question as to what exactly caused it and the metabolic disorder.
Ultimately I do understand why the researchers performed the study with mice as opposed to humans. An equivalent study with human subjects would have required real food, dieticians, coaches, accurate methods for determining increases in lean muscle mass in living subjects, and would be prohibitively expensive.
My personal experiment of one tells me that a ketogenic diet is excellent for gaining lean muscle mass while avoiding gains in fat mass (I have now gained a total of 15 pounds of lean muscle mass). I readily acknowledge that a ketogenic diet might not be ideal for everyone seeking to gain muscle mass and that my experiment of n=1 has little to no scientific value. Then again, ketogenic diets have been shown to improve insulin sensitivity in human beings and Dr. Mirkin acknowledges the importance of insulin sensitivity in muscle repair and recovery. It’s almost as if personal experience that is guided by legitimate scientific research trumps the nutritional advice offered by mice. Go figure.
(1) The high- and low-fat diets in the Sitnick study provided equal amounts of soybean oil. As such, we would have to assume that the lard originated from grain-fed pigs and contained a significant amount of omega-6 fatty acids as a result.
(2) Interesting side note; beta-lactam antibiotics are toxic to guinea pigs. We would have seriously missed out if we had initially tested these antibiotics on guinea pigs instead of humans.