Monday, August 31, 2015

Moy et al (Low Fat Diet for CHD Risk Factors)

Hooper et al included this trial by Moy et al as a reduced fat trial in their 2012 meta-analysis and as a reduced SFA trial in their 2015 meta-analysis


Participants were 235 ‘apparently healthy people’* aged 30-59 years old who were brothers and sisters of people who were diagnosed with CHD before 60 years of age.  The participants were randomised to ‘nurse counselling’ or to ‘usual care by their primary physicians’ for 2 years.  The purpose of this study was to examine whether dietary counselling from specially trained nurses would be more effective than the usual care provided by doctors

The nurse counselling involved individualised instructions to lower total fat intake usually to < 40g of total fat, as well as self-monitoring logs that were used for the first 1–2 months and regular check-ups approximately every 6–8 weeks to reinforce the diet, evaluate dietary compliance, and measure lipids.  For more information see the methods

* Additional eligibility criteria included at least one of the following: LDL-C ≥ 3.4 mmol/l (130 mg/dl), blood pressure ≥ 140/90 or current use of anti-hypertensive medication, or currently smoking

** The groups were similar at baseline except for significantly more smokers in the usual care group (39 vs. 53) and slightly lower LDL-C in the usual care group (4.7 vs. 4.3 mmol/l)


The nurse counselling group slightly reduced their calorie, total fat and saturated fat intake.  Protein and carbohydrate intake were not reported, but probably didn’t change much as the reduction in fat is responsible for almost all (85%) the reduction in calories.

Nurse Counselling
Usual Care
Energy (kcal)
Total fat
85.1 (38.0%)
85.0 (38.3%)
-14.3 (-3.9%)
4.7 (-0.27%)
Saturated fat
30.2 (13.5%)
29.7 (13.4%)
-4.9 (-1.4%)
1.9 (0.0%)

After 2 years triglycerides significantly decreased and HDL-C slightly increased in the low fat group perhaps due to some weight loss (which wasn’t significant, P = 0.4398) from the reduction in calories (P = 0.0085).  LDL-C was the only significant difference between the groups regarding BMI and blood lipids

Nurse Counselling
Usual Care

CHD/CVD events and mortality was not reported in the study.  Hooper et al reported the following outcomes:

Nurse Counselling
(N = 117)
Usual Care
(N = 118)
Myocardial Infarction
(all were non-fatal)
CHD Events
Combined CVD Events

Hooper et al assessed this trial as having a high risk of bias in ‘systematic difference in care’ and an unclear risk of bias in ‘dietary differences other than fat’

Sunday, August 23, 2015

An Upcoming Meta-Analysis on Saturated Fat, Polyunsaturated Fat and Coronary Heart Disease

I recently finished writing a meta-analysis on saturated fat, polyunsaturated fat and coronary heart disease.  It’s now ready for feedback before I revise it and submit it to a journal.  If you would like to read it and provide feedback please send me an email at and I’ll send it to you.  However, to avoid the possibility of plagiarism I will only send it to people I know well.

The meta-analysis includes some of the content I’ve posted on the blog in the last several months and is basically that many of the SFA vs. PUFA trials were confounded by differences between the groups not related to SFA or PUFA, and that the results after exclusion of those trials suggests no benefit (or harm) from replacing saturated fat with polyunsaturated fat, as well as discussing some related issues.

If you would like to provide feedback please do so by the 6th of September (2 weeks from now) as it would ideally be published in time for my presentation at the AHSNZ symposium in Queenstown on October 23rd-25th (which will give me a little over 6 weeks).

I’m also open to suggestions for a good/appropriate journal to publish in.  I’ve begun to have a look around and most journals have a word limit quite a bit lower than 10,000 words, not to mention reference and table/figure limits.

Wednesday, August 19, 2015

Unraveling the Mechanisms of Insulin Resistance

Insulin Signalling Defect

It’s well understood that insulin resistance typically precedes the development of type 2 diabetes by many years.  Not only is insulin resistance a key risk factor for type 2 diabetes, but it’s also a risk factor for many other chronic diseases, so understanding what causes insulin resistance and effectively dealing with it can solve many problems at once

Perhaps the dominant idea of what causes insulin resistance is that it’s due to a defect or inhibition in the insulin signalling pathway.  You might be familiar that one of the actions of insulin is to increase muscle glucose uptake by causing the translocation of a glucose transporter called GLUT4 to the cell membrane so that glucose can pass from the bloodstream to the muscle cell and either be used or stored.  The nitty gritty of physiology is complex, for insulin cause the translocation of GLUT4 to the cell membrane it must first activate a complex signalling pathway shown below.  In summary [1] [2]:

·         Insulin binds to insulin receptor (IR)
·         The insulin receptor tyrosine phosphorylates the insulin receptor substrate (IRS)
·         The tyrosine phosphorylated IRS associates with PI3K
·         PI3K activates Akt
·         Akt inhibits AS160 (Akt substrate 160) and GSK3 (glycogen synthase kinase 3)
o   Inhibition of AS160 increases the activity of certain enzymes that lead to the translocation of GLUT4 to the cell membrane
o   Inhibition of GSK3 promotes the activity of glycogen synthase, leading to increased glycogen synthesis


Several human studies have examined the activation of the insulin signalling pathway in muscle during a euglycemic-hyperinsulinemic clamp.  The euglycemic-hyperinsulinemic clamp is a technique whereby insulin is infused to high levels and glucose is also infused to maintain euglycemia (normal glucose levels, usually 5.0 mmol/l).  The euglycemic-hyperinsulinemic clamp is not a method that replicates physiological conditions*, but is the gold standard for measuring insulin resistance.  These studies have consistently found a defect in insulin-stimulated IRS-1 association with PI3K among people with type 2 diabetes, but are less consistent with other parts of the insulin signalling pathway [3] [4] [5] [6] [7] [8].  In addition, a defect in IRS-1 association with PI3K has also been found in women with gestational diabetes compared to pregnant women during an oral glucose tolerance test [9]

However, there are some issues with this mechanism:

·         While studies have consistently found that there is a defect in IRS-1 tyrosine phosphorylation/IRS-1 association with PI3K among individuals with IR/T2D during a euglycemic-hyperinsulinemic clamp, this has not been tested under more physiological conditions such as an oral glucose tolerance test or a standardised meal
·         While studies have consistently found that there is a defect in IRS-1 tyrosine phosphorylation/IRS-1 association with PI3K among individuals with IR/T2D, there is inconsistent evidence that this translates into impaired insulin signalling downstream of IRS-1/PI3K, with some studies finding that there is impaired insulin-stimulated Akt phosphorylation [3] [5] [10] and others finding that it is normal [7] [8] [11]
·         Even if insulin-stimulated Akt activation was impaired, very little Akt activation is needed for maximal GLUT4 translocation, therefore a defect in insulin-stimulated Akt activation of 50% may not be relevant in postprandial conditions and offset by higher fasting insulin in basal conditions [10]
·         Only a few of those insulin signalling studies have found that activation of the insulin signalling pathway is not reduced under basal conditions [4], whereas many find it to be normal or even slightly elevated [5] [7] [8] [10] [11]
·         Current glucose lowering drugs do not target any of the proteins in the insulin signalling pathway.  Thiazolidinediones promote fat storage, which lowers plasma NEFA, thereby increasing glucose oxidation.  Biguanides (such as metformin) reduce gluconeogenesis by inhibiting complex I of the mitochondrial electron transport chain [12] [13], rather than by AMPK dependent mechanisms [12] [14] [15].  And sulfonylureas reduce glucose levels by increasing insulin secretion
·         Growth hormone impairs glucose tolerance and insulin sensitivity without inhibiting the insulin signalling pathway [16] [17], suggesting that insulin resistance can develop through alternative mechanisms

* There are many differences between a meal or OGTT and the euglycemic-hyperinsulinemic clamp (see below)

Meal or OGTT
Glucose levels
Transiently high
Basal (~5 mmol/l)
Glucose delivery
Portal vein
Systemic circulation
Portal signal (very high glucose in portal vein)
Insulin levels
Transiently high
Chronically very high
Insulin delivery
Portal vein and systemic circulation
Systemic circulation

The unusual concentrations and delivery of both glucose and insulin into the system means that tissues respond differently to the clamp compared with a meal or an OGTT.  You may have heard that skeletal muscle is responsible for ~70-80% of insulin stimulated glucose uptake.  This is correct in relation to the euglycemic-hyperinsulinemic clamp, but is the not case following a meal or an OGTT, where muscle and liver each take up roughly a third of the ingested glucose [18].  The clamp doesn’t result in the high levels of glucose and insulin levels in the portal vein that enhance liver glucose uptake and the concentrations of glucose and insulin play to the strengths of skeletal muscle due to differences in the main glucose transporters and glucokinase enzymes in each tissues.  Regardless the extrapolation that skeletal muscle takes up 70-80% of glucose from a meal doesn’t make sense when you consider that the brain uses a lot of glucose and liver needs to store glycogen to supply it continuously.  Anyway, something you don’t need to know but may find interesting

Inflammation and Insulin Resistance

Whether insulin resistance is cause by a defect in the insulin signalling pathway is a relevant question because it will influence how insulin resistance is treated.  The good thing with that particular diagram is that it also shows how other things affect and/or are effected by the insulin signalling pathway, as well as showing the pathways responsible for the anabolic effects of insulin.  For example, note that TNF-α, JNK, inflammatory cytokines, SOCS proteins and PTP1B are shown to inhibit the insulin signalling pathway.  The inflammatory cytokines do this by serine phosphorylating IRS-1, which inhibits tyrosine phosphorylation, thereby inhibiting the ability of IRS-1 to associate with PI3K [2].  As such, one of the prime candidates for this defect or inhibition of the insulin signalling pathway is chronic inflammation

An inflammatory cause of insulin resistance and type 2 diabetes is supported by several observations:

·         Elevated markers of inflammation precede and can predict the development of type 2 diabetes [19] [20] [21] [22]
·         Anti-TNFa drugs for rheumatoid arthritis is inversely associated with development of T2D [23] [24]
·         Gene knockouts of some inflammatory cytokines protect against diet induced insulin resistance in rodents [25]

While inflammation seems to be an attractive mechanism for insulin resistance, it’s unfortunately not so simple:

·         An IL-1R antagonist (anakinra) lowers fasting glucose, but does so by improving β-cell function, rather than by improving insulin sensitivity* [26]
·         Salicylates (such as aspirin) also lower fasting glucose, but do so by either increasing insulin secretion [27] or by reducing insulin clearance [28]
·         Inhibiting TNF-a does not improve insulin sensitivity, fasting glucose or glucose tolerance in humans [29] [30] (does improves insulin sensitivity in rodents [31])

* If inflammation is involved in the development of type 2 diabetes, it’s probably not because inflammation causes insulin resistance, but rather than inflammation, in particular IL-1, impairs β-cell function and/or promotes the destruction of β-cells [32].  There are properties of the pancreatic β-cells that make some particularly vulnerable to IL-1, and therefore a ‘weak link’.  Pancreatic islet cells express the highest density of IL-1 receptors among all body tissues, and the pancreatic β-cells seem to be particularly sensitive to IL-1 [32].  In addition there are also some unique aspects of β-cell metabolism and their vulnerability to oxidative stress that also makes them a ‘weak link’ [33]

Measuring Metabolism

Metabolomics and lipidomics are methods to simultaneously measure the concentrations of many different metabolites and lipids in biological tissues or fluids (such as plasma, saliva and urine).  These methods are often used in research to try to develop non-invasive biomarkers of disease, to better predict who is at risk.  For example, some common metabolites and lipids associated with insulin resistance and/or type 2 diabetes include:

·         Higher levels of most amino acids and fatty acids [34] [35] [36] [37]
·         Lower levels of glycine [34] [36] [37] [38] [39]
·         Higher levels of α-hydroxybutyrate (a by-product of elevated glutathione synthesis that is indicative of oxidative stress) [36] [39]

An advantage of metabolomics and lipidomics is that you can measure lots of metabolites and lipids at once.  As such there are many other metabolites and lipids I could mention such as carnitines, bile acids and glycerophosphocholine species (and other obscure lipids).  But a limitation with an approach like this is that it can be difficult to determine cause and effect.  For example, elevated levels of certain amino acids and fatty acids in plasma have consistently been found to be associated with the development of type 2 diabetes.  But do these amino acids and fatty acids promote diabetes, or are they merely elevated due to the insulin resistance that develops long before type 2 diabetes is diagnosed

A few studies have also examined how plasma metabolites and/or lipids change following an OGTT [35] [40] [41] [42] [43] or a high fat meal (P:F:C = 12:59:30) [44].  And fewer still have compared how individuals with insulin resistance and/or type 2 diabetes [35] [40] [41] [42].  To my knowledge, no studies measured tissue metabolites and lipids under fasting conditions or following a meal or OGTT, or whether they are altered in individuals with insulin resistance and/or type 2 diabetes.  Studying this could improve our understanding of metabolism and the causes of insulin resistance and type 2 diabetes as changes and alterations in tissue metabolites and lipids more closely reflects metabolism than changes in plasma metabolites and lipids.

For example, a proposed mechanism whereby insulin resistance impairs glucose uptake by tissues is that insulin resistance ultimately impairs GLUT4 translocation, which is related to the idea that insulin resistance is caused by a defect in the insulin signalling pathway.  This mechanism would predict that individuals with insulin resistance would have lower intracellular glucose compared with healthy controls and either lower or normal glucose-6-phosphate (the first step of glycolysis).  This has been demonstrated by a research group using the euglycemic-hyperinsulinemic clamp technique and nuclear magnetic resonance imaging (NMR) to measure metabolite concentrations in skeletal muscle [44], but has not been replicated.  An alternative proposed mechanism is that impaired glucose uptake is due to impaired glucose metabolism.  This mechanism would predict that individuals with insulin resistance have either normal or elevated intracellular glucose levels, but a lower concentration of glucose-6-phosphate due to reduced activity of the enzyme that catalyses the first step of glycolysis, which is seen in mice on a HFD [45]

My PhD Project

So the aim of the first study for my PhD is to characterise the insulin signalling and metabolic responses to a meal in skeletal muscle and adipose tissue in healthy individuals.  And the aim of the second study is to investigate whether these responses are altered in individuals with insulin resistance and/or type 2 diabetes.  (The third project will likely be on cancer metabolism)

We are recruiting for the first study at the moment and I’ve copied the content from our flyer below.  Feel free to ask any questions and I can send though more detailed information about the study if you’re interested.  If you’re eligible and interested to participate (or know someone that would be) we would really appreciate it.  Just please note we are at Deakin University's Burwood Campus in Melbourne, Australia:

Healthy males (18-35 years old), with a BMI of 23-30, are required for a study examining the physiological responses to a meal.

You will be screened prior to involvement to ensure eligibility.  Participation would involve 1 visit to our laboratory for approximately 4 hours, which will include:

·         Eating a standardised breakfast consisting of an egg, mozzarella cheese and a glucose drink
·         A total of 3 small muscle and fat biopsies that will be taken under local anaesthetic, before the meal and at 30 minutes and 2 hours after the meal
·         A total of 9 small blood samples (3 ml) that will be taken before the meal and at regular intervals after the meal
·         A total of 5, 5 minute periods of sitting underneath a metabolic hood before the meal and at regular intervals after the meal

You will receive $60 for your time and would benefit by receiving feedback on your: metabolic rate and glucose, insulin and other physiological responses to a meal.

Results from this study will provide crucial information about the normal physiological responses to a meal that may assist in the treatment of pre-diabetes and type 2 diabetes

This study has been approved by the Deakin University Human Research Ethics Committee (Ref No: 2015-078)

If you are interested in participating or would like more information please contact:
Dr. Clinton Bruce or Mr. Steven Hamley
School of Exercise and Nutrition Sciences, Deakin University
Email: or

Monday, August 17, 2015

Thoughts on the Trans Fat Ban, Food Taxes and Carbon Emissions from Animal Agriculture

The Ban on Trans Fats

The FDA has decided to remove artificial sources of trans fats (aTFA) from the US food supply.  This began in 2013 Fred Kummerow wrote a citizen’s petition to the FDA to ban aTFA.  The petition was ignored, so he filed a lawsuit against the FDA.  The FDA responded and now plans to progressively remove aTFA from the US food supply which will be complete in 2018

One problem is the discussion of the health effects of aTFA usually begins and ends with it increasing LDL-C and decreasing HDL-C*.  This trivialises the health effects of aTFA, making it ‘a worse saturated fat’ in the eyes of the public, and turns it into something you can offset by taking unsaturated fats, phytosterols or drugs.  But the adverse health effects of aTFA go beyond increasing the total-C:HDL-C ratio.  Fred Kummerow has discussed some other mechanisms in his petition and some other papers he has written, but I suspect the story with aTFA won’t end there either

Despite the consistent body of evidence that aTFA are harmful, there have been some concerns about the ban, mainly about government intervention and personal freedoms.  I doubt the FDA will ban butter (for example) next because the US probation on alcohol (with alcohol being another easy to make substance that people want) was a pretty big failure.  I also doubt people actually want to knowingly consume partially hydrogenated oils over other fats or oils (that taste better and make you feel better)

Unfortunately it doesn’t seem like Australia or New Zealand will ban aTFA soon.  A spokesperson for the New Zealand Heart Foundation said that saturated fat is a bigger issue.  This ignores the evidence that SFA is not associated with CHD whether you look at its effect on cholesterol levels (unless you selectively use total-C or LDL-C and ignore the effect SFA has on HDL-C), meta-analyses of observational studies or the fact that there don’t appear to be any reduced SFA clinical trials**.  But whether saturated fat is a bigger problem or not is irrelevant to whether banning aTFA will improve health.  It is a false dichotomy that you have to choose between efforts to reduce saturated fat and banning trans fats

* aTFA has this effect by increasing the activity of cholesterol ester transfer protein (CETP).  Since genetic mutations of CETP have inconsistent effects regarding CHD and CETP inhibitors don’t reduce CHD, the adverse health effects of aTFA are probably not related to its effects on LDL-C and HDL-C

** All of the included trials in the recent Cochrane meta-analysis regarding reducing SFA for CHD were either SFA > PUFA, reduced fat or Mediterranean diet trials

Fat, Sugar and Salt Taxes

It has often been said that the ‘the western diet is high in fat, sugar and salt’, thereby blaming these nutrients for the high incidence of obesity and chronic disease in ‘western’ countries.  The blame is probably being placed on pizza and soft drinks (etc), but it could just as easily be placing the blame on eggs, cheese and fruit.  Grouping foods based on macronutrient composition and salt for the purposes of dietary research, recommendations or policy can sometimes be ineffective at targeting and discriminating between healthy or unhealthy foods

‘Good’ nutrients, ‘bad’ nutrients plays into the hands of the processed food industry who will adapt and reformulate their products to meet nutrient-based guidelines, taxes or food fads (see 3:29 of this video and note the low SFA chips), whereas the sellers of whole foods will be more strongly affected as they have a very limited capacity to adapt.  Adding or removing nutrients in processed foods is unlikely to achieve much, particularly if the nutrients themselves have a pretty neutral effect of health as is the case of fat, SFA and salt (unless they’re actually toxic such as aTFA and certain other food additives).  It’s still a processed food, low in nutrients and other beneficial substances.   A cake fortified with vitamins and low in saturated fat is still a cake.  It’s still low in protein, fibre, minerals, the vitamins they didn’t add, LCO3s, phytonutrients and zoonutrients and largely made up of refined starches, added sugar and added fats.  These ‘healthy’ reformulations by industry will be aggressively marketed and may give a bit of a green light (or a poor basis for rationalisation) for consumers to eat quantities of these foods as if they were actually healthy (an excellent example of this is all the low fat/low carb/low sugar/Paleo*/etc junk food).

A possible unintended consequence is that taxes that lower the consumption of certain whole foods will likely reduce the overall consumption of other nutrients associated with those whole foods.  For example a saturated fat tax would likely lead to reduced intakes of vitamin K2 (which is found almost exclusively in animal fats) and may also reduce intake of other nutrients like protein, iron and zinc (due to a lower consumption of meat), which most people could do with more of

With a large percentage of calories in the average western diet coming from processed foods, refined starches, added sugar and added fats, surely the focus should be on replacing these with whole, minimally processed foods regardless of macronutrient consumption (within reason)

All that being said, I’m currently in favour of a tax on sugar sweetened beverages (SSBs) (I’m undecided on fruit juice too as it’s more than just liquid sugar).  Nutrition education programs/initiatives, at least the way they are currently done, is proving to not be very effective.  Intensive nutrition advice, such as a 1-on-1 or small groups with the dietician/nutritionist, isn’t a long term solution as it will be expensive and people tend to revert to their old eating behaviours when that support is gone (such as in clinical trials).  Besides, consumption of SSBs remains high despite basically everyone thinking they are unhealthy.  This is unfortunate, as ideally education would be sufficient to change behaviour, but I only need to look at myself to realise that’s not the case.  Real life is not as black and white as that and environmental pressures can be powerful even when you’re conscious of them.  A tax on SSBs is likely to reduce consumption and slightly improve health or at very least provide some funding for healthcare, and there shouldn’t be much risk of unintended consequences.  Some people may just switch to artificially sweetened beverages, which may or may not be a healthier choice and probably depends on which sweetener is used

* Paleo is not the sum total of grain free, dairy free and sugar free and in my opinion ‘Paleo’ cupcakes (and other baked goods and desserts) is an oxymoron

Carbon Emissions from Animal Agriculture and Conservation of Carbon

Food sustainability is becoming more commonly discussed, such as the new USDA dietary guidelines (discussed here), Sweeden’s dietary guidelines and also in a recent paper looking at the effects of a CO2 tax.  There are a few key issues in these discussions of sustainability including the obvious (overfishing and eating local and season food) and the idea that high consumption of animal foods is not sustainable

There are many components to ‘sustainability’, but perhaps the main reason animal foods are singled out is for greenhouse gas emissions, where animal foods are proposed to emit more greenhouse gases than plant foods.  However, some aspects of this idea don’t make much sense in the context of animals on pasture (CAFOs are obviously absurd), though I can understand how this idea may have come about

To begin with, another argument against animal foods is that much more water is needed to produce a kilogram of beef (for example) compared to a kilogram of plant food.  I’ve been very sceptical of relevance of this claim because it’s not like this water is lost, but rather almost all the water is excreted as urine.  Think about us: if we drink 2 L per day and live for 75 years, that means we drink about 55,000 L in a lifetime.  Where does that water go?  Almost all of it leaves our body in urine and sweat and we only hold onto a very small amount of that water, about 0.1%.  So long as the consumption of water exists in a closed system or is sourced sustainably then the higher water requirement shouldn’t be a problem as the water isn’t lost and just gets recycled.  However, excessive irrigation water from rivers or lakes in drought prone areas is definitely a problem (such as the Murray-Darling basin)

This leads to the greenhouse gas issue.  Very simply, plants take CO2 from the atmosphere (and water from their roots) and convert it into carbohydrates like glucose.  Animals eat the plants and extract energy from glucose first by breaking down the glucose molecule into carbons and hydrogens, where the carbon atoms are combined with oxygen to form CO2 and are removed from the body.

Imagine in a closed system like a biosphere where grass grows by taking up atmospheric CO2, cows then eat the grass and release almost all of the carbon via respiration.  In isolation, if the carbon stored as biomass (grass + soil + cows) is identical from year to year, then atmospheric carbon must also be identical.  However, atmospheric carbon will increase if the cows overgraze in this biosphere due to less carbon being stored as biomass as there will be less carbon stored as grass and not an equivalent increase in carbon stored as biomass in the cows as most of the carbon the cows eat is exhaled as CO2 (this cycle is occasionally mentioned, but I think it needs to be reinforced)

Two laws of physics include the ‘conservation of energy’ and the ‘conservation of mass’.  I think we need a similar law here, the law of ‘conservation of carbon’

With carbon being conserved, agricultural systems that store more carbon as biomass (usually in the form of topsoil and plants), such as include perennial crops (fruits, nuts and legumes) and managed grazing would reduce atmospheric carbon.  While those systems that deplete this biomass, such as annual crops due to tillage and systems with high oil use (transport, fertiliser, etc), would increase atmospheric carbon.  Why then is animal agriculture portrayed as being so harmful for the environment?

Higher carbon flux: people seem to only measure and report the rates of CO2 released into the atmosphere.  This is understandable because the increase in atmospheric carbon appears to be mostly driven by the release of stored carbon due to deforestation and the burning of fossil fuels.  However, without also measuring and reporting the rate of carbon taken up in a system you can’t know the overall balance of carbon.  If only CO2 emissions are being measured and not also CO2 uptake, then a carbon neutral system with higher carbon flux would appear worse than carbon neutral system with lower carbon flux.  Animals grazing on rapidly growing plants like grass would almost certainly have a higher carbon flux than most other forms of agriculture based on: growing plants having a higher net rate of CO2 uptake (to support growth) compared to mature plants; and that animals have a higher metabolic rate than plants.

Methane: another reason is for the focus on animal agriculture is the methane released from animals, particularly ruminants.  Methane is formed by enteric fermentation, where carbohydrates are broken down by microorganisms in an anaerobic environment (such as the rumen in ruminants), which actually occurs to a greater extent if the animals are fed grass.  Less methane released compared to CO2 and there is less methane than CO2 in the atmosphere.  However, methane is a more potent greenhouse gas (~20-30x depending on source), but has a shorter half-life (~2-10 vs. 50-200 years depending on source).  So it’s argued that reducing the release methane from livestock would be an effective quick fix for climate change, giving us more time to reduce other greenhouse emissions (assuming that happens sufficiently).  But how much of impact does enteric fermentation actually have?  Enteric fermentation has been estimated to contribute to about 13% of global methane released between 2000-2009 [1] or 26% of US methane released between 1990-2013 [2].  Currently methane contributes about 9.5% of CO2 equivalents of greenhouse gases in the atmosphere* (calculated from [3]), so the enteric fermentation contributes 1.2% to total global or 2.5% to US greenhouse gases.  This may very slightly (probably a few percent) overestimate the impact of enteric fermentation as the carbon originally came from the atmosphere.  And of course, this doesn’t reflect animal agriculture as a whole as enteric fermentation is only one component and other relevant issues include oil use, whether deforestation occurred and the effect on topsoil.  Therefore, while enteric fermentation is a notable source of greenhouse gases, I think its role has been over-exaggerated and the question of whether this meat is sustainable is probably more related to farming practices and transportation

* I don’t think it is correct to calculate the contribution of methane from enteric fermentation to overall greenhouse gases by using rates of methane emitted relative to CO2 and other greenhouse gases.  This is because the relevant measure to compare the contribution of different gases is the concentration of the gases in the atmosphere rather than how many CO2 equivalents of each gas is emitted.  Calculating the contribution of methane (including enteric fermentation) by using methane emissions (which is what often happens) overestimates the contribution as methane has a higher relative flux than CO2.  To use an extreme example to make a point, imagine you have gas A and gas B with the following characteristics.  Clearly gas B contributes more to the overall concentration of gases at year 10 even though gas A has a higher flux.  However, in this example, reducing emissions of gas A by X% is likely to be more effective than reducing emissions of gas B by X% to reduce the overall concentration of gases.

Gas A
Gas B
Concentration year 0
Emissions per year
Degradation per year
Concentration year 10