It's surprising, isn't it. Brazil nuts are unique in their ability to concentrate selenium. They absorb huge amounts of selenium partly because their deep roots reach selenium-rich soil layers in the Amazon. But there’s likely an evolutionary edge, too: high selenium concentrations in their seeds and tissues may help defend against fungi, insects, and over-hungry animals. I suspect few people are aware of the potential toxicity.
You blend cashews with coffee.. There's a combination I wouldn't have expected. Is this hot or iced coffee? What's it like?
Glad you mention toward the end the preference for raw or lightly roasted nuts. But I would think there is a very large difference between nuts eaten raw and nuts that are highly salted & roasted in seed oils, higher both in calories and omega 6. How much higher? That would be good to know.
As an Oregonian who grows hazelnuts, I say “eat more hazelnuts”! But also, two little-known facts about this true nut. First, the new varieties of hazelnuts (aka filberts) taste different than the nuts grown a generation ago; you might like them better—I know I do! Second, I volunteer at a wildlife rescue/rehab clinic, where I was thrilled to learn that the macro- and micronutrients in hazelnuts make them the single BEST nut to feed to baby squirrels! 🐿️ Not all nuts are created equal, even to our squirrel friends.
I have only tasted them twice, but I can confirm they are highly likely to be the tastiest chestnut on Earth, as American Chestnuts were historically reputably far superior to Eurasian. To me, they were like a slightly sweeter and more flavorful chestnut, but Dunstan is the only other chestnut of which I can recall the flavor. I’m glad you looked them up, the Ozark Chinquapin Foundation has created blight-resistant non-GMO non-transgenic trees, with natural novel breeding techniques, and they’re available as permitting to Americans for a small annual donation. Last year I had lab tests done on the southern subspecies (South of the Arkansas River), and those confirmed the reported high protein, but had traditional chestnut fat levels! Making it the only low-fat, high-protein tree nut. Those nutrition tests were a very small sample size, but from two labs, and supplies are so rare it will probably be a few years before I can do additional tests.
You haven’t mentioned about the cost. It is a factor for a majority of populations. In terms of balanced nutrition at affordable cost, peanuts should rank at the top. Highest protein content among the profiled lot, good oil content with good quality oil, quite rich in w-9 and lower levels of w-6.
Why should saturated fatty acids from animal fats be problematic compared to those from plant oils ? Body metabolism will see them as only palmitic and stearic acids or the triglycerides, regardless of their origin.
Hey, Moro. In fact, there's a whole cost section toward the end!
You're right. Peanuts are the cheapest, coming in at around 15c in the US for a 15g handful. Almonds, walnuts acorns and cashews take the next spots.
As for animal vs plant saturated fats, the data clearly shows worse health outcomes for animal vs plant saturated fats, same with monounsaturates.
As you know better than most, saturated fats are a family of fats rather than a single entity. The profile of saturated fats from animals vs plants is different. For example, just looking at carbon chain length, that for beef fat skews to longer chain length [BC10-03.%, C12-0.2%, C14-6.5%, C15-1.2%, C16-57.3%, C17-2.8%, C18-31.2%, C20-0.2%, C24-0.3%] compare to coconut oil [C6-0.6%, C8-8.2%, C10-6.5%, C12-50.7%, C14-20.2%, C16-10.5%, C18-3.1%, C20-0.1%].
This matters because chain length affects how these fats behave in the body.
Medium-chain saturated fats (like lauric, caprylic, and capric acid - abundant in coconut oil) are absorbed directly via the portal vein and rapidly oxidised in the liver. They’re less likely to be incorporated into lipoproteins like LDL, and more likely to be used as energy.
Long-chain saturated fats (especially palmitic and stearic acids - dominant in beef fat) are absorbed via the lymphatic system, packaged into chylomicrons, and tend to raise LDL cholesterol more significantly. Palmitic acid in particular has been shown to increase hepatic lipogenesis and insulin resistance in metabolic studies.
The epidemiological data backs this up: diets rich in animal-derived saturated fats are consistently associated with higher rates of cardiovascular disease and all-cause mortality compared to diets high in plant-based saturated fats, even after adjusting for confounders.
So, while “saturated” is a useful shorthand, the source and structure of the fatty acid matter, and our bodies seem to respond differently depending on the profile.
I suspect there's also an issue in the way we typically cook meat vs plants, especially in Western countries. Meat is typically grilled, roasted or fried, processes that create a variety of toxic compounds like polycyclic aromatic hydrocarbons, heterocyclic amines and lipid oxidation products, all of which are associated with poorer health outcomes. Plants tend to be boiled or steamed, methods that don't produce these toxic compounds.
Well, there are a whole range of different saturated fats. It’s easy to get lost down the rabbit hole of specific fats and their properties, however, study after study shows that those consuming more saturated fats of animal origin have worse health outcomes, including higher all-cause mortality rates.
While that suggests animal saturated fats may be bad for us, that’s never been proven. It’s likely that those who consume the most animal saturated fats also consume more of the toxic products we produce when we grill, fry or roast meat, and that may be an underlying cause. It’s frustrating how much there still is to understand.
As for which fatty acids are better for us, we don’t get to pick the fatty acids directly, say palmitic or stearic acid since they come mixed up with other constituents of meat, dairy or vegetables. Whatever the underlying cause, the data is clear that those who get more of their fat (and calories) from plants rather than animals typically have better health outcomes and live longer.
Most of the data is from people eating a Western diet. With an Asian diet things may be different, in part because of the way foods are cooked and served.
For example, consuming whole eggs is associated with poorer health outcomes in the US but not in Asia. In the US, eggs are frequently fried and served with processed meats like bacon or sausages. In many Asian cuisines, eggs are more likely to be poached in broth and served with vegetables. It may be the differing health effects are nothing to do with the eggs after all!
I feel that on the point of saturated fatty acids ( SFA), the comparison between coconut oil and beef tallow is inappropriate. Coconut oil ( and Palm kernel oil) are somewhat unique category, they are lauric acid oils - rich in lauric ( C 12) and Myrtistic (C 14) oils - not largely present in the saturated content of many common oils and fats. C 16 (palmitic) and C 18 (stearic) are more bench mark fatty acids, present in the saturated content of all common oils and fats, particularly stearic acid. So, a better comparison for Beef Tallow ( 32% saturates split majorly as 2% of C14, 18% of C16 and 10% of C 18) would be something like Cottonseed oil ( over all 23 % C 16 and 1% C 18). To complete the comparison, beef tallow is 38/20 and cottonseed 48/23 respectively of w6/w9. Even the saturate content content of nuts, the context here, is predominantly stearic and palmitic. So, the individual C length split of saturates in beef tallow is not unique, could be found in the saturate content of many common plant and animal fats. I don’t see any reason, therefore, to put down the saturate content of animal fats, while giving the benefit of doubt to the saturate content of common plant oils. Animal fats - saturated FA - cardiovascular linked studies appear to be common. Are there any studies on saturates from vegetable sources versus cardiovascular condition - to show that they are not gravely detrimental ? The point I want to make is that almost all studies based on cardiovascular health versus saturated fats could be based on animal fats. As you said, the problem could be the common high temperature processing of meat in kitchens and we don’t know the role of heat damaged meat proteins in the overall cardiovascular considerations, their fat content is the overtly visible culprit.
Sorry, I overlooked your cost comparison of various nuts. Like India, USA too is a major peanut growing country, but why the nuts cost so much. Here, raw groundnut kernels are the equivalents of about $3 a kg, the processed, branded, small pack ones about 5 $ a kg, max.
Thanks again, Moro. I always appreciate your thoughtful lens on lipid chemistry.
On the beef vs coconut fat point: I agree it’s not ideal to conflate all saturated fats — but that’s exactly why I think comparing coconut oil and beef tallow is useful. While both contain palmitic (C16:0) and stearic (C18:0), their overall fatty acid profiles are clearly distinct.
So while both contain some C16 and C18, beef fat is overwhelmingly long-chain, while coconut is dominated by medium-chain SFAs (lauric, myristic). These are metabolised differently, and likely have different health impacts. I think it’s fair to say their “saturated” labels hide meaningful differences.
On the broader issue: I completely agree that many studies linking saturated fat to CVD risk are based on animal fats. However, I don’t see this as evidence that plant-based saturated fats are harmless - just that they’re understudied.
In fact, I noted this in an earlier piece on coconut and palm oils. While coconut oil raises both LDL and HDL, the net effect on cardiovascular outcomes remains unknown. And the best data we have on palm oil (on a population level) suggest potential harms, particularly in developing countries where intake has risen steeply.
So I don’t argue that plant saturated fats are “safe.” I argue that, in the context of modern diets, animal-based saturated fats have a stronger and more consistent link to cardiovascular risk, while the jury is still out on the plant side.
In the meantime, given the lack of evidence for benefit, and the growing consumption of palm oil via ultra-processed foods, caution seems warranted.
As ever, I appreciate you keeping the bar high and the definitions clear.
In times to come, we will get to know that no fat is unsafe, no component of any fat is unsafe. Moderation is the key.
Coconut being such a healthy thing, befitting its name, how to consume it really as a nut. In the normally known and consumed coconut products, like oil, tender coconut, coconut milk etc, even coconut cookies, we don’t get to consume the whole nut, or the flesh of the nut. There is an option practised in the cuisines of South India, but used only in limited ways, because we use freshly grated coconut almost daily. This form of coconut is normally eaten in blended light snacks. Here is how, it is done in households.
This is from copra, the dry form of coconut, the feed stock for oil extraction. It is a general grocery item for us and used in some of our recipes, usually after grating it. Typically, take a half copra, using a potato slicer, get slices of it around the circumference, thin slices. If the slices break, it doesn’t matter, we eventually need one-two inch lengths only. It could be about 1 cm width, the width of the flush in the half copra. You can eat them, very chewy, oily, because the dry copra has 60% oil content, the rest could be protein plus carbohydrates. This chewing is very relieving for mouth ulcers. Now, lightly dry roast these pieces in a pan, to light browning. This is ready to eat, toasted but version of coconut, less chewy, a bit crunchy. This will keep well for a few weeks in normal storage at about 25 C. Anyway, coconut oil doesn’t get rancid that easily. Along with your 15 gm of recommend daily nuts, you can add a few pieces of this, or some 3 gms in the 15 gms serving of any nut. You will get all the health benefits of coconut through this.
You can get copra in Indian grocery stores or in Indian section of large supermarkets. Please give it a try once and let me know how you feel.
Thanks, Moro, your sliced, toasted copra sounds absolutely delicious. I’m sure it’s far richer and more satisfying than the closest Western equivalent: coconut flakes.
For readers unfamiliar with copra, it’s essentially the dried flesh of mature coconuts, traditionally used to make coconut oil. But when it’s thinly sliced and toasted, as Moro describes, it becomes a wonderfully chewy, flavour-packed snack.
How does it differ from standard coconut flakes?
- Copra is oil-rich, dense, and intensely coconutty — because it hasn’t been defatted or overly processed.
- Coconut flakes, by contrast, are usually lower in fat, milder in taste, and often sweetened or sulfited for baking, rather than snacking.
Think of it this way, copra is to coconut flakes what dark chocolate is to white chocolate: deeper, richer, closer to the source.
I don't hold out much hope of finding any in any store near me, but I'm going to be looking out for it now!
I don’t know about the coconut flakes available there. They could be what they have described. Dry toast them a bit in a flat pan and see if it enhances the flavour and taste. The flavour and the slight sweetishness come from toasting and of course it is also unmolested whole coconut flesh in dry form. These flakes, if they are finer or smaller, will not suit as a nut analogue. You need thin slices. Decades ago, while living in England, we used to get desiccated coconut ( coarse powder) which we could use in place of grated fresh coconut, which is the mainstay in our cooking.
We also make sometimes these toasted copra flakes at home for certain blended dry snack preparations. Look for whole copra or half round. The inner surface won’t be pure white, slightly yellowish. It should be available in large Indian grocery stores or possibly on line these days. Though some cultures in S. E.Asia like Thailand use coconut in their recipes, mostly they are fresh. Philippine is perhaps the only other country producing and using Coconut oil than India, Sri Lanka and a couple of Polynesian Island countries. So they must be familiar with Copra, but I don’t know how far their cuisines use copra in their preparations. They may not be aware of these toasted copra flakes. Properly done, it is both munchy and crispy, of course nutty. You are perhaps hearing first time as a snacking nut. It must be used only as a complimenting nut component in small quantities with other nuts. Its oil composition makes it unique among the commonly consumed nuts.
I just casually made an interesting observation in the kitchen and immediately looked into my notes of oil composition of various nuts. We had a small lot of 11 months old red walnut kernels brought from the USA last year, as also some pecan nuts and almonds of similar age. While the latter two still taste good, I noticed a small rancid flavour in the walnut kernels. All three are high oil content nuts. The PUFA content seems to be the reason. Almond oil has 26% w-6 and 67% w-9. Pecan oil has about 36.5 % of w-6, 1.5% of w-3 and 52% of w-9. Walnut oil has 58% w-6, 13% w-3 and 22% w-9. While walnuts are commended because of the w-3 content among the common nuts, it is part of the PUFA content of total 71%, liable for rancid on aging. It’s high w-3 could be an accelerant for this aging phenomenon. Even though, the other two have medium to medium high levels of PUFA, their high MUFA ( w-9) could be an aging stabiliser. I think peanuts and cashew nuts also stay well with aging.
Great article, Ben! I really appreciated your detailed comparisons, although I wonder if it is fair to include peanuts, a legume, against the tree nuts. Technically, pistachios and cashews are seeds, rather than nuts. BTW, the way that cashews grow is really fascinating.
You might want to say something about the oxalate content of the species you've mentioned, since renal calculi can be a problem for susceptible individuals. I'm thinking that calcium citrate or dairy products taken at the same time might help to mitigate that risk.
Personally, I eat lots of peanuts, a Brazil nut a day, and ample quantities of walnuts, pecans, and almonds. Hazelnuts (aka filberts) are a seasonal treat where I live. I was never fond of the oily macadamia nuts when I lived in Hawaii, but I frequently enjoy pistachios and cashews.
I forgot to mention earlier, Mick, when I worked in West Africa, some local friends had a small grove of cashew trees. One day they showed me how they got the nuts to eat.
They’d pull the seed-containing pods from the fruit and put them in an open can which they’d then heat over a fire. Once it got hot enough, the oils in the cashew pod would catch fire. It would go off like a rocket engine! They delighted in my reaction the first time I saw it. It was truly spectacular.
Then they’d tip out the ‘roasted’ (more like fire-bombed) pods and pull out the cashew nuts which would be warm, smoky and delicious!
Incidentally, the cashew fruit was then used to brew a very potent hooch!
On a less light-hearted note, commercial cashew growing is associated with a lot of poor ethical practices. The workers are paid a pittance for harvesting, and the acidic sap burns their skin. Often they’re not given any protective clothing. In some countries, prisoners are forced to harvest cashews. It’s definitely worth looking for the Fairtrade logo.
Thank you, Ben. I didn’t know about the harvesting practices. From Wikipedia: The cashew seed forms at the base of the “cashew apple.” It is surrounded by a double-shell that contains anacardic acid, a potent allergen and skin irritant.
Thanks, Mick! Always a pleasure to read your thoughtful replies (and admire your snack lineup - strong choices across the board).
You’re quite right about peanuts being legumes and cashews/pistachios being seeds - which is exactly why I included a comment on nut taxonomy in the article. Here’s the bit:
“Before you remind me: yes, peanuts are legumes, not nuts… Almonds, cashews, walnuts? Technically seeds of drupes — the same botanical family as peaches and plums. Brazil nuts? Seeds, too, though from a different kind of fruit.
The only true nuts in your snack bowl? Hazelnuts and chestnuts.”
Nut pedantry brings me joy too, but if I’d titled the piece “A Comparative Nutritional Overview of Commonly Consumed Tree Nuts, Drupe Seeds, Legumes, and Other Botanically Misunderstood Edible Plant Ovaries”, I fear even the keenest readers would’ve clicked away by the third taxonomic footnote!
So I kept the title simple and snuck the botany into the body for those of us who care (read: you and me, and maybe three others!).
Excellent point about oxalates, and one that does pop up now and again in more medically focused circles.
For most people, the tiny 15g nut portion I recommend won’t deliver enough oxalate to pose any risk, especially if they have no personal history of calcium oxalate stones. Spinach, beet greens, and rhubarb are far more significant sources. That said, your note on pairing high-oxalate foods with calcium is spot-on and worth keeping in mind for stone-formers. Appreciate you raising it.
As you know, One Health Tweak a Week is designed to be a five-minute, practical health nudge for busy readers - not a full systematic review. There’s always more detail I could dive into (and often do elsewhere), but part of the challenge I set myself here is deciding what not to include. Oxalates just didn’t make the cut this time. But I’m glad the comments give us space to explore them!
Sounds like you should get a can of mixed nuts. (LOL) My post from a couple days ago shared a study on diverticulitis risk from seeds, nuts and corn. Observational and self reporting but does hint that this may not be a risk for diverticulitis. Maybe this challenge the old idea about the risk but note: "All participants were free of inflammatory bowel disease, diverticulitis, and cancer at baseline." Best check with your health care provider. Was surprised that Macadamia was low in carbs though high in calories?
I completely agree. Eating a variety of nuts is likely to offer the widest benefits, just like eating a wide variety of vegetables.
That old chestnut that nuts can cause diverticulitis has indeed been debunked. Multiple large, prospective studies found that eating nuts does not increase the risk of diverticulitis or diverticular bleeding. In fact, higher nut and popcorn consumption was linked to a lower risk of diverticulitis, with hazard ratios suggesting a protective effect for those consuming nuts at least twice per week compared to those eating them less than once per month.
As for macadamias, they're unusually high in total fat (~75% by weight), most of it monounsaturated, which explains their calorie density. Since fat yields 9 kcal/g, compared to 4 kcal/g for carbs and protein, a food that's low in carbs but high in fat can be especially calorically dense.
In fact, their low carbohydrate profile (~4g/100g) is part of what makes macadamias attractive for those following low-carb or ketogenic diets. So while they’re energy-dense, the metabolic effects are quite different from, say, high-carb snacks with the same calorie count.
I have at least one handful of nuts daily. My favorites for snacking are pistachios, cashews, almonds, and peanuts. For cooking it's walnuts, pecans, cashews and hazelnuts.
That's a great variety of nuts, David, and I'm sure that offers the widest range of health benefits. No Brazil nuts to top up your selenium?
Selenium levels in soil, and thus in local crops and livestock vary geographically across the US as you know. They're low in the Northeast, Pacific Northwest and parts of Florida, Michigan and the Ohio River Valley. Conversely, the Great Plains and some parts of the Midwest and Mountain West have high levels. Of course, this depends on how much food is sourced locally. It's one of those rare occasions where you're less likely to be deficient if you get all your food from national supermarkets rather than farmers' markets!
It was a very exhaustive response. I am glad that you could find so much time to frame it. I am also learning a lot through this conversation. I agree that all the factors, the number of double bonds, their positions in the chain etc seem important on present understandings. Perhaps the w-3 position is more important. But this is not a closed topic for research and we will have to wait for what future work would reveal.
On linoleic acid on the Indian urban scene, sunflower oil came into the picture about 40 years ago ( when I was in my first job, in edible oil industry), and has gained a dominant position here. There is lot of sunflower cultivation and lot of oil imported too. That has flipped the scene for the worse. There is lot of rice bran oil use too in the same decades. There was perhaps no alternatives for the burgeoning population. The need to limit w-6 consumption in the population has not registered in the minds of food authorities here. There was one large company here that began promoting safflower oil and its blends, under an umbrella brand of Saffola.,about 20-25 years ago. Their tag line was it was a very heart friendly oil and many sub brands were created under the cardio name. These were being sold at a premium and there was enough patronage in the rich urban markets. About 10 years ago, I noticed that they began withdrawing Saffola oil, both as pure oil and in their blends. While retaining their parent brand and sub brand names, their products were getting based on rice bran oil, soy bean oil and to a small extent on sunflower oil. I don’t know if this was a commercial decision or a health decision, the public had no idea about these changes and they continued to give a premium. But in the shelves here, I find sunflower oil is the most abundant one. The long term implications have not sunk in the minds of any one including the fact that we have practically no options for w-3.
It’s fascinating to hear the history of edible oils in India, and how sunflower and rice bran oils have become so dominant.
Here in the UK, the story is a little different. For decades rapeseed oil (canola) was widely used, particularly in margarines and salad dressings, but more recently I’ve noticed ingredients lists are vaguer, often just “vegetable oils including palm.” That gives manufacturers freedom to switch blends depending on price and availability, and leaves consumers with no idea whether they’re mostly getting omega-6-rich oils or something with a bit more omega-3 like rapeseed.
I seem to remember that mustard seed oil is quite popular in parts of India and often marketed as “heart healthy.” In Western countries, however, pure mustard oil is banned for human consumption because of its erucic acid content. At high intakes, erucic acid (a monounsaturated omega-9, 22:1 n-9) has been linked in animal studies to myocardial lipidosis (fat deposition in the heart muscle) and possible cardiotoxicity. That’s why in the EU, US, and UK, mustard oil is legally sold “for external use only.” Rapeseed/canola was actually bred to be low in erucic acid (“double zero” varieties), precisely to get around that concern.
Here also the labelling is vague. Often they say ‘vegetable oil’. Except for a rare hot head like me, I don’t think any one cares. Many people here are sensitive about palm oil ( in their imagination, not physically), so some manufacturers put ‘No Palm Oil used”, to milk the irrelevant sentiments of people. I have just written a letter to some leading newspapers that the Govt must stop this practice. Clinically and nutritionally there is nothing wrong about Palm oil, in fact being an healthier option to Sunflower oil. India began importing refined palmolein from Malaysia about 60 years ago, when it wasn’t self sufficient in cooking oils. For some decades, this used to be sold at very low price via public distribution shops, ration shops as we call it, I think the practice continues. But it is sold in the open market also. In my large group of family and friends, I always advocate peanut oil or palm oil for every day cooking. Yes, north India predominantly uses classical unrefined mustard oil from times unknown. Yes, some of us know about the erucic acid story, but the large users have remained unfazed for generations.
This brings me to another point. Many studies, including the one on mustard oil, are normally done on populations for 2, 5, 20 years. I feel many of these effects tend to stabilise to safer levels over generations, time scale needed for metabolism to absorb the changes. For example, the seed oil story in the West, in the USA in particular is less than 100 years, before that it was all animal fat only for generations. The metabolism got used to saturated fats. Suddenly unsaturated fats became an issue, because the metabolism was not quickly ready. This is the cause for the rise of life style diseases, purely from food or edible oil perspective. Over several centuries, this smoothens out in populations. We are used to w-6 oils, but even for us the high LA sunflower oil is a big shift. It will be a few generations before the population level metabolisms stabilise and the negative impacts disappear. This could be true for adaption of any new or different food ingredients. On this premise, a large section of world population could be undergoing a metabolic churn now, because increased international travel, trade and commerce in the last 200 hundred years had brought about an intermingling or new exposures in the cuisines, that is mostly in the form of food ingredients and the process is still on. And then there is this industrial and processed food everywhere. If this process stays at where we are now, in another 200 years, the metabolisms would have stabilised against the changes from the past, life style diseases might slowly disappear. We should be wary of other external impacts, like climate changes.
I was raising this point in the context of a story by someone else on how the perceptions have changed over saturated fats. It was mentioned how the Intuits in arctic Canada or the Masai maras in central Africa, despite their total dependence on hunted/reared meat and thus saturated fats have not been very prone to cardiac diseases. My point was that their metabolisms got adjusted over centuries to this food available to them in their midst and the likely negative impacts were smoothened out.
I like how you’ve pulled this conversation towards the bigger evolutionary picture. You’re right: populations can and do adapt over time to new foods. The classic cases are things like lactase persistence in dairying populations, or the spread of extra amylase genes in high-starch farming cultures - but those shifts took thousands of years and dozens of generations.
The modern story of seed oils is much younger - under 100 years in most countries - so there hasn’t been time for genetic adaptation. What we do see is metabolic strain: obesity, diabetes, and CVD risk rising as high-LA oils and ultra-processed foods displace traditional diets. That’s less an adaptation process, and more a case of metabolic plasticity being pushed to its limits.
The Inuit and Masai are interesting examples too. Their apparent resilience to high saturated fat seems to come not from genetic rewiring in a few centuries, but from lifestyle context: high activity, low sugar, low processed food. Once those populations shift to modern diets, their rates of chronic disease rise rapidly, which shows the “protection” wasn’t permanent adaptation.
So I’d agree with you that we may see gradual smoothing out over many centuries, but in the near term, I believe the safest stance is still to limit excess omega-6 and ultra-processed oils, and to rebalance with more omega-3 where possible.
It is a pleasure seeing your engaging views. What I meant by speeding up ( of metabolic equilibrium) is the kind of things you have mentioned in the end. But seed oils are here to stay, even in North America. The traditional fats cannot generate enough volumes for the risen population in these decades. Moreover, the seeds (corn, soybean, cotton) let the American agriculture prosper big in a century and the story cannot be reversed. Oils from these cultivation actually boost the economics of agriculture substantially. It is a virtuoso cycle in a sense and it doesn’t make sense breaking it. There are two ways in which this seed oil conundrum can be turned upside. The first is to turn to peanut oil, canola oil and high oleic sunflower oil for major every day uses along with dairy and animal fats. And processed food industries must be directed to use them by decree, in place of soybean/corn/cotton seed oils. These measures can be adopted on the go and this will substantially limit LA consumption. The second one is the big oil industries like P&G, General Mills, Levers ( UK and Europe) must be asked to research and develop directed hydrogenation that will convert these old school seed oils to oleic acid rich oils, similar to olive oil. The catalyst development is the key. Though I left this industry long ago, my guess is that catalyst developments have taken place that will avoid trans acids formation, but the end point is still general, only iodine value, which would mean converting part of polyunsaturation into saturation, leaving some PUFA still. I feel this will be a very critical technical development, should it happen. At once, huge volumes of problematic seed oils can become safe MUFA rich oils, very affordable too. But who will nudge these oil giants ? Food authorities world over do not have people who can think and direct on those lines.
Animal farming , already facing global pollution and warming labels , cannot expand massively to meet full scale needs of dairy and animal fats, on present volumes. That is a limitation to returning to the past. So, the world, North America in particular, has to work around seed oils.
This is an historical point about the Indian food scene. Wheat in North India and rice in the south have been the dominant grains for millennia. But millets also had a strong place traditionally. Despite some inroads by these two mainstream grains in rural India, millets continue to be a choice there. In urban India, for a few decades in the second half of last century, millets had lost their flavour, but are making a return, as health food and backed by some promotion by the government boards. I feel this could be a permanent trend. This small flip flop at population level should not matter from the larger metabolism angle.
You’re absolutely right: seed oils aren’t just a nutritional question, they’re an economic and agricultural reality. They’re deeply embedded in North American farming, and it’s unrealistic (and undesirable) to imagine simply “going back” to animal and dairy fats to feed large populations.
The healthier path is exactly what you suggest: shifting the balance of fats towards MUFA-rich oils (like peanut, canola/rapeseed, and high-oleic sunflower) while keeping appropriate amounts of saturates and polyunsaturates. That pattern aligns much more closely with the evidence we have from Mediterranean-style diets and other long-lived populations.
On your second idea - as you say, it's technically feasible to convert omega-6–rich oils into MUFA-rich ones. The most practical approach is through crop breeding and biotechnology, which have already given us high-oleic variants of sunflower, safflower, soybean, and canola. These oils can be 70–80% oleic acid with much less linoleic acid, and they’re already used in some food manufacturing because they’re more stable for frying and have a longer shelf life. Directed hydrogenation is another route, but while modern catalysts may reduce the risk of trans fat formation, it’s less elegant and less widely adopted than breeding.
So in practice, I think the big opportunity is encouraging the food industry to switch from commodity high-LA oils to high-oleic oils - something that can be done at scale, is economically feasible, and would immediately improve the fatty acid profile of processed foods.
Your point about millets is a hopeful one: it shows that traditional staples can make a comeback when health, culture, and policy align. Perhaps the same can be done for fats - not by going backwards, but by steering the market towards healthier options and maybe engineering smarter oils.
We can look at this food evolutionary picture from another angle. Carbohydrates, Proteins and Fats are the three food components that are key to the metabolic state in populations. So any big change even in one of them by a population group is going to have a major health impact in the near term - a few decades atleast in the context of what we are discussing. Historically, population groups that we have today and that have existed for millennia around the world have not changed their choices of these three much over millennia with a recent origin exception or two on fats. Large population groups that traditionally lived on animal and dairy fats have to accommodate seed oils. Geographically these are North American populations mainly, possibly some in South America. Much of elsewhere, where a mix of fats has reigned, have also not seen big changes in their fat profile, again except for the invasion of sunflower oil in their consumption pattern. This could be a much smaller metabolic churn compared to the North American story - the switch in predominance from animal/dairy fats to the seed oils. This is a big change metabolically too and that is the reason why this region is the one facing maximum incidence of life style diseases. I don’t know if this geography realises the implications of the choice before them, but with some clever decisions, they can speed up the process of metabolic stability.
The largest intermingling has taken place in the baskets of fruits, vegetables, nuts etc that contribute to the needs of vitamins and minerals largely. Even here, I would say that two thirds of these lot are now common world over and their contribution to the metabolic instability picture in populations is not significant.
I think you’ve framed it very clearly: when whole populations shift one of the “big three” macronutrients, the effects are bound to be disruptive. The North American fat story is a striking example - in less than a century, a dietary base built on animal and dairy fat flipped to one dominated by seed oils, especially soy and corn. That’s a profound change in both the quality of fat (saturated vs high-LA polyunsaturated) and the quantity (as industrial oils made it into almost every processed food).
As you say, that’s a very different kind of metabolic shock than broadening the fruit and vegetable basket, which humans have always adapted to quite easily.
I’d agree this helps explain why the “lifestyle disease” burden is so high in North America: the scale and speed of the fat transition is historically unprecedented. Of course, it hasn’t happened in isolation - the rise of ultra-processed foods (where those oils are combined with refined starches, sugars, and additives) has amplified the metabolic disruption. Other regions with a more mixed fat intake (animal + plant oils) may face a gentler adjustment, though the sunflower oil boom is still significant.
Where I’d add a note of caution is on the idea of “speeding up” metabolic stability. Our biology does have a degree of plasticity - enzymes can adjust within a lifetime, and the microbiome shifts within weeks - but true genetic adaptation is much slower, usually over many centuries. Which means public health levers (shifting the n-6:n-3 balance, limiting industrial oils, encouraging traditional fats where appropriate) may be more effective than waiting for evolution to catch up.
I read recently about the health profile of pine nuts in another presentation, rechecked yours of this and found that you haven’t included it in your discussion here. The nuts have an oil content of over 70%. Even though the oil is rich in linoleic acid, close to 50%, the speciality of the oil seems to be the 18-20% of Pinolenic acid, a tri unsaturated w-6 (C 18, 5,9,12 double bond positions) which seems to have significant anti inflammatory properties. The oil also contains upto 5% of amino acids ( free amino acids or low mol wt peptides ?) which is another significant factor of its health quotient.
The important thing here is a w-6 being so beneficial, pinolenic acid, is it because it is triunsaturated), similar to gamma Linolenic acid ( GLA, 6,9,12). What is the mechanism here ? ALA is another outlier, a linolenic analogue, but with a different double bond position (12,15) and thus a w-3. Which is important, being w-3 regardless of the number of double bonds OR multiple double bonds, 3 or more, regardless of their position ?
Isn’t it ironic that the most abundant fatty acid type in the triglyceride world (oils and fats), Linoleic acid, double unsaturated and w-6, is now regarded as the least desirable molecule ? Being just w-6 is not a downside, but the number of double bonds is. Would the Linoleic acid story also come to pass in future ? That would eventually be the irony of reality, because it is the most abundantly occuring fatty acid in nature in triglyceride oils and fats.
Moro - brilliant observations, as always! You're expert on oil chemistry, not me, so I'll explain my understanding, but let me know if I get the chemistry wrong.
You’re right that pine nuts are unusual: their oil carries ~15–20% pinolenic acid (PNLA), an omega-6 triene (18:3 n-6, Δ5,9,12). PNLA isn’t just “another omega-6” - the double-bond pattern changes how enzymes handle it.
Mechanistically, PNLA has been shown to raise satiety hormones (CCK, GLP-1), improve hepatic LDL receptor activity, and dampen pro-inflammatory cytokines in cell and animal models. It also acts as a dual agonist at FFA1/FFA4 receptors, which may explain the metabolic and anti-diabetic signals seen in preclinical studies. Evidence in humans is still modest - small appetite and lipid studies, mostly - but intriguing enough to warrant more work.
On your broader question, does “omega-6” vs “omega-3,” bond count, or bond position matter most? I’d argue it’s all three:
• Family (n-3 vs n-6): determines which eicosanoids and mediators you can make, and which enzymes compete for substrates.
• Number of double bonds: influences membrane fluidity and susceptibility to oxidation.
• Position (Δ5 vs Δ6, etc.): alters how desaturases/elongases handle the molecule, and therefore which downstream lipids you end up with.
That’s why GLA (18:3 n-6, Δ6,9,12) and PNLA (Δ5,9,12) behave differently, despite both being omega-6 trienes. And why ALA (18:3 n-3, Δ9,12,15) is in another family altogether, with different fates.
On linoleic acid (LA, 18:2 n-6, Δ9,12): I don’t think it’s fair to call LA inherently undesirable. Substituting LA for saturated fat lowers LDL and reduces cardiovascular risk, and most RCTs don’t show LA raising inflammatory markers in healthy people. Where the concern arises, and I share it, is in the Western diet (and now urban Indian diet) context: very high LA intake from seed oils (often >10% of energy) combined with very low omega-3 intake. That imbalance means omega-6 dominates the shared metabolic pathways, tilting the system towards pro-inflammatory arachidonic acid-derived mediators. So while the molecule itself isn’t “bad,” in practice the context is.
Bottom line: pine nuts are a solid addition to a varied nut intake, but their high omega-6 content makes me wary of suggesting excess use. PNLA makes them especially interesting, but the best health signal still comes from a balance of mixed tree nuts and from keeping omega-3 intake up (ideally from oily fish) while not letting omega-6 dominate the plate.
As a lifelong nut lover, I’m truly grateful for this fantastic nutritional guide to eating nuts. (I eat too much.) And I will share it with friends & family members. Thank you! [h/t to Dr. Chawla because i read your comment on her June 27th post!] 🙏🏽
You’re right, Steve, I didn’t mention ‘activating’ nuts. While it’s discusssed in wellness circles, there’s no real evidence that soaking and dehydrating nuts significantly improves their nutritional value or digestibility.
You do right to eat all of them! I’m sure eating a wide variety maximises the health benefits, so you won’t miss the peanuts. I must admit though, I find peanuts irresistible!
Thank you. That is important information about Brazil nuts. It is easy to eat more than one a day. I blender cashews with coffee.
It's surprising, isn't it. Brazil nuts are unique in their ability to concentrate selenium. They absorb huge amounts of selenium partly because their deep roots reach selenium-rich soil layers in the Amazon. But there’s likely an evolutionary edge, too: high selenium concentrations in their seeds and tissues may help defend against fungi, insects, and over-hungry animals. I suspect few people are aware of the potential toxicity.
You blend cashews with coffee.. There's a combination I wouldn't have expected. Is this hot or iced coffee? What's it like?
So interesting about why they are so concentrated in selenium. Many nut mixes have included them.
About 10 raw cashews in hot coffee. It just tastes good and is frothy.
Thanks for this outstanding guide!
Thank you for those kind words! I’m so pleased you found the article helpful.
Glad you mention toward the end the preference for raw or lightly roasted nuts. But I would think there is a very large difference between nuts eaten raw and nuts that are highly salted & roasted in seed oils, higher both in calories and omega 6. How much higher? That would be good to know.
Hi Paul, you're right. What happens to the nuts between their leaving the plant and entering our mouths matters a lot.
You ask a good question about how much the nutritional data changes. Here's what the USDA data says:
Raw peanuts: Fat - 43g/100g, calories - 588, omega-6 - 17g, sodium - 22mg
Roasted & salted peanuts: Fat 49g, calories - 579, omega-6 - 17g, sodium - 433mg
Raw cashews: Fat - 44g, calories - 553, omega-6 - 7.8g. sodium - 12mg
Roasted & salted cashews: Fat - 48g, calories - 581, omega-6 - 8.5g, sodium - 308mg
Interesting results. Based on these limited samples:
- Fat content increases by about 10%
- Calories don't change much (-2% and +9%)
- Omega-6 doesn't change much (0% and +9%)
- Sodium increases hugely (1870% and 2470%)
I must admit I expected fat and calories to increase more.
Are the findings in line with what you had expected, or are you surprised?
Thank you, Ben! I would also have thought calories and omega-6 would be more significant. So that's rather reassuring. But the salt: holy cow!
As an Oregonian who grows hazelnuts, I say “eat more hazelnuts”! But also, two little-known facts about this true nut. First, the new varieties of hazelnuts (aka filberts) taste different than the nuts grown a generation ago; you might like them better—I know I do! Second, I volunteer at a wildlife rescue/rehab clinic, where I was thrilled to learn that the macro- and micronutrients in hazelnuts make them the single BEST nut to feed to baby squirrels! 🐿️ Not all nuts are created equal, even to our squirrel friends.
I didn’t know about new hazelnut/filbert varieties. How interesting.
I must admit, I do find them a bit dry, but there’s no doubting their excellent nutritional profile.
If they’re good enough for squirrels, the ultimate nut connoisseurs, they’re certainly good enough for me!
Ozark chinquapin chestnuts are a true chestnut tree and are the only high protein chestnut.
That’s a new one for me, Conrad! I’ve never heard of them before, so I had to look them up.
They look to be impressive trees. Do they taste the same as regular chestnuts?
I have only tasted them twice, but I can confirm they are highly likely to be the tastiest chestnut on Earth, as American Chestnuts were historically reputably far superior to Eurasian. To me, they were like a slightly sweeter and more flavorful chestnut, but Dunstan is the only other chestnut of which I can recall the flavor. I’m glad you looked them up, the Ozark Chinquapin Foundation has created blight-resistant non-GMO non-transgenic trees, with natural novel breeding techniques, and they’re available as permitting to Americans for a small annual donation. Last year I had lab tests done on the southern subspecies (South of the Arkansas River), and those confirmed the reported high protein, but had traditional chestnut fat levels! Making it the only low-fat, high-protein tree nut. Those nutrition tests were a very small sample size, but from two labs, and supplies are so rare it will probably be a few years before I can do additional tests.
You haven’t mentioned about the cost. It is a factor for a majority of populations. In terms of balanced nutrition at affordable cost, peanuts should rank at the top. Highest protein content among the profiled lot, good oil content with good quality oil, quite rich in w-9 and lower levels of w-6.
Why should saturated fatty acids from animal fats be problematic compared to those from plant oils ? Body metabolism will see them as only palmitic and stearic acids or the triglycerides, regardless of their origin.
Hey, Moro. In fact, there's a whole cost section toward the end!
You're right. Peanuts are the cheapest, coming in at around 15c in the US for a 15g handful. Almonds, walnuts acorns and cashews take the next spots.
As for animal vs plant saturated fats, the data clearly shows worse health outcomes for animal vs plant saturated fats, same with monounsaturates.
As you know better than most, saturated fats are a family of fats rather than a single entity. The profile of saturated fats from animals vs plants is different. For example, just looking at carbon chain length, that for beef fat skews to longer chain length [BC10-03.%, C12-0.2%, C14-6.5%, C15-1.2%, C16-57.3%, C17-2.8%, C18-31.2%, C20-0.2%, C24-0.3%] compare to coconut oil [C6-0.6%, C8-8.2%, C10-6.5%, C12-50.7%, C14-20.2%, C16-10.5%, C18-3.1%, C20-0.1%].
This matters because chain length affects how these fats behave in the body.
Medium-chain saturated fats (like lauric, caprylic, and capric acid - abundant in coconut oil) are absorbed directly via the portal vein and rapidly oxidised in the liver. They’re less likely to be incorporated into lipoproteins like LDL, and more likely to be used as energy.
Long-chain saturated fats (especially palmitic and stearic acids - dominant in beef fat) are absorbed via the lymphatic system, packaged into chylomicrons, and tend to raise LDL cholesterol more significantly. Palmitic acid in particular has been shown to increase hepatic lipogenesis and insulin resistance in metabolic studies.
The epidemiological data backs this up: diets rich in animal-derived saturated fats are consistently associated with higher rates of cardiovascular disease and all-cause mortality compared to diets high in plant-based saturated fats, even after adjusting for confounders.
So, while “saturated” is a useful shorthand, the source and structure of the fatty acid matter, and our bodies seem to respond differently depending on the profile.
I suspect there's also an issue in the way we typically cook meat vs plants, especially in Western countries. Meat is typically grilled, roasted or fried, processes that create a variety of toxic compounds like polycyclic aromatic hydrocarbons, heterocyclic amines and lipid oxidation products, all of which are associated with poorer health outcomes. Plants tend to be boiled or steamed, methods that don't produce these toxic compounds.
Wow an encyclopedia of facts about fats
Thanks, Jennifer! I’m glad you found it interesting.
I thought the question was on palmitic/stearic acids... is it true other acids r butter than these two?
Well, there are a whole range of different saturated fats. It’s easy to get lost down the rabbit hole of specific fats and their properties, however, study after study shows that those consuming more saturated fats of animal origin have worse health outcomes, including higher all-cause mortality rates.
While that suggests animal saturated fats may be bad for us, that’s never been proven. It’s likely that those who consume the most animal saturated fats also consume more of the toxic products we produce when we grill, fry or roast meat, and that may be an underlying cause. It’s frustrating how much there still is to understand.
As for which fatty acids are better for us, we don’t get to pick the fatty acids directly, say palmitic or stearic acid since they come mixed up with other constituents of meat, dairy or vegetables. Whatever the underlying cause, the data is clear that those who get more of their fat (and calories) from plants rather than animals typically have better health outcomes and live longer.
Most of the data is from people eating a Western diet. With an Asian diet things may be different, in part because of the way foods are cooked and served.
For example, consuming whole eggs is associated with poorer health outcomes in the US but not in Asia. In the US, eggs are frequently fried and served with processed meats like bacon or sausages. In many Asian cuisines, eggs are more likely to be poached in broth and served with vegetables. It may be the differing health effects are nothing to do with the eggs after all!
I feel that on the point of saturated fatty acids ( SFA), the comparison between coconut oil and beef tallow is inappropriate. Coconut oil ( and Palm kernel oil) are somewhat unique category, they are lauric acid oils - rich in lauric ( C 12) and Myrtistic (C 14) oils - not largely present in the saturated content of many common oils and fats. C 16 (palmitic) and C 18 (stearic) are more bench mark fatty acids, present in the saturated content of all common oils and fats, particularly stearic acid. So, a better comparison for Beef Tallow ( 32% saturates split majorly as 2% of C14, 18% of C16 and 10% of C 18) would be something like Cottonseed oil ( over all 23 % C 16 and 1% C 18). To complete the comparison, beef tallow is 38/20 and cottonseed 48/23 respectively of w6/w9. Even the saturate content content of nuts, the context here, is predominantly stearic and palmitic. So, the individual C length split of saturates in beef tallow is not unique, could be found in the saturate content of many common plant and animal fats. I don’t see any reason, therefore, to put down the saturate content of animal fats, while giving the benefit of doubt to the saturate content of common plant oils. Animal fats - saturated FA - cardiovascular linked studies appear to be common. Are there any studies on saturates from vegetable sources versus cardiovascular condition - to show that they are not gravely detrimental ? The point I want to make is that almost all studies based on cardiovascular health versus saturated fats could be based on animal fats. As you said, the problem could be the common high temperature processing of meat in kitchens and we don’t know the role of heat damaged meat proteins in the overall cardiovascular considerations, their fat content is the overtly visible culprit.
Sorry, I overlooked your cost comparison of various nuts. Like India, USA too is a major peanut growing country, but why the nuts cost so much. Here, raw groundnut kernels are the equivalents of about $3 a kg, the processed, branded, small pack ones about 5 $ a kg, max.
Thanks again, Moro. I always appreciate your thoughtful lens on lipid chemistry.
On the beef vs coconut fat point: I agree it’s not ideal to conflate all saturated fats — but that’s exactly why I think comparing coconut oil and beef tallow is useful. While both contain palmitic (C16:0) and stearic (C18:0), their overall fatty acid profiles are clearly distinct.
Here’s the split we used last time:
Beef fat:
C10 – 0.3%, C12 – 0.2%, C14 – 6.5%, C15 – 1.2%, C16 – 57.3%, C17 – 2.8%, C18 – 31.2%, C20 – 0.2%, C24 – 0.3%
Coconut oil:
C6 – 0.6%, C8 – 8.2%, C10 – 6.5%, C12 – 50.7%, C14 – 20.2%, C16 – 10.5%, C18 – 3.1%, C20 – 0.1%
So while both contain some C16 and C18, beef fat is overwhelmingly long-chain, while coconut is dominated by medium-chain SFAs (lauric, myristic). These are metabolised differently, and likely have different health impacts. I think it’s fair to say their “saturated” labels hide meaningful differences.
On the broader issue: I completely agree that many studies linking saturated fat to CVD risk are based on animal fats. However, I don’t see this as evidence that plant-based saturated fats are harmless - just that they’re understudied.
In fact, I noted this in an earlier piece on coconut and palm oils. While coconut oil raises both LDL and HDL, the net effect on cardiovascular outcomes remains unknown. And the best data we have on palm oil (on a population level) suggest potential harms, particularly in developing countries where intake has risen steeply.
So I don’t argue that plant saturated fats are “safe.” I argue that, in the context of modern diets, animal-based saturated fats have a stronger and more consistent link to cardiovascular risk, while the jury is still out on the plant side.
In the meantime, given the lack of evidence for benefit, and the growing consumption of palm oil via ultra-processed foods, caution seems warranted.
As ever, I appreciate you keeping the bar high and the definitions clear.
In times to come, we will get to know that no fat is unsafe, no component of any fat is unsafe. Moderation is the key.
Coconut being such a healthy thing, befitting its name, how to consume it really as a nut. In the normally known and consumed coconut products, like oil, tender coconut, coconut milk etc, even coconut cookies, we don’t get to consume the whole nut, or the flesh of the nut. There is an option practised in the cuisines of South India, but used only in limited ways, because we use freshly grated coconut almost daily. This form of coconut is normally eaten in blended light snacks. Here is how, it is done in households.
This is from copra, the dry form of coconut, the feed stock for oil extraction. It is a general grocery item for us and used in some of our recipes, usually after grating it. Typically, take a half copra, using a potato slicer, get slices of it around the circumference, thin slices. If the slices break, it doesn’t matter, we eventually need one-two inch lengths only. It could be about 1 cm width, the width of the flush in the half copra. You can eat them, very chewy, oily, because the dry copra has 60% oil content, the rest could be protein plus carbohydrates. This chewing is very relieving for mouth ulcers. Now, lightly dry roast these pieces in a pan, to light browning. This is ready to eat, toasted but version of coconut, less chewy, a bit crunchy. This will keep well for a few weeks in normal storage at about 25 C. Anyway, coconut oil doesn’t get rancid that easily. Along with your 15 gm of recommend daily nuts, you can add a few pieces of this, or some 3 gms in the 15 gms serving of any nut. You will get all the health benefits of coconut through this.
You can get copra in Indian grocery stores or in Indian section of large supermarkets. Please give it a try once and let me know how you feel.
Thanks, Moro, your sliced, toasted copra sounds absolutely delicious. I’m sure it’s far richer and more satisfying than the closest Western equivalent: coconut flakes.
For readers unfamiliar with copra, it’s essentially the dried flesh of mature coconuts, traditionally used to make coconut oil. But when it’s thinly sliced and toasted, as Moro describes, it becomes a wonderfully chewy, flavour-packed snack.
How does it differ from standard coconut flakes?
- Copra is oil-rich, dense, and intensely coconutty — because it hasn’t been defatted or overly processed.
- Coconut flakes, by contrast, are usually lower in fat, milder in taste, and often sweetened or sulfited for baking, rather than snacking.
Think of it this way, copra is to coconut flakes what dark chocolate is to white chocolate: deeper, richer, closer to the source.
I don't hold out much hope of finding any in any store near me, but I'm going to be looking out for it now!
I don’t know about the coconut flakes available there. They could be what they have described. Dry toast them a bit in a flat pan and see if it enhances the flavour and taste. The flavour and the slight sweetishness come from toasting and of course it is also unmolested whole coconut flesh in dry form. These flakes, if they are finer or smaller, will not suit as a nut analogue. You need thin slices. Decades ago, while living in England, we used to get desiccated coconut ( coarse powder) which we could use in place of grated fresh coconut, which is the mainstay in our cooking.
We also make sometimes these toasted copra flakes at home for certain blended dry snack preparations. Look for whole copra or half round. The inner surface won’t be pure white, slightly yellowish. It should be available in large Indian grocery stores or possibly on line these days. Though some cultures in S. E.Asia like Thailand use coconut in their recipes, mostly they are fresh. Philippine is perhaps the only other country producing and using Coconut oil than India, Sri Lanka and a couple of Polynesian Island countries. So they must be familiar with Copra, but I don’t know how far their cuisines use copra in their preparations. They may not be aware of these toasted copra flakes. Properly done, it is both munchy and crispy, of course nutty. You are perhaps hearing first time as a snacking nut. It must be used only as a complimenting nut component in small quantities with other nuts. Its oil composition makes it unique among the commonly consumed nuts.
I just casually made an interesting observation in the kitchen and immediately looked into my notes of oil composition of various nuts. We had a small lot of 11 months old red walnut kernels brought from the USA last year, as also some pecan nuts and almonds of similar age. While the latter two still taste good, I noticed a small rancid flavour in the walnut kernels. All three are high oil content nuts. The PUFA content seems to be the reason. Almond oil has 26% w-6 and 67% w-9. Pecan oil has about 36.5 % of w-6, 1.5% of w-3 and 52% of w-9. Walnut oil has 58% w-6, 13% w-3 and 22% w-9. While walnuts are commended because of the w-3 content among the common nuts, it is part of the PUFA content of total 71%, liable for rancid on aging. It’s high w-3 could be an accelerant for this aging phenomenon. Even though, the other two have medium to medium high levels of PUFA, their high MUFA ( w-9) could be an aging stabiliser. I think peanuts and cashew nuts also stay well with aging.
Great article, Ben! I really appreciated your detailed comparisons, although I wonder if it is fair to include peanuts, a legume, against the tree nuts. Technically, pistachios and cashews are seeds, rather than nuts. BTW, the way that cashews grow is really fascinating.
You might want to say something about the oxalate content of the species you've mentioned, since renal calculi can be a problem for susceptible individuals. I'm thinking that calcium citrate or dairy products taken at the same time might help to mitigate that risk.
Personally, I eat lots of peanuts, a Brazil nut a day, and ample quantities of walnuts, pecans, and almonds. Hazelnuts (aka filberts) are a seasonal treat where I live. I was never fond of the oily macadamia nuts when I lived in Hawaii, but I frequently enjoy pistachios and cashews.
I forgot to mention earlier, Mick, when I worked in West Africa, some local friends had a small grove of cashew trees. One day they showed me how they got the nuts to eat.
They’d pull the seed-containing pods from the fruit and put them in an open can which they’d then heat over a fire. Once it got hot enough, the oils in the cashew pod would catch fire. It would go off like a rocket engine! They delighted in my reaction the first time I saw it. It was truly spectacular.
Then they’d tip out the ‘roasted’ (more like fire-bombed) pods and pull out the cashew nuts which would be warm, smoky and delicious!
Incidentally, the cashew fruit was then used to brew a very potent hooch!
On a less light-hearted note, commercial cashew growing is associated with a lot of poor ethical practices. The workers are paid a pittance for harvesting, and the acidic sap burns their skin. Often they’re not given any protective clothing. In some countries, prisoners are forced to harvest cashews. It’s definitely worth looking for the Fairtrade logo.
Thank you, Ben. I didn’t know about the harvesting practices. From Wikipedia: The cashew seed forms at the base of the “cashew apple.” It is surrounded by a double-shell that contains anacardic acid, a potent allergen and skin irritant.
Ah. I didn’t know the fruit was called an apple. I wonder if that makes the hooch they brewed a form of cider?! It certainly had people rolling drunk.
Thanks, Mick! Always a pleasure to read your thoughtful replies (and admire your snack lineup - strong choices across the board).
You’re quite right about peanuts being legumes and cashews/pistachios being seeds - which is exactly why I included a comment on nut taxonomy in the article. Here’s the bit:
“Before you remind me: yes, peanuts are legumes, not nuts… Almonds, cashews, walnuts? Technically seeds of drupes — the same botanical family as peaches and plums. Brazil nuts? Seeds, too, though from a different kind of fruit.
The only true nuts in your snack bowl? Hazelnuts and chestnuts.”
Nut pedantry brings me joy too, but if I’d titled the piece “A Comparative Nutritional Overview of Commonly Consumed Tree Nuts, Drupe Seeds, Legumes, and Other Botanically Misunderstood Edible Plant Ovaries”, I fear even the keenest readers would’ve clicked away by the third taxonomic footnote!
So I kept the title simple and snuck the botany into the body for those of us who care (read: you and me, and maybe three others!).
Excellent point about oxalates, and one that does pop up now and again in more medically focused circles.
For most people, the tiny 15g nut portion I recommend won’t deliver enough oxalate to pose any risk, especially if they have no personal history of calcium oxalate stones. Spinach, beet greens, and rhubarb are far more significant sources. That said, your note on pairing high-oxalate foods with calcium is spot-on and worth keeping in mind for stone-formers. Appreciate you raising it.
As you know, One Health Tweak a Week is designed to be a five-minute, practical health nudge for busy readers - not a full systematic review. There’s always more detail I could dive into (and often do elsewhere), but part of the challenge I set myself here is deciding what not to include. Oxalates just didn’t make the cut this time. But I’m glad the comments give us space to explore them!
Ben, I apologize for breezing through the taxonomy part and restating what you had already made perfectly clear.
Sounds like you should get a can of mixed nuts. (LOL) My post from a couple days ago shared a study on diverticulitis risk from seeds, nuts and corn. Observational and self reporting but does hint that this may not be a risk for diverticulitis. Maybe this challenge the old idea about the risk but note: "All participants were free of inflammatory bowel disease, diverticulitis, and cancer at baseline." Best check with your health care provider. Was surprised that Macadamia was low in carbs though high in calories?
I completely agree. Eating a variety of nuts is likely to offer the widest benefits, just like eating a wide variety of vegetables.
That old chestnut that nuts can cause diverticulitis has indeed been debunked. Multiple large, prospective studies found that eating nuts does not increase the risk of diverticulitis or diverticular bleeding. In fact, higher nut and popcorn consumption was linked to a lower risk of diverticulitis, with hazard ratios suggesting a protective effect for those consuming nuts at least twice per week compared to those eating them less than once per month.
As for macadamias, they're unusually high in total fat (~75% by weight), most of it monounsaturated, which explains their calorie density. Since fat yields 9 kcal/g, compared to 4 kcal/g for carbs and protein, a food that's low in carbs but high in fat can be especially calorically dense.
In fact, their low carbohydrate profile (~4g/100g) is part of what makes macadamias attractive for those following low-carb or ketogenic diets. So while they’re energy-dense, the metabolic effects are quite different from, say, high-carb snacks with the same calorie count.
I have at least one handful of nuts daily. My favorites for snacking are pistachios, cashews, almonds, and peanuts. For cooking it's walnuts, pecans, cashews and hazelnuts.
That's a great variety of nuts, David, and I'm sure that offers the widest range of health benefits. No Brazil nuts to top up your selenium?
Selenium levels in soil, and thus in local crops and livestock vary geographically across the US as you know. They're low in the Northeast, Pacific Northwest and parts of Florida, Michigan and the Ohio River Valley. Conversely, the Great Plains and some parts of the Midwest and Mountain West have high levels. Of course, this depends on how much food is sourced locally. It's one of those rare occasions where you're less likely to be deficient if you get all your food from national supermarkets rather than farmers' markets!
Dr Jones, have you seen this comment on pine nuts and pinolenic acid ? I am eagerly waiting for your views.
Hi Moro. I did, but I was travelling and needed some time to sit and think about your comment. You always stretch my biochemistry to its limits!
I’ve now replied directly to your comment.
It was a very exhaustive response. I am glad that you could find so much time to frame it. I am also learning a lot through this conversation. I agree that all the factors, the number of double bonds, their positions in the chain etc seem important on present understandings. Perhaps the w-3 position is more important. But this is not a closed topic for research and we will have to wait for what future work would reveal.
On linoleic acid on the Indian urban scene, sunflower oil came into the picture about 40 years ago ( when I was in my first job, in edible oil industry), and has gained a dominant position here. There is lot of sunflower cultivation and lot of oil imported too. That has flipped the scene for the worse. There is lot of rice bran oil use too in the same decades. There was perhaps no alternatives for the burgeoning population. The need to limit w-6 consumption in the population has not registered in the minds of food authorities here. There was one large company here that began promoting safflower oil and its blends, under an umbrella brand of Saffola.,about 20-25 years ago. Their tag line was it was a very heart friendly oil and many sub brands were created under the cardio name. These were being sold at a premium and there was enough patronage in the rich urban markets. About 10 years ago, I noticed that they began withdrawing Saffola oil, both as pure oil and in their blends. While retaining their parent brand and sub brand names, their products were getting based on rice bran oil, soy bean oil and to a small extent on sunflower oil. I don’t know if this was a commercial decision or a health decision, the public had no idea about these changes and they continued to give a premium. But in the shelves here, I find sunflower oil is the most abundant one. The long term implications have not sunk in the minds of any one including the fact that we have practically no options for w-3.
It’s fascinating to hear the history of edible oils in India, and how sunflower and rice bran oils have become so dominant.
Here in the UK, the story is a little different. For decades rapeseed oil (canola) was widely used, particularly in margarines and salad dressings, but more recently I’ve noticed ingredients lists are vaguer, often just “vegetable oils including palm.” That gives manufacturers freedom to switch blends depending on price and availability, and leaves consumers with no idea whether they’re mostly getting omega-6-rich oils or something with a bit more omega-3 like rapeseed.
I seem to remember that mustard seed oil is quite popular in parts of India and often marketed as “heart healthy.” In Western countries, however, pure mustard oil is banned for human consumption because of its erucic acid content. At high intakes, erucic acid (a monounsaturated omega-9, 22:1 n-9) has been linked in animal studies to myocardial lipidosis (fat deposition in the heart muscle) and possible cardiotoxicity. That’s why in the EU, US, and UK, mustard oil is legally sold “for external use only.” Rapeseed/canola was actually bred to be low in erucic acid (“double zero” varieties), precisely to get around that concern.
Here also the labelling is vague. Often they say ‘vegetable oil’. Except for a rare hot head like me, I don’t think any one cares. Many people here are sensitive about palm oil ( in their imagination, not physically), so some manufacturers put ‘No Palm Oil used”, to milk the irrelevant sentiments of people. I have just written a letter to some leading newspapers that the Govt must stop this practice. Clinically and nutritionally there is nothing wrong about Palm oil, in fact being an healthier option to Sunflower oil. India began importing refined palmolein from Malaysia about 60 years ago, when it wasn’t self sufficient in cooking oils. For some decades, this used to be sold at very low price via public distribution shops, ration shops as we call it, I think the practice continues. But it is sold in the open market also. In my large group of family and friends, I always advocate peanut oil or palm oil for every day cooking. Yes, north India predominantly uses classical unrefined mustard oil from times unknown. Yes, some of us know about the erucic acid story, but the large users have remained unfazed for generations.
This brings me to another point. Many studies, including the one on mustard oil, are normally done on populations for 2, 5, 20 years. I feel many of these effects tend to stabilise to safer levels over generations, time scale needed for metabolism to absorb the changes. For example, the seed oil story in the West, in the USA in particular is less than 100 years, before that it was all animal fat only for generations. The metabolism got used to saturated fats. Suddenly unsaturated fats became an issue, because the metabolism was not quickly ready. This is the cause for the rise of life style diseases, purely from food or edible oil perspective. Over several centuries, this smoothens out in populations. We are used to w-6 oils, but even for us the high LA sunflower oil is a big shift. It will be a few generations before the population level metabolisms stabilise and the negative impacts disappear. This could be true for adaption of any new or different food ingredients. On this premise, a large section of world population could be undergoing a metabolic churn now, because increased international travel, trade and commerce in the last 200 hundred years had brought about an intermingling or new exposures in the cuisines, that is mostly in the form of food ingredients and the process is still on. And then there is this industrial and processed food everywhere. If this process stays at where we are now, in another 200 years, the metabolisms would have stabilised against the changes from the past, life style diseases might slowly disappear. We should be wary of other external impacts, like climate changes.
I was raising this point in the context of a story by someone else on how the perceptions have changed over saturated fats. It was mentioned how the Intuits in arctic Canada or the Masai maras in central Africa, despite their total dependence on hunted/reared meat and thus saturated fats have not been very prone to cardiac diseases. My point was that their metabolisms got adjusted over centuries to this food available to them in their midst and the likely negative impacts were smoothened out.
I like how you’ve pulled this conversation towards the bigger evolutionary picture. You’re right: populations can and do adapt over time to new foods. The classic cases are things like lactase persistence in dairying populations, or the spread of extra amylase genes in high-starch farming cultures - but those shifts took thousands of years and dozens of generations.
The modern story of seed oils is much younger - under 100 years in most countries - so there hasn’t been time for genetic adaptation. What we do see is metabolic strain: obesity, diabetes, and CVD risk rising as high-LA oils and ultra-processed foods displace traditional diets. That’s less an adaptation process, and more a case of metabolic plasticity being pushed to its limits.
The Inuit and Masai are interesting examples too. Their apparent resilience to high saturated fat seems to come not from genetic rewiring in a few centuries, but from lifestyle context: high activity, low sugar, low processed food. Once those populations shift to modern diets, their rates of chronic disease rise rapidly, which shows the “protection” wasn’t permanent adaptation.
So I’d agree with you that we may see gradual smoothing out over many centuries, but in the near term, I believe the safest stance is still to limit excess omega-6 and ultra-processed oils, and to rebalance with more omega-3 where possible.
It is a pleasure seeing your engaging views. What I meant by speeding up ( of metabolic equilibrium) is the kind of things you have mentioned in the end. But seed oils are here to stay, even in North America. The traditional fats cannot generate enough volumes for the risen population in these decades. Moreover, the seeds (corn, soybean, cotton) let the American agriculture prosper big in a century and the story cannot be reversed. Oils from these cultivation actually boost the economics of agriculture substantially. It is a virtuoso cycle in a sense and it doesn’t make sense breaking it. There are two ways in which this seed oil conundrum can be turned upside. The first is to turn to peanut oil, canola oil and high oleic sunflower oil for major every day uses along with dairy and animal fats. And processed food industries must be directed to use them by decree, in place of soybean/corn/cotton seed oils. These measures can be adopted on the go and this will substantially limit LA consumption. The second one is the big oil industries like P&G, General Mills, Levers ( UK and Europe) must be asked to research and develop directed hydrogenation that will convert these old school seed oils to oleic acid rich oils, similar to olive oil. The catalyst development is the key. Though I left this industry long ago, my guess is that catalyst developments have taken place that will avoid trans acids formation, but the end point is still general, only iodine value, which would mean converting part of polyunsaturation into saturation, leaving some PUFA still. I feel this will be a very critical technical development, should it happen. At once, huge volumes of problematic seed oils can become safe MUFA rich oils, very affordable too. But who will nudge these oil giants ? Food authorities world over do not have people who can think and direct on those lines.
Animal farming , already facing global pollution and warming labels , cannot expand massively to meet full scale needs of dairy and animal fats, on present volumes. That is a limitation to returning to the past. So, the world, North America in particular, has to work around seed oils.
This is an historical point about the Indian food scene. Wheat in North India and rice in the south have been the dominant grains for millennia. But millets also had a strong place traditionally. Despite some inroads by these two mainstream grains in rural India, millets continue to be a choice there. In urban India, for a few decades in the second half of last century, millets had lost their flavour, but are making a return, as health food and backed by some promotion by the government boards. I feel this could be a permanent trend. This small flip flop at population level should not matter from the larger metabolism angle.
You’re absolutely right: seed oils aren’t just a nutritional question, they’re an economic and agricultural reality. They’re deeply embedded in North American farming, and it’s unrealistic (and undesirable) to imagine simply “going back” to animal and dairy fats to feed large populations.
The healthier path is exactly what you suggest: shifting the balance of fats towards MUFA-rich oils (like peanut, canola/rapeseed, and high-oleic sunflower) while keeping appropriate amounts of saturates and polyunsaturates. That pattern aligns much more closely with the evidence we have from Mediterranean-style diets and other long-lived populations.
On your second idea - as you say, it's technically feasible to convert omega-6–rich oils into MUFA-rich ones. The most practical approach is through crop breeding and biotechnology, which have already given us high-oleic variants of sunflower, safflower, soybean, and canola. These oils can be 70–80% oleic acid with much less linoleic acid, and they’re already used in some food manufacturing because they’re more stable for frying and have a longer shelf life. Directed hydrogenation is another route, but while modern catalysts may reduce the risk of trans fat formation, it’s less elegant and less widely adopted than breeding.
So in practice, I think the big opportunity is encouraging the food industry to switch from commodity high-LA oils to high-oleic oils - something that can be done at scale, is economically feasible, and would immediately improve the fatty acid profile of processed foods.
Your point about millets is a hopeful one: it shows that traditional staples can make a comeback when health, culture, and policy align. Perhaps the same can be done for fats - not by going backwards, but by steering the market towards healthier options and maybe engineering smarter oils.
We can look at this food evolutionary picture from another angle. Carbohydrates, Proteins and Fats are the three food components that are key to the metabolic state in populations. So any big change even in one of them by a population group is going to have a major health impact in the near term - a few decades atleast in the context of what we are discussing. Historically, population groups that we have today and that have existed for millennia around the world have not changed their choices of these three much over millennia with a recent origin exception or two on fats. Large population groups that traditionally lived on animal and dairy fats have to accommodate seed oils. Geographically these are North American populations mainly, possibly some in South America. Much of elsewhere, where a mix of fats has reigned, have also not seen big changes in their fat profile, again except for the invasion of sunflower oil in their consumption pattern. This could be a much smaller metabolic churn compared to the North American story - the switch in predominance from animal/dairy fats to the seed oils. This is a big change metabolically too and that is the reason why this region is the one facing maximum incidence of life style diseases. I don’t know if this geography realises the implications of the choice before them, but with some clever decisions, they can speed up the process of metabolic stability.
The largest intermingling has taken place in the baskets of fruits, vegetables, nuts etc that contribute to the needs of vitamins and minerals largely. Even here, I would say that two thirds of these lot are now common world over and their contribution to the metabolic instability picture in populations is not significant.
I think you’ve framed it very clearly: when whole populations shift one of the “big three” macronutrients, the effects are bound to be disruptive. The North American fat story is a striking example - in less than a century, a dietary base built on animal and dairy fat flipped to one dominated by seed oils, especially soy and corn. That’s a profound change in both the quality of fat (saturated vs high-LA polyunsaturated) and the quantity (as industrial oils made it into almost every processed food).
As you say, that’s a very different kind of metabolic shock than broadening the fruit and vegetable basket, which humans have always adapted to quite easily.
I’d agree this helps explain why the “lifestyle disease” burden is so high in North America: the scale and speed of the fat transition is historically unprecedented. Of course, it hasn’t happened in isolation - the rise of ultra-processed foods (where those oils are combined with refined starches, sugars, and additives) has amplified the metabolic disruption. Other regions with a more mixed fat intake (animal + plant oils) may face a gentler adjustment, though the sunflower oil boom is still significant.
Where I’d add a note of caution is on the idea of “speeding up” metabolic stability. Our biology does have a degree of plasticity - enzymes can adjust within a lifetime, and the microbiome shifts within weeks - but true genetic adaptation is much slower, usually over many centuries. Which means public health levers (shifting the n-6:n-3 balance, limiting industrial oils, encouraging traditional fats where appropriate) may be more effective than waiting for evolution to catch up.
I read recently about the health profile of pine nuts in another presentation, rechecked yours of this and found that you haven’t included it in your discussion here. The nuts have an oil content of over 70%. Even though the oil is rich in linoleic acid, close to 50%, the speciality of the oil seems to be the 18-20% of Pinolenic acid, a tri unsaturated w-6 (C 18, 5,9,12 double bond positions) which seems to have significant anti inflammatory properties. The oil also contains upto 5% of amino acids ( free amino acids or low mol wt peptides ?) which is another significant factor of its health quotient.
The important thing here is a w-6 being so beneficial, pinolenic acid, is it because it is triunsaturated), similar to gamma Linolenic acid ( GLA, 6,9,12). What is the mechanism here ? ALA is another outlier, a linolenic analogue, but with a different double bond position (12,15) and thus a w-3. Which is important, being w-3 regardless of the number of double bonds OR multiple double bonds, 3 or more, regardless of their position ?
Isn’t it ironic that the most abundant fatty acid type in the triglyceride world (oils and fats), Linoleic acid, double unsaturated and w-6, is now regarded as the least desirable molecule ? Being just w-6 is not a downside, but the number of double bonds is. Would the Linoleic acid story also come to pass in future ? That would eventually be the irony of reality, because it is the most abundantly occuring fatty acid in nature in triglyceride oils and fats.
Moro - brilliant observations, as always! You're expert on oil chemistry, not me, so I'll explain my understanding, but let me know if I get the chemistry wrong.
You’re right that pine nuts are unusual: their oil carries ~15–20% pinolenic acid (PNLA), an omega-6 triene (18:3 n-6, Δ5,9,12). PNLA isn’t just “another omega-6” - the double-bond pattern changes how enzymes handle it.
Mechanistically, PNLA has been shown to raise satiety hormones (CCK, GLP-1), improve hepatic LDL receptor activity, and dampen pro-inflammatory cytokines in cell and animal models. It also acts as a dual agonist at FFA1/FFA4 receptors, which may explain the metabolic and anti-diabetic signals seen in preclinical studies. Evidence in humans is still modest - small appetite and lipid studies, mostly - but intriguing enough to warrant more work.
On your broader question, does “omega-6” vs “omega-3,” bond count, or bond position matter most? I’d argue it’s all three:
• Family (n-3 vs n-6): determines which eicosanoids and mediators you can make, and which enzymes compete for substrates.
• Number of double bonds: influences membrane fluidity and susceptibility to oxidation.
• Position (Δ5 vs Δ6, etc.): alters how desaturases/elongases handle the molecule, and therefore which downstream lipids you end up with.
That’s why GLA (18:3 n-6, Δ6,9,12) and PNLA (Δ5,9,12) behave differently, despite both being omega-6 trienes. And why ALA (18:3 n-3, Δ9,12,15) is in another family altogether, with different fates.
On linoleic acid (LA, 18:2 n-6, Δ9,12): I don’t think it’s fair to call LA inherently undesirable. Substituting LA for saturated fat lowers LDL and reduces cardiovascular risk, and most RCTs don’t show LA raising inflammatory markers in healthy people. Where the concern arises, and I share it, is in the Western diet (and now urban Indian diet) context: very high LA intake from seed oils (often >10% of energy) combined with very low omega-3 intake. That imbalance means omega-6 dominates the shared metabolic pathways, tilting the system towards pro-inflammatory arachidonic acid-derived mediators. So while the molecule itself isn’t “bad,” in practice the context is.
Bottom line: pine nuts are a solid addition to a varied nut intake, but their high omega-6 content makes me wary of suggesting excess use. PNLA makes them especially interesting, but the best health signal still comes from a balance of mixed tree nuts and from keeping omega-3 intake up (ideally from oily fish) while not letting omega-6 dominate the plate.
Correction…….ALA is a linoleic analogue, not linolenic.
As a lifelong nut lover, I’m truly grateful for this fantastic nutritional guide to eating nuts. (I eat too much.) And I will share it with friends & family members. Thank you! [h/t to Dr. Chawla because i read your comment on her June 27th post!] 🙏🏽
Thanks for those kind words. I'm glad I've been able to give you an excuse to eat more nuts!
Thx! After all that, I didn’t see anything about activating nuts.
You’re right, Steve, I didn’t mention ‘activating’ nuts. While it’s discusssed in wellness circles, there’s no real evidence that soaking and dehydrating nuts significantly improves their nutritional value or digestibility.
Thank you for sharing!
Embrace the fuel.
Have you had your nuts today?
I have. 😌
Shockingly, I haven’t! I forgot. Time to put that right.
Yay! Love the nuts!
Fantastic! Now you have more reasons to enjoy them.
I will eat all of them.
Except peanuts. I don't trust them.
Don't ask me why.
You do right to eat all of them! I’m sure eating a wide variety maximises the health benefits, so you won’t miss the peanuts. I must admit though, I find peanuts irresistible!