Is It Reasonable To Think Seed Oils Cause Obesity?
tl;dr: Nick Norwitz is right, his position is moderate. But is it correct?
Nick Norwitz (if you don’t know him) has a PhD from Oxford, is currently pursuing an MD at Harvard, and manages in his copious spare time (hah!) to have a popular Youtube channel and to publish papers with Dave Feldman. Last year we discussed seed oils a bit, and have chatted a little bit off-line.
Today Nick published a short explanation of why he think concern about seed oils is over-hyped in some respects, and “reasonable” in others.
Substack:
And X:
Nick is a very smart guy, but as he has made clear, this is not his specialty.
So I thought it might be helpful to address his points one by one. Do read his post first, I’ll not be fully reproducing it here.
“1. Defining Seed Oils”
“Highly oxidized, refined heated/damaged oils, e.g. French Fry Oil at MacDonald’s, is NOT the same as the oil in a raw sesame seed or walnut.
“That should be obvious but it deserves emphasis.”
This is correct, as far as it goes.
Linoleic acid on its own is harmless, but when it becomes oxidized it is toxic. This is indisputable.
However, the dose makes the poison. We have extensive systems to cope with oxidized PUFAs like linoleic acid, but we do not have an unlimited capacity to deal with PUFA intake. This is known as oxidative stress, which is an excess, essentially, of oxidized PUFA beyond our capacity to detoxify it. For instance:
“…there is a vast excess of polyunsaturated fatty acids in LDL, in relationship to the content of natural, endogenous antioxidants.” (Witztum, 1991)
PUFA are very susceptible to oxidation, as Nick notes, so the question becomes, if we consume un-oxidized PUFA, can they oxidize in the body and cause oxidative stress?
Can the PUFA in nuts exceed our evolved capacity?
So for instance, seed oils reduce LDL levels, as do nuts. The mechanism appears to be a common one to PUFA: that the oxidation of these fats prevents VLDL secretion and thus lowers LDL levels (Pan, 2004; Krauss, 2004). Seed oils in RCTs nevertheless increase CVD mortality, in proportion to their LDL lowering (Ramsden, 2016). Would nuts do the same, or would the lower levels of oxidation in nut fats be protective? Some level of protection seems likely. Sadly we don’t have great evidence for benefit or harm from nut consumption (Cofán, 2020).
The only hunter-gatherer group that shows signs of obesity and diabetes on their traditional diet eat a diet high in nuts (Joffe, 1971).
Nick doesn’t offer a definition of seed oils. I do think it’s important, as he suggests, to distinguish between seed oils in whole-food nuts and what I like to call industrial seed oils, which have been refined and oxidized in the course of production. But I think that seed oils as oils from seeds (to distinguish from healthier “fruit” oils like olive or avocado) is a pretty reasonable definition.
“2. Do Not Equate Seed Oils to Omega-6/Linoleic Acid”
“Building on the prior point, “seed oils” should not be equated with polyunsaturated fatty acids (PUFA), omega-6, or linoleic acid.”
This is in error. A defining trait of Ω-6 linoleic acid (18:2 n-6 in the image) is that it is most commonly produced by seeds of plants.
It is true that some seed oils contain Ω-3 plant fats (18:3 n-3 above) and that these have distinct properties (and risks), but I don’t think that terminology is the reason “it has been impossible to resolve this issue.”
“3. Overload of Omega-6 Can Cause Inflammation. However…”
“That said, an imbalance of omega-6 and 3 in the body can create problems, contributing to inflammation. (*See Nuance Note, below)”
We agree on that!
His nuance note is the Fat-1 mouse model (Kaliannan, 2019), which I discussed in this post:
More about that later.
“However, dietary omega-6 and 3 intake does not directly translate into blood/membrane omega-6/3 levels.”
This isn’t quite true. Animal-based Ω-3 intake does map pretty well, PUFA from seeds does not.
“Plasma concentrations of eicosapentaenoic acid (EPA), docosahexaenoic acid (DHA) and (n-3) highly unsaturated FA (HUFA) derived from marine foods, but not linoleic and α-linolenic acid from plant origins, demonstrated positive correlations with dietary intakes (r = 0.303– 0.602, P < 0.05) in both genders.” (Kuriki, 2003)
The care that Nick takes with his diet is demonstrated by his AA:EPA ratio of 1. We should all do so well.
“4. Real Foods for the Win… Thanks Nature!”
“Real whole foods area often “packed” with compounds that may compensate for and/or protect against oxidation of fragile fats. For example, sesame oil and sesame products (tahini, etc.)…”
For sure. Nuts are nice packages of compounds designed to keep fats from going rancid. There’s an experiment I read about (can’t find it now) where they store seeds and then germinate them, and some of the seeds are still viable after decades.
A refined seed oil goes rancid from the moment production commences, and it’s rapidly downhill after that.
But sesame oil and tahini are not “whole” foods. If you want to oxidize a fat, refining or blending it is a fine way to do so. Sesame is also a rather unique oil, due to its strong antioxidant component. One shouldn’t extrapolate that oil to other seed oils.
“Intakes of lard, peanut oil, and refined blended plant oil but not soybean oil, canola oil, and sesame oil are associated with higher T2D risk.”
Sesame is superior because in addition to the antioxidants it contains, it is used sparingly, and not used for frying.
“On the contrary, we found that sesame oil, which was rarely used for frying (7.7%), showed no significant association with T2D risk. In addition to abundant PUFAs and MUFAs, sesame oil also contains vitamin E and profuse sesame lignans, including episesamin, sesamin, sesamol, and sesamolin, which possess antioxidant properties (39, 40)…. In our study, the nonsignificant association observed for sesame oil may result from the overall low intake amount in this population given that 95% of participants consumed <1.8 g [per day].” (Zhuang, 2020)
I also use sesame oil in my cooking, but I add it at the end, so it won’t oxidize. And I love tahini, but don’t eat it frequently.
(And before you get excited about soybean and canola (rapeseed) oils, frying with those oils correlates with lung cancer risk in China (Ko, 2000).)
“5. Randomized Controlled Trials are Not Up To the Task”
“Short-term randomized controlled trials (including ‘calorie- and macro- controlled’ trials) are not equipped to properly assess the potential harms of ‘seed oils.’
“Whether you define seed oils here as highly oxidized oils or just very high omega-6 content vs control, the physiological changes induced by an omega-6/3 ‘imbalance’ and/or oxidative stress take many months and more likely years to manifest.
“Thus, a ‘negative’ finding in a relative short-term RCT can pose a false negative for health effects.”
Agreed.
We do have some older RCTs that show consistent signs of harm from seed oils consumption (Ramsden, 2016); but even those are confounded by a problem Nick doesn’t mention here: ubiquity.
Seed oils have been a significant and increasing factor in the human diet since the 1860s:
“The garbage of 1800 became the fertilizer of 1870, the cattle food of 1880, and is now made to yield table food and useful articles of industrial pursuits.” (Grimshaw, 1889)
And now they are up to 21% of our daily calories consumed. This makes RCTs impossible: where are you going to find the control arm that doesn’t consume them? To take two groups and add seed oils to one, is to beg the question: at what point does consumption become a problem, if ever?
“So can the role of near-ubiquitous risk factors be studied and their importance for disease causation quantified? Yes they can, but it’s not easy, and they cannot always be studied with our ‘standard’ methods such as randomized controlled trials, cohort studies or case–control studies.” (Pierce, 2011)
“6. Metabolic (Macronutrient) Context Matters”
“As mentioned earlier, the degree to which omega-6 and omega-3 end up getting stored in your body and/or used for structural purposes depends on many metabolic factors.”
This is absolutely, and annoyingly, true.
It would be nice if this was a simple story, but it’s not. Not only the amount, but other factors like fats consumed alongside seed oils (including saturated and monounsaturated fats), carbohydrates, vitamins and minerals, and genetic variance can all play a role in the effects of Ω-6 fats.
“In summary of my perspective”
“Based on basic physiology, it’s reasonable to believe ‘seed oils’ (see point 1-2) are harmful beyond the calorie load, and paucity/absence of human RCT data is not all that meaningful (see point 5).”
I’m in agreement with that, as far as it goes.
“However, it’s also not reasonable to claim, as many do, ‘seed oils are poison and/or THE reason for the obesity epidemic.’ It’s also not reasonable to claim all high omega-6/LA foods are ‘bad’ because of their PUFA content.”
This is a false equivalence, an example of poor logic I am surprised to see from Nick.
“Seed oils are poison”/“not reasonable to claim all high omega-6/LA foods are ‘bad’ because of their PUFA content.”
The dose makes the poison. This is true for water, oxygen, and it’s certainly true for seed oils. It’s obvious that at some point we will be at a toxic level, where they have become a poison.
The fact that in the animal models we use to understand chronic disease, seed oils will induce all of them, at once, should suggest that we’re into the toxic level in the human population experiencing those same diseases.
“THE reason for the obesity epidemic.”
Physicians have been prescribing seed oils to induce weight gain since the 1950s
And it is how Nick might do it (I hope not) when he begins practicing medicine, as it’s standard.
So yes, the fact that the Dietary Guidelines bear a striking resemblance to Ensure, which bears a striking resemblance to the diet scientists use to make animals fat, makes seed oils THE leading candidate for our obesity epidemic. Furthermore, our leading treatments all affect the pathway by which seed oils are known (and used) to induce obesity.
Nick mentioned the fat-1 rodent model.
“Nuance Note: The direct impact of omega-6/3 ratio can be best assessed through transgenic animal since, in humans, omega-3 and 6 can't be synthesized de novo. So, I order to manipulate 6/3 ratio, you need to change other dietary variables. This can be overcome in animal models with genetic manipulation.”
That model has been used to demonstrate (see my post above) that Ω-6 fats are obesogenic, and that the Fat-1 mouse’s ability to change Ω-6 fats to Ω-3 provides protection against obesity.
I can’t think of a better model for determining the role of Ω-6 seed oils in obesity, and I’m glad Nick agrees with me. It’s the most reasonable, parsimonious explanation.
Conclusion
I hope you all, and especially Nick, whom I respect a lot, find this useful. Nick ends his post with a summary of what he thinks is right in a dietary context, and I agree with them.
And I should add, “Stay Curious!”
References
Cofán, M., Rajaram, S., Sala-Vila, A., Valls-Pedret, C., Serra-Mir, M., Roth, I., Freitas-Simoes, T. M., Bitok, E., Sabaté, J., & Ros, E. (2020). Effects of 2-Year Walnut-Supplemented Diet on Inflammatory Biomarkers. Journal of the American College of Cardiology, 76(19), 2282–2284. https://doi.org/10.1016/j.jacc.2020.07.071
Grimshaw, R. (1889). Industrial Applications of Cottonseed Oil. Journal of the Franklin Institute, 127(3), 191–203. https://doi.org/10.1016/0016-0032(89)90146-4
Hao, L., Nie, Y.-H., Chen, C.-Y., Li, X.-Y., Kaliannan, K., & Kang, J. X. (2022). Omega-3 Polyunsaturated Fatty Acids Protect against High-Fat Diet-Induced Morphological and Functional Impairments of Brown Fat in Transgenic Fat-1 Mice. International Journal of Molecular Sciences, 23(19), Article 19. https://doi.org/10.3390/ijms231911903
Joffe, B. I., Jackson, W. P. U., Thomas, M. E., Toyer, M. G., Keller, P., Pimstone, B. L., & Zamit, R. (1971). Metabolic Responses to Oral Glucose in the Kalahari Bushmen. British Medical Journal, 4(5781), 206–208.
Kaliannan, K., Li, X.-Y., Wang, B., Pan, Q., Chen, C.-Y., Hao, L., Xie, S., & Kang, J. X. (2019). Multi-Omic Analysis in Transgenic Mice Implicates Omega-6/Omega-3 Fatty Acid Imbalance as a Risk Factor for Chronic Disease. Communications Biology, 2(1), Article 1. https://doi.org/10.1038/s42003-019-0521-4
Ko, Y. C., Cheng, L. S., Lee, C. H., Huang, J. J., Huang, M. S., Kao, E. L., Wang, H. Z., & Lin, H. J. (2000). Chinese Food Cooking and Lung Cancer in Women Nonsmokers. American Journal of Epidemiology, 151(2), 140–147. https://doi.org/10.1093/oxfordjournals.aje.a010181
Kuriki, K., Nagaya, T., Tokudome, Y., Imaeda, N., Fujiwara, N., Sato, J., Goto, C., Ikeda, M., Maki, S., Tajima, K., & Tokudome, S. (2003). Plasma Concentrations of (n-3) Highly Unsaturated Fatty Acids Are Good Biomarkers of Relative Dietary Fatty Acid Intakes: A Cross-Sectional Study. The Journal of Nutrition, 133(11), 3643–3650. https://doi.org/10.1093/jn/133.11.3643
Pan, M., Cederbaum, A., Zhang, Y., Ginsberg, H., Williams, K., & Fisher, E. (2004). Lipid Peroxidation and Oxidant Stress Regulate Hepatic Apolipoprotein B Degradation and VLDL Production. The Journal of Clinical Investigation, 113(9). https://doi.org/10.1172/JCI19197
Pearce, N. (2011). Epidemiology in a Changing World: Variation, Causation and Ubiquitous Risk Factors. International Journal of Epidemiology, 40(2), 503–512. https://doi.org/10.1093/ije/dyq257
Ramsden, C. E., Zamora, D., Majchrzak-Hong, S., Faurot, K. R., Broste, S. K., Frantz, R. P., Davis, J. M., Ringel, A., Suchindran, C. M., & Hibbeln, J. R. (2016). Re-Evaluation of the Traditional Diet-Heart Hypothesis: Analysis of Recovered Data from Minnesota Coronary Experiment (1968-73). BMJ, 353. https://doi.org/10.1136/bmj.i1246
Various. (1958). Advertising section. British Medical Journal. https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC2028440&blobtype=pdf
Witztum, J. L., & Steinberg, D. (1991). Role of Oxidized Low Density Lipoprotein in Atherogenesis. Journal of Clinical Investigation, 88(6), 1785–1792. https://doi.org/10.1172/JCI115499
Zhuang, P., Mao, L., Wu, F., Wang, J., Jiao, J., & Zhang, Y. (2020). Cooking Oil Consumption Is Positively Associated with Risk of Type 2 Diabetes in a Chinese Nationwide Cohort Study. The Journal of Nutrition. https://doi.org/10.1093/jn/nxaa103
This is helpful, and regarding your reference to lung cancer connected with use of rapeseed oil in China, it'd be worth knowing more about the rest of Asia, which uses all kinds of oils. China uses primarily soy and peanut (among others) but obesity is rare. Singapore and Malaysia use a lot of palm oil, and obesity is more common and seems to be getting worse. Seems like not all seed oils are created equally. Maybe has to do with the heating and oxidation?
Huge Thanks Tucker,
Followed your work quite awhile, searching for more the whole time.
Your impact is far greater than your numbers! My health included.
Follow Nick too and enjoy him immensely also, I don't know what it takes to produce as much as he does while studying at such a high level. I'd love to see this work combined with Dave and Nick's work! I'd love to see hundreds of rats, normal vs 1 oil added while keto, and build thresholds of bad, or something like that for all the common oils. Oil alone, oil at bad threshold, the later oil with glucose or fructose as the accelerant factor etc. Whatever makes sense to get an RDA.