Outline
Recommendation
How to minimize oxidation and reduce your exposure
Background and key terms
Why eating oxidized fats is unhealthy
At what level do these fats become harmful?
Is there any safe limit?
Our body’s natural defense mechanism
The cause of oxidation
What determines “smoke point”?
Smoke points of common cooking oils
Final thoughts
Recommendation: Avoid heating oils and fats to the point they begin smoking. Once that happens, it’s a sign that the oil has changed significantly from a process known as “lipid peroxidation”. This degradation is a process which begins well before the oil gets hot enough to see smoke, but using smoking as the sign your oil is too hot is an easy rule-of-thumb to follow.
The products created by heating oil make up a large category of harmful compounds. These products are absorbed by our bodies and carried in our blood. While we have natural defense against them, it is clearly the case that this natural defense doesn’t completely protect us from harm, even from a typical serving of french fries. Relax, this harm won’t cause cancer in the near term. Still, if you’re willing to trade some fried food now for (potentially) more time with your grandchildren later, this article shares the science behind why that is a reasonable decision. We have two choices regarding the amount of oxidized oils we eat: either we can wait for multiple randomized, controlled, triple-blind crossover studies before we reduce our consumption, or we can recognize that the mechanism of how they damage cells in our body is well understood, and adjust our diets accordingly. It’s about what’s more important to us–quality of life based on food today, or quality of life based on health down the road.
How to minimize oxidation and reduce exposure
If you do fry food, it would likely be healthier to use glass or aluminum cookware, as opposed to iron and copper (iron/copper ions can damage oil more easily). Also, eating a balanced diet of quality food provides adequate levels of the vitamins and minerals we need to support our natural defense against free radicals. We see issues when diets are deficient in certain nutrients. ((Free radicals, antioxidants, and nutrition (2002))) However, we might need additional support against free radicals for two reasons. First, even the veggies and fruits we eat have wildly different levels of nutrients depending on how long they’ve been out of the ground and their farming method. I previously wrote an article covering the importance of freshness here. Second, certain activities expose us to exceptionally high levels of free radicals–e.g. exercising in polluted city air, sunburns, and eating oxidized foods. Thus, while a “balanced diet” is certainly helpful, I think it’s probably based on research using cells which aren’t under equivalent oxidative attack compared to our cells, today.
Next time you buy bulk nuts, smell a few of them; if they’ve oxidized enough (likely due to air exposure) they may smell like a box of Crayons! I can’t say eating them causes more harm than good, but definitely try to find nuts that are as fresh as possible.
Protect lipids against excess oxidation by storing them in the absence of oxygen, and also by avoiding undue exposure to high temperatures or light. ((Chemical reactions…))
Turmeric alone provides antioxidant benefits when added to burger meat (before cooking) and black pepper seems to amplify this benefit. ((Turmeric and black pepper spices decrease lipid peroxidation in meat patties during cooking (2015))) Note: black pepper alone wasn’t able to show significant antioxidative ability.
Background and key terms
You’ve maybe heard chefs, doctors, or talk-show hosts say cooking oils can create “harmful compounds” if the oil gets too hot. Rarely do they tell us what those compounds are, how hot the oil has to get, let alone any research or explanation of the science. So, here’s what I found to explain the issues with cooking oils. Even though, in the kitchen, we use the terms “fat” for solids (butter) and “oil” for liquids (canola oil), for the sake of this topic I’ll just use the term “fats” to refer to oils, too.
To understand this topic, I had to learn some basic chemistry and a few key terms, some of which I’ll define here. “Fats” are a type of lipid. The “Lipids” category also includes waxes, sterols, and other natural chemical compounds unrelated to cooking. The terms we need to clarify are:
“Edible Fats/oils”–a specific type of lipid. At room temperature and normal air pressure, fats are solid and oils are liquid; further distinction between fats & oils isn’t needed since we will discuss saturation.
“Oxidation”–Oxidation of edible fats, in simple terms, means damage due to light, heat or air along with the production of certain molecules, some of which are “free radicals”. Each fat has a different temperature at which it will begin to smoke due to heat, though moisture, cookware, oxygen, and antioxidants in the environment can raise or lower this temperature (a.k.a. “smoke point”). As oxidation increases, it causes rancidity, usually recognized as a bad smell or taste. For now, we’ll assume oxidation is bad, but later in this article I’ll explain exactly how it’s actually harmful.
“Fatty acids”–chains of basic Carbon-based units (methylene bridges and alkenes); these chains are what oxidize when heated. They’re considered either short, medium, or long and either “saturated” or “unsaturated”, which I’ll describe later.
“Glyceride”/”Acylglycerol”–Structure comprised of 1, 2, or 3 fatty acids connected to a single glycerol molecule called a “backbone”. See here:
“Triglyceride”/”Triacylglycerol”–A glyceride with 3 fatty acids, similar in shape to the letter “E”. Each fatty acid can be the same or different. The fats we use for cooking are 90-100% (by volume) made up of various triglycerides, which is why you’ll hear some people refer to fats generally as “triglycerides”, though fats can contain a few mono- and diglycerides floating around, too.
Example:
The big picture: Fatty acids combine to make triglycerides, and triglycerides combine to make the fat we eat.
Why is eating oxidized fats unhealthy?
Here are two pretty good videos:
One explaining the damage free radicals can do and one on how antioxidants protect against this damage.
Here’s an excellent, succinct description (paraphrased) of the initiation and propagation steps in lipid peroxidation:
When free radicals and other reactive species extract a hydrogen atom from an unsaturated fatty acid, that fatty acid becomes a radical itself. Free oxygen joins to create an oxidized lipid radical. It further propagates the peroxidation chain reaction by abstracting a hydrogen atom from a nearby unsaturated fatty acid (of which there are plenty around since this happens in a bottle of oil, for example). ((Free radicals, antioxidants, and nutrition (2002))) This chain reaction continues until the radicals are neutralized by each other or antioxidants in the environment.
When a fatty acid oxidizes, those “harmful compounds” everyone warns us about are indeed created. That catchall term is used to refer to all of the various molecules created by oxidation for a good reason–there are tons of different names for these things. It seems to depend on what context dictates–since we’re speaking broadly, we can call them “lipid peroxides”, of which there are several categories (e.g. “reactive aldehydes”, “dicarbonyls”).
A study on two “reactive aldehydes” created by oxidation explains that they remain in the blood stream until our body neutralizes them with it’s own antioxidants. While circulating, they damage the walls of our blood vessels, causing inflammation and, over time, atherosclerosis. ((Lipid peroxides and human diseases (1987))) Since food is known to be one way to get radicals into the blood, ((4-Hydroxynonenal (HNE), a Toxic Aldehyde in French Fries from Fast Food Restaurants (2015))) limiting foods which contain these two compounds seems like a decent strategy for long-term health. Considering that study specifically speaks about those two chemicals, I want to avoid assuming that all of the harmful compounds follow these rules. This goes for all studies. Still, the basic insights suggest we should limit foods cooked in this way even if we’re only limiting these two (out of many) harmful compounds.
At what level do these molecules become harmful?
It seems unclear currently, but in the range “expected to be found in vivo”, animal studies of cultured cells show increased tumor frequency and incidence of atherosclerosis, as well as disrupted function of blood vessel, liver, and lymphatic cells. They add: “Recent findings strongly suggest that in vivo modification of low-density lipoprotein by certain lipid-peroxidation products (eg, 4-hydroxynonenal and malonaldehyde) renders this lipoprotein more atherogenic”. ((Cytotoxicity and genotoxicity…)) Many studies I skimmed referenced more-dangerous LDL as a result of consuming oxidized fats.
In the french fry study cited above, results from restaurant No. 1 ranged between 19.07 and 32.15 µg HNE/100 g FF. The average value was 22.23 µg HNE/100 g FF. For restaurant No. 2, the fries contained between 7.47 and 10.21 µg HNE/100 g FF, averaging 8.71. Another study using fresh (not frozen) fries found they also absorbed the compound, though they absorbed less than the pre-frozen fries did (4.90 ± 0.47 mg HNE per 100 g fried potato). ((Incorporation of the toxic aldehyde…)) It should be noted they fried the fries in soybean oil (about 50% linoleic fatty acid) for 10 minutes at 365 degrees Fahrenheit. Having worked in restaurants, I can say those aren’t unusual conditions for frying fries. They also note that corn oil, which contains about 58% linoleic acid, has a similarly high concentration of HNE compared to soybean oil under similar heat treatment. Based on other research showing the small-intestine’s ability to absorb these fats, I feel comfortable concluding we are exposed to harmful amounts when our diet pattern includes standard servings of foods fried in oil.
Is there any “safe” limit?
So, there seem to be two answers to the question “how much oxidized fat can we safely eat?”. Simply put: any free radical will damage a nearby cell. Thus, the ideal diet would minimize oxidized fatty acids since they are absorbed by the small intestine and packaged into LDL. This atherogenic LDL is taken up by macrophages and leads to increased foam cell creation (early precursors to heart disease). ((Antioxidant defenses and lipid peroxidation in human blood plasma (1988))) & ((Nutritional, Dietary and Postprandial Oxidative Stress (2005))) Again, considering diet can increase the radicals in our blood, limiting our oxidized foods as much as possible seems smart, if long-term health is important to us.
The complex, more practical answer is that we can’t definitively know the limit of our natural defenses. We simply can’t prove that our body can handle 9 french fries per hour, but the radicals from the 10th fry are left to wreak havoc on our cells. It’s left to our reason–after some research on our own–to determine how much we’ll give our bodies to handle, taking into account the radicals in our environment beyond our diet. One study showed that lipid hydroperoxides (products of fatty acid oxidation) aren’t likely to pass through intact epithelium, due to adequate innate antioxidant defense. ((Gastrointestinal glutathione peroxidase prevents transport of lipid hydroperoxides in CaCo-2 cells (2000))) However, they noted several other studies showing that the degradation products of lipid hydroperoxides (like the reactive aldehydes mentioned above) do get absorbed in the presence of normal antioxidant levels. It’s also been suggested that adding antioxidants to high-fat meals can alleviate the amount of oxidative stress incurred after eating. ((The postprandial effect of components of the mediterranean diet on endothelial function (2000)))
Another way to say it: Are we overloading our natural defenses when we smoke oil?
Quite simply, we incur excess oxidative stress after eating oxidized fats. A 2007 study looking at red wine polyphenols put it best (citing 4 studies):
High-fat, high-cholesterol foods containing oxidized products affect endogenous lipoprotein production and catabolism, and lead to transient exposure of arteries to cytotoxic chylomicron remnants and ALEs.
“ALE” stands for “advanced lipid peroxidation end product” and is a correlate to an “AGE”, the G being for “glycation”. The main finding from this study was that red wine (specifically, the polyphenols in it) reduces plasma and urine MDA Levels. ((A novel function of red wine polyphenols in humans: prevention of absorption of cytotoxic lipid peroxidation products (2007))) They hypothesize, citing supporting research, potential mechanisms for this. Skeptics should note they freeze-dried the wine, thus it contained no alcohol, and they froze the prepared meals for a few weeks. Because of this, it was a different type of preparation than would be done in a real setting. However, considering the mechanisms of autoxidation, I’d guess any processing only increases oxidation levels at consumption. Still, I suspect alcohol might cause some oxidative effect, had they left it in. They also used commercially available wine and turkey cooked “well-done” thus creating at least as many radicals as would likely be found when the layperson prepares this meal. Note: recent research suggests MDA may not be a huge concern, though it is reactive. ((Lipid Peroxidation: Chemical Mechanism, Biological Implications and Analytical Determination (2012)))
One rat study found that intestinal cells have the ability to absorb and repackage oxidized fatty acids from linoleic acid. Linoleic acid is the main fatty acid in soybean, corn, peanut, and sunflower oils, so it is probably the most commonly-consumed oxidized fat. ((Dietary oxidized fatty acids: an atherogenic risk? (2000)))
One group twice proved that we do ingest reactive aldehydes created in burger meat. The key finding in both studies was that we can significantly reduce these harmful compounds if we use antioxidant-containing spices. Healthy ((Antioxidant-rich spice added to hamburger meat during cooking results in reduced meat, plasma, and urine malondialdehyde concentrations (2010))) and diabetic ((Decrease of postprandial endothelial dysfunction by spice mix added to high-fat hamburger meat in men with Type 2 diabetes mellitus (2013))) human subjects were tested.
Not only are harmful compounds created, but the other ingredients in foods we’re cooking experience oxidation as their exposure to free radicals increases. When our cooking method oxidizes fats, any nutrients nearby are susceptible to attack. Thus, the propagation of the chain reaction should continue, thus reducing the nutrition and antioxidant capacity of any vegetables and herbs we’re using. ((Lipid Peroxidation and Coupled Vitamin Oxidation in Simulated and Human Gastric Fluid Inhibited by Dietary Polyphenols: Health Implications (2005)))
The cause of oxidation (a.k.a. weak hydrogen-carbon bonds in unsaturated fatty acids)
Fatty acids come in many different types. To understand how to cook with fats properly, we need to quickly cover length and saturation.
Length of a fatty acid is denoted by the letter “C” followed by the number of carbons in the chain. C8 (caprylic acid) has 8 carbons and looks like this:
While length is simple to understand, saturation can be confusing. It only refers to how many double-bonds exist between carbons in the chain. Since “saturated” fatty acids contain zero double-bonds, think of saturation as meaning the carbons are all “saturated” by hydrogen atoms–all the Cs have an H on either side (as seen in the example of caprylic acid above). The key to lipid peroxidation is the weak hydrogen-carbon single bond next to a carbon-carbon double-bond. That H-C bond can be only half as strong than the H-C single-bonds which aren’t next to a C=C double-bond. This weakness allows the hydrogen to be easily stolen from its carbon by a free radical. Thus, saturated fats are less prone to oxidation because they don’t contain any double bonds which weaken the neighboring H-C bond.
Unsaturated fatty acids contain one or more double bonds; the more double bonds they have, the more susceptible to oxidation they are. Monounsaturated fats have one double bond, while polyunsaturated fats have two or more.
Induction period and relative rate of oxidation for fatty acids at moderate temperatures is strongly dependent on unsaturation.
(From the same source) “The observation that saturated lipids only autoxidise very slowly means that this reaction occurs involving the isolenic unsaturated double bonds present. When subject to high temperature as during cooking and especially frying operations, reactions of oils with food components can greatly accelerate oxidation. Eventually the foaming from the frying oil increases and smoke is liberated upon heating even at low temperatures. At this stage the use of oil in frying is no longer only a long term health risk but also represents, especially if using an open flame, a dangerous behaviour because of it’s increased flamability and the risk of fire.
The -scission reaction may leave carbonyl groups attached to the triacylglyceride moiety (core aldehydes). Core aldehydes are non-volatile and extremely reactive species, capable of adding to amine groups from protein material (available from other components of food) and starting processes which eventually lead to policyclic compounds, which may be heterocyclic and will contribute to the light absorption pattern of the polymeric mass. Total solids content is one of the criteria for discarding a cooking oil, especially because this may be determined experimentally with some ease.” ((Chemical Reactions of Oil, Fat, and Fat Based Products))
Finally, many different compounds are created as a byproduct of lipid peroxidation:
So what determines smoke point?
The smoke point is simply the lowest temperature at which oxidized compounds reach a high-enough volume to appear. Generally, unrefined polyunsaturated oils have lower smoke points than monounsaturated and saturated fats because of their susceptibility to oxidation.
In fresh oils (before they’re cooked) the smoke point is mostly connected to the quality of the oil. Specifically, the more low molecular weight molecules (e.g. free fatty acids) there are floating around, the lower it’s smoke point will be. ((Evaluation of physicochemical changes in cooking oil during heating (1987))) This is why refined oils have higher smoke points as compared to their “dirtier” natural varieties. Also, how much the fat has oxidized matters (in storage or extraction, for example). ((Effects of Antioxidant and Cholesterol on Smoke Point of Oils (1997)))
Also, different fats are made up of different triglycerides. Different triglycerides have different combinations of fatty acids. Here’s the triglyceride from above, but note it has one saturated and two unsaturated fatty acids:
As a side note, monoglycerides and diglycerides also exist, but triglycerides make up the vast majority of the fats we consume, and use for cooking. Furthermore, they follow the same rules since they’re made of up of the same fatty acids as triglycerides.
Where do common oils fall on the spectrum?
See coconut, avocado, macadamia, sesame, corn, canola, olive, butter, ghee, sunflower, rapeseed here:
This is a chart showing the percentage of each fatty acid in common cooking oils. Note the breakdown of saturated vs unsaturated at the bottom of the table:
Final thoughts
While we wait for randomized controlled triple-blind crossover studies to prove we can reduce our risk of cancer by limiting fried food, we must behave with respect to the information we have. There is no evidence that oxidized fats is supportive of health. Meanwhile, there is an increasing amount of evidence that it is detrimental to health. At least, understanding this, we can make more informed decisions.