Why You Can't Safely Cut Fat the Way You Can Cut Carbs

Diets that go too low in fat (typically below roughly 20% of total calories) risk hormonal and micronutrient consequences that low-carbohydrate diets generally do not carry to the same degree. That's because, unlike carbohydrate, dietary fat contains genuinely essential components: linoleic acid (omega-6) and alpha-linolenic acid (omega-3) cannot be synthesised by the body and must come from food. Fat is also required to absorb the fat-soluble vitamins A, D, E, and K, and is the primary substrate for steroid hormone production, including testosterone and oestrogen.

Saturated, unsaturated, and the cholesterol question

Saturated fat (animal fats, butter, coconut oil) and its link to heart disease is one of nutrition's real, ongoing arguments: some large studies show weaker links than once assumed[26], while trials swapping saturated fat for polyunsaturated fat (not refined carbs) do show fewer cardiovascular events[27]. The mainstream view — a 2017 American Heart Association advisory — is that what you replace saturated fat with matters as much as cutting it[28]. A vocal group of cardiologists disagree, arguing the case against saturated fat is weaker than the public health messaging suggests[29]. Either way, the food source matters more than the nutrient total — a steak and a packet of biscuits aren't equivalent just because they share a saturated fat number.

Red meat, processed meat, and cancer risk

The IARC classifies processed meat (bacon, sausages, deli meats — anything smoked, cured, or salted) as a known cause of cancer, based on solid evidence linking it to colorectal cancer. Unprocessed red meat gets the weaker "probably" label, reflecting thinner, less consistent evidence[30]. As with any IARC hazard classification, the label tells you how confident the evidence is that a risk exists at all, not how big that risk actually is: processed meat raises colorectal cancer risk by around 18% per 50g eaten daily[31] — real, but starting from a baseline lifetime risk of only a few percent, so the actual increase in your odds is modest, not dramatic. The practical takeaway: treat processed meat as occasional, not a daily staple, and favour unprocessed cuts when you do eat red meat regularly. An occasional bacon sandwich isn't a cigarette.

A separate consideration: environmental footprint

This is a distinct argument from the health one above and shouldn't be conflated with it: the largest cross-study analysis of food production found that, per gram of protein, beef and lamb generate roughly 20–100 times more greenhouse gas emissions and use vastly more land and water than legumes, grains, or other plant protein sources[32]. None of this changes the health verdict above — the cancer-risk and environmental arguments rest on different evidence and don't need to point the same way for the same person. This guide treats environmental impact as a secondary, values-based consideration for the reader to weigh, not something baked into its Science Verdict recommendations.

Seed oils: the inflammation claim versus the evidence

A common online claim: seed oils (sunflower, soybean, corn, canola) drive chronic inflammation via their omega-6 content and should be avoided. The mechanism is plausible-sounding, but controlled trials directly testing omega-6 intake generally don't show raised inflammatory markers in humans — several show mildly favourable cholesterol effects instead[33]. Observational links to disease are confounded by the fact that high seed oil intake correlates strongly with high ultra-processed food intake (Section 1)[7] — separating the two is genuinely hard, and online discourse often attributes effects to the oil that may belong to the processed-food pattern it travels in.

Omega-3 and omega-6: ratio versus absolute intake

Omega-3 fatty acids (EPA and DHA, found mainly in oily fish, with ALA in flaxseed and walnuts converting only inefficiently to EPA/DHA) are associated with reduced inflammation markers and modest cardiovascular benefit. Popular claims about needing a strict omega-6:omega-3 ratio are not well supported; current evidence favours focusing on absolute omega-3 intake (roughly 250–500mg combined EPA/DHA per day for general health, achievable through 2 portions of oily fish weekly) over obsessing about ratios, or about seed oils as a category, discussed above. Vegetarians and vegans, for whom oily fish is off the table, can reach the same EPA/DHA target directly with an algae-oil supplement rather than relying on the inefficient ALA-to-EPA/DHA conversion from flaxseed and walnuts alone.

Trans fats: the one genuinely settled villain

Industrially produced trans fats, formed through partial hydrogenation (a chemical process that turns liquid vegetable oil solid) of vegetable oils, are the one category of fat with essentially no scientific defenders. They raise LDL cholesterol while lowering protective HDL cholesterol, and are associated with significantly increased cardiovascular disease risk per gram compared with any other fat type[34]. Most high-income countries have heavily restricted or banned them in commercial food production, though they can still appear in some processed and fried foods, particularly outside such regulation.