Seed Oils

What it is

Seed oils are a group of cooking oils pressed from plant seeds — canola (rapeseed), soybean, sunflower, corn, safflower, grapeseed, rice bran and cottonseed. They are a subset of vegetable oils high in polyunsaturated fat, especially linoleic acid, an omega-6 fat [1]. Omega-6 fats are polyunsaturated fats needed for many body functions [2].

What the research says

In controlled feeding trials, replacing saturated fat with polyunsaturated fat lowers LDL cholesterol: a 2025 meta-analysis of 24 trials (1,011 adults) found diets higher in polyunsaturated fat lowered LDL by 9.8 mg/dL on average, and by 15.7 mg/dL when the two diets really differed in saturated fat [3]. A 2026 meta-analysis of 53 trials (1,690 people) found high-polyunsaturated diets also lowered LDL, total cholesterol and triglycerides a little more than high-monounsaturated diets [4]. Cochrane's omega-6 review found higher omega-6 intake lowered total cholesterol by 0.33 mmol/L, with no clear change in LDL [2].

Long-term trials that replaced saturated fat with unsaturated oils found fewer heart problems. A 2020 Cochrane review of 15 randomised trials (about 59,000 people) found reducing saturated fat cut combined cardiovascular events by 21% (risk ratio 0.79) over two years or more, with the biggest benefits where cholesterol fell most; replacing the energy with polyunsaturated fat or carbohydrate worked similarly [5]. An earlier meta-analysis of 8 trials (13,614 people, 1,042 heart events) found replacing saturated fat with polyunsaturated fat reduced heart attacks and heart-disease deaths by 19% (RR 0.81) [6]. A 2026 risk-stratified review of 17 trials (66,337 people) found replacing saturated fat with polyunsaturated fat reduced non-fatal heart attacks (RR 0.75), with meaningful benefits mainly in people at higher cardiovascular risk [7]. Cochrane's omega-6-specific review of 19 trials (6,461 people, one to eight years) found little or no difference in cardiovascular events (RR 0.97) or deaths (RR 1.00) and a possible reduction in heart attacks (RR 0.88) [2].

Large observational studies link higher linoleic acid to better heart outcomes. Pooled individual data from 30 studies (68,659 people, 15,198 cardiovascular events) found higher circulating or tissue linoleic acid was associated with lower total cardiovascular disease (HR 0.93), cardiovascular death (HR 0.78) and stroke (HR 0.88) [8]. A meta-analysis of 38 studies (811,069 people) found higher linoleic acid intake or levels linked to about 13% lower all-cause mortality and 11-13% lower cardiovascular and cancer mortality [9]. In the 135,335-person PURE study across 18 countries, higher polyunsaturated fat intake was linked to 20% lower total mortality (HR 0.80), though not to fewer heart attacks or cardiovascular deaths [10]. A 2025 analysis of 172,891 people found higher blood linoleic acid linked to lower stroke risk, while some breakdown products of linoleic acid were linked to higher heart-disease risk in one cohort [11]. Observational studies can show links, not cause and effect.

The claim that seed oils cause inflammation is not supported by randomised trials. A 2017 meta-analysis of 30 trials (1,377 people) found no significant effect of higher linoleic acid intake on C-reactive protein, interleukin-6, TNF, fibrinogen or adhesion molecules; a rise in CRP was suggested only among people whose intake rose the most [12]. An earlier review of 15 trials found no significant effect on any inflammatory marker tested [13]. A 2026 review of the controversy concluded that claims about inflammation, oxidative stress and harmful manufacturing by-products lack scientific support [1].

Two re-analyses of old trials are the main counterweight. In the Sydney Diet Heart Study, 458 men who had recently had a heart attack had higher death rates when their saturated fat was replaced with safflower oil (17.6% versus 11.8%); adding the recovered data to a meta-analysis of linoleic-acid trials showed non-significant trends toward more heart-disease and cardiovascular deaths [14]. Recovered data from the Minnesota Coronary Experiment (9,423 people) found cholesterol fell on the corn-oil diet but there was no mortality benefit, and deaths were higher per unit of cholesterol lowering [15]. A 2010 meta-analysis that split trials by the type of polyunsaturated fat found that omega-6-only interventions did not reduce heart attacks or heart deaths, while diets that raised both omega-6 and omega-3 did (RR 0.78) [16].

Doses used in studies

In the trials pooled by Mozaffarian and colleagues, the polyunsaturated-fat intervention supplied about 15% of daily energy versus about 5% in controls, for at least a year [6]. Cochrane's omega-6 review covered 19 trials of one to eight years in which omega-6 fats usually replaced saturated or monounsaturated fat [2]. In the LDL feeding trials, the higher-polyunsaturated diets supplied about 10.6% of energy from polyunsaturated fat (median polyunsaturated-to-saturated ratio 1.2) versus 4.4% (ratio 0.4), for at least three weeks [3]. In the 2020 Cochrane review of saturated-fat reduction, trials lasted at least two years, and 56 people would need to reduce saturated fat for about four years for one person to avoid a cardiovascular event [5].

Safety and interactions

In 19 randomised trials that increased omega-6 intake for one to eight years, there was no reported increase in deaths, heart attacks or strokes [2]. Higher-polyunsaturated diets that supplied around 15% of daily energy for years did not increase heart attacks or heart deaths [6]. Fully refined seed oils contain only trace amounts of protein, but a laboratory analysis detected peanut proteins able to bind allergic antibodies in some peanut-oil extracts; people with severe soy or peanut allergy should ask a doctor which oils are safe for them [17]. Repeatedly heating polyunsaturated oils at high temperatures forms oxidation products; concerns about these compounds come mostly from laboratory and animal studies, and the reviewers called for human trials [18]. Concerns that seed-oil manufacturing itself creates harmful substances were not supported by a 2026 review of the human evidence [1]. Trials have used polyunsaturated fat intakes averaging about 15% of daily energy, up to about 21% in some studies, for a year or more, and no toxic level has been identified in these human trials [6]. No studies in this review tested seed-oil diets during pregnancy or breastfeeding, and no medicine interactions with seed oils at food amounts were found in the research searched here.

Caveats and open questions

Cochrane's omega-6 review suggested the protection was greatest in people whose baseline omega-6 intake was lowest [2]. In one large cohort, some breakdown products of linoleic acid were linked to higher heart-disease risk even though linoleic acid itself was linked to lower overall risk [11]. For cancer, an umbrella review of 150 cohort papers found higher omega-6 intake linked to lower overall cancer incidence but higher risks of ovarian and endometrial cancers [19]. The 2026 review of the saturated-fat trials noted that trials varied a lot in how much saturated fat they removed and what replaced it, and called for new trials to clarify the uncertainty [7].