Household Ingredients Compared › Guides
"Good cholesterol" sounds like something you can eat your way to more of — a food that raises it, an ingredient that boosts it. That is the premise this page starts by correcting, because the evidence does not support it. HDL — the "good" cholesterol — is a marker, not a lever. People with higher HDL tend to have fewer heart events, but raising the HDL number itself has repeatedly failed to prevent any. The drugs designed to do exactly that were tested in tens of thousands of people, and they did not work. So this page is not about ingredients that raise your good-cholesterol number. It is about the ingredients that support a healthy cholesterol profile overall — mostly by lowering LDL ("bad") cholesterol, and by supporting HDL as part of an overall healthy pattern — with the amounts where the evidence gives one, and the honest limits where it does not.
This page does not score or rank anything, and it does not tell you what to eat. It explains labels and evidence. For personal dietary advice — especially if you are on cholesterol medication — talk to a doctor or dietitian. Nothing here replaces prescribed treatment.
Three things it is for:
HDL (high-density lipoprotein) carries cholesterol away from the body's tissues and back to the liver, where it can be removed — the process called reverse cholesterol transport. That is a genuinely useful job, and it is why higher HDL is associated with lower heart risk in population studies: people with higher HDL have fewer heart attacks. But an association is not a mechanism you can push. The question the last twenty years of trials asked was whether raising HDL prevents heart events — and the answer, repeatedly, was no.
The evidence, in order:
The lesson the cardiology literature draws from all of this is that HDL function matters, not the HDL number — and that the number on a blood test is a readout of an overall pattern (diet, exercise, weight, smoking, genetics), not a dial any single ingredient turns. The levers that do reliably raise HDL are not ingredients at all: regular aerobic exercise, maintaining a healthy weight, and not smoking are the ones public-health guidance names. This page is about the ingredients, so it covers what ingredients can do — which is mostly about LDL, and about triglycerides — and it does not pretend any of them is an HDL booster.
Each entry says what it does, how it works, how much the evidence uses, and where it actually appears on this site — nothing here is invented, and every example is linked to the page that panelled it.
What it does: lowers LDL cholesterol. It does not raise HDL — the trials found no effect on HDL or triglycerides.
How it works: beta-glucan is a viscous soluble fibre. In the gut it forms a gel that traps cholesterol-rich bile acids, so they are excreted instead of reabsorbed. The liver then has to pull LDL out of the bloodstream to make new bile acids, and blood LDL falls.
How much: about 3 g of oat beta-glucan a day. That is the dose in both authorised health claims — the FDA's (1997) and EFSA's (Regulation 432/2012, following its 2010 opinion). The meta-analyses agree on the size of the effect: Whitehead et al. 2014 (28 randomised controlled trials, at least 3 g/day) found LDL about 0.25 mmol/L lower than control; Ho et al. 2016 (58 trials) found about 0.19 mmol/L lower. That is roughly a 5% LDL reduction from a breakfast food, sustained daily. About 75 g of dry rolled oats — a large bowl — delivers roughly 3 g of beta-glucan, though the content varies by oat type and processing, so that figure is approximate, not a guarantee for any one pack. Psyllium husk is the other viscous fibre with the same mechanism and an authorised claim; its dose is not stated here because it was not verified in this pass.
On this site: Quaker Instant Oatmeal 400g lists a single ingredient — "100% Australian Wholegrain Oats" — see oats and instant oatmeal. The plain oats are the example; the flavoured 3-in-1 oat drinks on the same page add sugar and flavouring, which is a different product. Note also that the beta-glucan claim is about oats, not about oat milk: the plant-milks page shows oat drinks are mostly water with a small percentage of oats, so an oat drink is not a reliable way to reach 3 g.
The full section is in the fats and cholesterol guide; this is the one-paragraph version. Plant sterols and stanols are compounds from plants that look enough like cholesterol to block its absorption in the gut. The meta-analysis (Ras et al. 2014, 124 randomised controlled trials) found 2.0–2.5 g a day lowers LDL by about 8.4% on average, with the effect levelling off above about 3 g/day; the EFSA-authorised claim covers 1.5–2.4 g a day taken with meals. The honest limit, stated in the fats guide too: no long-term outcome trial has shown that sterols themselves prevent heart attacks or strokes — the evidence is a surrogate win (lower LDL), not an events win. They also do not raise HDL.
On this site: the sterol products that are actually panelled are milks, not spreads. The fresh-milk page carries Magnolia Cholest Care (plant stanol esters) and a second cholesterol-care milk listing plant sterols — see fresh milk. No sterol-fortified spread was panelled on the margarine page — the "cholesterol free" spreads there (Naturel Cholesterol Free, for example) are a different claim entirely, and the margarine page explains why: any spread made only from vegetable oils is cholesterol-free by definition, and Naturel Soft still contains palm fraction and skimmed milk powder. If a sterol spread is on the shelf, it needs a pack check before it is claimed here. And check the dose on the milk: the claim requires roughly 2 g a day of sterols or stanols, taken with meals — a single small glass may not deliver it.
What it does: replacing saturated fat with unsaturated fat lowers LDL. The key word is replacing: adding olive oil on top of butter does not help; cooking with olive oil instead of butter does. The AHA's 2017 advisory and the WHO's 2023 guideline both rest on this — the evidence is summarised in the fats guide, and the mechanism (less saturated fat in, less LDL out) is the same one that guide explains.
How much: there is no single dose, because this is a swap across the whole diet, not a supplement. The trial that tested it as a pattern is PREDIMED: 7,447 Spanish adults at high cardiovascular risk ate a Mediterranean diet supplemented with either extra-virgin olive oil (about 1 litre a week) or 30 g a day of mixed nuts (15 g walnuts, 7.5 g almonds, 7.5 g hazelnuts), against a low-fat control, for a median of 4.8 years. The nut arm had 28% fewer major cardiovascular events (hazard ratio 0.72) and the olive-oil arm 31% fewer (0.69). Two cautions: the trial tested a whole diet pattern, not nuts or oil alone, so the benefit cannot be credited to one ingredient; and the original 2013 report was retracted and republished in 2018 after protocol corrections — the conclusions held in the revised analysis, and this page cites the 2018 version. A separate meta-analysis of 61 controlled trials (Del Gobbo et al.) found tree nuts in general lower LDL; walnuts, almonds, pistachios, pecans and hazelnuts showed the effect, while cashews and peanuts did not reach significance.
On this site: Bertolli Extra Virgin olive oil — 13 g saturated fat per 100 g on the Singapore listing, against 52 g for a block of butter — and Cobram Estate Classic, see olive oil. For nuts, the mixed-nuts page has Camel Roasted Almonds 400g (almonds, one ingredient) and Planters Lightly Salted Mixed Nuts, see mixed nuts — but read the list: the honey-butter almond on the same page carries sugar, corn starch syrup and honey, so "nuts" and "sweetened nuts" are different products. And the swap matters more than the nut: a handful of plain nuts instead of a processed snack is the pattern the trials tested, not nuts on top of everything else.
What it does: mainly lowers triglycerides, not LDL. This is the most mis-sold point on the page: fish oil is marketed as a cholesterol product, but its best-documented lipid effect is on triglycerides. The AHA's 2019 science advisory (Skulas-Ray et al., Circulation) states that 2–4 g a day of EPA+DHA lowers triglycerides by roughly 20–30% in people with elevated levels — and that those are prescription-level doses, not food doses. At food levels, the guidance is simpler: eat oily fish about twice a week.
The drug evidence, kept separate from the food evidence: REDUCE-IT (8,179 statin-treated patients with elevated triglycerides, 2018) found 4 g a day of icosapent ethyl — purified EPA, a prescription drug — cut major cardiovascular events by about 25%. STRENGTH (13,078 patients, 2020), testing 4 g a day of combined EPA+DHA, did not show a benefit. The two trials disagree, and the difference (EPA alone vs EPA+DHA, and the placebo used) is still argued about. Neither trial is evidence for a fish-oil capsule from a supermarket shelf: the doses are 4 g a day of purified product, taken by high-risk patients under a doctor's care.
ALA is not EPA/DHA. Alpha-linolenic acid, the omega-3 in walnuts, flaxseed and chia, is a different molecule. The body converts only a small fraction of ALA into EPA and DHA, so a walnut is not a substitute for oily fish on this measure — walnuts earn their place through the unsaturated-fat and fibre evidence above, not through omega-3.
On this site: Atlantic salmon fillet (Norway), see fresh salmon; Ayam Brand Brisling Sardines in Extra Virgin Olive Oil, see canned sardines; Ayam Brand Mackerel Fillet in Extra Virgin Olive Oil 125g, see canned mackerel and other canned fish. The fish is the point; the sauce is not — the tomato-sauce variants on the same pages add salt and sugar, so check which can you are holding.
What it does: a modest LDL reduction, and the size of it is the argument. This is the one ingredient on this page where the regulators have gone back and forth, so both sides are stated.
For the claim: a 1995 meta-analysis found about a 13% LDL reduction when 50 g of soy protein a day replaced animal protein, and on that basis the FDA authorised a health claim in 1999 — "25 g of soy protein a day, as part of a diet low in saturated fat and cholesterol, may reduce the risk of heart disease." A cumulative meta-analysis by Jenkins et al. (2019), re-analysing the same 46 trials the FDA reviewed, found the LDL reduction — about 4–6 mg/dL — stayed statistically significant at every point from 1999 to 2013, and argued the claim should stand.
Against the claim: the AHA's 2006 advisory concluded the effect was small relative to the dose needed for a clinical benefit. In 2017 the FDA proposed revoking the authorised claim, saying studies since 1999 had presented inconsistent findings and that significant scientific agreement had not been reached; the proposal was still pending as a final rule at the time of writing. EFSA concluded that a cause-and-effect relationship has not been established between soy isoflavones and maintaining healthy LDL cholesterol. And the dose is the practical problem: 25 g of soy protein is roughly 680 g of tofu or nearly 2 litres of soy drink a day — far more than a normal serving.
On this site: Fortune Japanese Silken Tofu 300g and Unicurd Tau Kwa 400g, see tofu and tempeh; NutriSoy No Sugar Added (soya beans, calcium carbonate, stabiliser, vitamin D3, flavouring — five items) and Vitasoy Soya Bean Drink, see plant milks. Note the plant-milks page's warning: the Australian and European UHT soy drinks are water-led with 8.7% or less soya, while the Singapore and Asian fresh soya drinks are soya-bean-led — the same "soy milk" name covers very different products, so the protein per glass differs.
What they do: legumes (beans, chickpeas, lentils) and whole grains bring soluble fibre, and — just as importantly — they displace foods higher in saturated fat. There is no single dose for "eat more beans", because the evidence is at the level of dietary patterns: the trials that show benefit test a pattern rich in these foods, not a gram target for one of them. The mechanism is the same soluble-fibre one as oats, plus replacement: beans instead of processed meat is a swap in the same sense as olive oil instead of butter.
On this site: Ayam Brand Mixed Beans (50% beans — navy beans, chickpeas, red kidney beans), Biona Red Kidney Beans and Biona Black Beans (beans and water, no salt — the plainest panels on the page), see canned beans and legumes. One trap on that page: baked beans in tomato sauce are a different product — sugar and tomato paste are in the list, so a baked bean is not a plain bean. And the oats on the oats page belong here too — the same beta-glucan, in a whole-grain food rather than a supplement.
Only where the evidence gives a dose. Where it does not, the table says so.
| Ingredient | What it moves | Dose where established | Evidence |
|---|---|---|---|
| Oat beta-glucan | LDL down (~5%) | 3 g/day (≈ 75 g dry oats, approximate) | Authorised claims (FDA, EFSA Reg 432/2012); Whitehead 2014, 28 RCTs; Ho 2016, 58 RCTs. No HDL effect. |
| Plant sterols / stanols | LDL down (~8.4%) | 1.5–2.4 g/day with meals (EFSA claim); 2.0–2.5 g/day in the meta-analysis | Authorised claim; Ras 2014, 124 RCTs. No outcome trial showing fewer heart events. |
| Mixed nuts (walnuts, almonds, hazelnuts) | LDL down; fewer events as part of a pattern | 30 g/day (15 g walnuts, 7.5 g almonds, 7.5 g hazelnuts) | PREDIMED 2018, 7,447 adults, 4.8 years — whole-diet trial, not nuts alone. |
| Oily fish (EPA/DHA as food) | Triglycerides down; pattern-level heart benefit | Oily fish about twice a week (food guidance); 2–4 g/day EPA+DHA is prescription territory for triglycerides | AHA 2019 advisory. REDUCE-IT and STRENGTH disagree on purified high-dose EPA/DHA. |
| Soy protein | LDL down modestly (~4–6 mg/dL) | 25 g/day in the 1999 FDA claim — contested; FDA proposed revoking the claim in 2017 | Contested. Jenkins 2019 re-analysis vs FDA 2017 proposal and EFSA. |
| Legumes and whole grains | LDL down as part of a pattern | No single dose established — pattern-level evidence | Soluble-fibre mechanism (as oats); displacement of higher-saturated-fat foods. |
| Unsaturated fats (olive oil, nuts) | LDL down when they replace saturated fat | No single dose — a swap across the diet | AHA 2017 advisory; WHO 2023 guideline; PREDIMED. |
Doses are as the cited evidence gives them. "Approximate" means the figure varies by product and is not a guarantee for any one pack. Nothing in this table is a recommendation for any individual — if you are on cholesterol medication, these foods sit alongside it, not instead of it.
Two labelling schemes matter for a Singapore shopper reading a cholesterol claim, and they are different things. The Healthier Choice Symbol (HCS), run by the Health Promotion Board, is voluntary and category-specific: a product earns it by meeting HPB's nutrient guidelines for its food category, and carries a tagline for what it does — "Lower in Saturated Fat" means at least 25% less saturated fat than similar products in the same category; "Higher in Wholegrains" means at least 20% more wholegrains; "Lower in Sodium" means at least 25% less sodium. Products with the symbol are generally lower in saturated fat, sodium and sugar and higher in dietary fibre, calcium and wholegrains than similar products in the same category — similar products is the key phrase, because the comparison is within a category, not across the whole shop. A "Lower in Saturated Fat" instant noodle is still an instant noodle.
Nutri-Grade is the mandatory front-of-pack grading for beverages, A to D by sugar and saturated fat per 100 ml: grade A is sugar at most 1 g and saturated fat at most 0.7 g per 100 ml; grade D is sugar above 10 g or saturated fat above 2.8 g. The grade is the worse of the two. It is being extended to cooking oil, with effect in 2028 as announced — re-check that date before publish, as the announcement has already moved once. Neither scheme is a cholesterol claim: a product can be grade A or carry the HCS and still have nothing to do with LDL. The schemes tell you about sugar, saturated fat, sodium, fibre and wholegrains; the cholesterol evidence on this page is a separate question, and the two should not be confused.
Where to check: HPB's Healthier Choice Symbol nutrient guidelines (as of July 2025) carry the category-by-category thresholds; HealthHub carries the tagline definitions. This page does not reproduce the full guidelines — they are category-specific and HPB revises them.
Three checks, in order, for any product making a cholesterol-related claim:
Pages where these checks are most useful: oats and instant oatmeal (beta-glucan dose), fresh milk (sterol/stanol milks), margarine ("cholesterol free" vs sterol claims), mixed nuts (plain vs sweetened), olive oil (extra virgin vs blends), fresh salmon, canned sardines and canned mackerel and other canned fish (oily fish), tofu and tempeh and plant milks (soy), and canned beans and legumes (plain beans vs baked beans).