Why Cholesterol Rises in Menopause, and What Food Actually Does

I am a nutritionist, not your doctor, and I cannot tell you what your own result means. What I can give you is what happens to groups of women who eat in particular ways, and how large those changes are. Every figure is linked below, including the ones funded by the industry that profits from them.

Your cholesterol went up and it may not be your fault

Of every cardiovascular risk factor measured in 3,302 midlife women, the ones that moved on menopause's schedule rather than the calendar's were the cholesterol numbers.

That is the SWAN cohort, Matthews 2009.1 Of those women, 1,054 reached a natural final menstrual period without hormone therapy, which lined everyone up on menopause rather than on birthdays. Total cholesterol, LDL and apolipoprotein B accelerated distinctly in the one year window bracketing that period. Blood pressure, glucose, insulin, CRP and fibrinogen tracked chronological ageing instead, across every ethnic group.

So the thing that changed is the thing menopause moves, on a timetable unrelated to your plate. If your number climbed while your cooking stayed the same, the usual framing, that high cholesterol happens to people who eat badly, is wrong about you, and it sends you into restriction for a problem restriction will not fix.

If reading about food and health has ever been difficult for you, stop here and take one line: add more plants, change the oils you cook with, and talk to your doctor about your number. That is most of it, and none of the rest is a rule.

What your cholesterol printout actually says

Both units, because half of you read a European lab report and half an American one.

To convert, divide cholesterol in mg/dL by 38.67, and triglycerides by 88.57.

Non-HDL is already on your printout

Total cholesterol minus HDL is your non-HDL. It captures every atherogenic particle in one figure and needs neither fasting nor a second test. The 2026 ACC/AHA and multisociety guideline, in Circulation under DOI 10.1161/CIR.0000000000001423, returned LDL and non-HDL to co-primary status with numeric goals: for very high risk in secondary prevention, meaning people who already have established disease, LDL below 1.4 mmol/L (55 mg/dL) and non-HDL below 2.2 (85). Apolipoprotein B refines that for people who have met both. It has not replaced LDL.

There is no single number that means high

The same LDL is on target for one woman and double target for another, because the target moves with total cardiovascular risk: age, blood pressure, smoking, diabetes, family history. That is the honest answer to the question you typed at eleven at night.

HDL is a reading, not a lever

Almost every article here tells you to raise your good cholesterol. That instruction has been tested and it failed. Niacin trials and the CETP inhibitors, torcetrapib, dalcetrapib and evacetrapib, raised HDL substantially and cut no events, some causing harm. Mendelian randomisation agrees: low HDL marks risk, it does not cause it. Chasing it wastes effort.

Lipoprotein(a) gets one mention. It is roughly 80 to 90 percent genetically set, guidelines suggest measuring it once in a lifetime, and food does not move it. One line your kitchen was never going to change, and that is not a failure.

What food moves LDL, and by how much

The biggest failure in this category is the phrase lowers cholesterol with no number attached. That leaves a woman with an LDL of 5.5 believing porridge will get her to target. Here are the sizes.

How much does each food actually lower LDL?

Who paid for these

The larger 0.25 mmol/L figure you will see quoted elsewhere comes from Whitehead 2014, which DSM Nutritional Products supported and whose search included an oat supplier's unpublished reports,3 so the figure above comes from Ho. The tree nut figure, Del Gobbo 2015, was part-funded through the International Tree Nut Council. Two of the three sterol authors work for Unilever, which sells sterol spreads, and their paper tested neither heterogeneity nor publication bias. Sterols have no outcome trial either: the European Atherosclerosis Society panel, Gylling 2014, says there are no randomised data with hard end points,6 so every event claim is extrapolated from LDL. The 2025 almond analysis was funded by the Almond Board of California, and I kept its figure only because Lee-Bravatti 20199 and Asbaghi 202110 agree.

A lipid panel repeated sooner than about 6 to 12 weeks after you change how you eat tells you little. That is how long the picture takes to settle.

Why you cannot add them up

The table is not a shopping calculator. Add every row together and you get about 1.17 mmol/L. The trial that actually fed people that whole stack got 0.73. Some rows are the same food twice, since almonds sit inside the tree nut row above them. Others are the same mechanism twice, since oats and psyllium are both viscous fibre doing one job. And the more you stack, the harder the whole thing is to keep doing.

You do not need to turn any of it into a weight-loss plan. The Portfolio trials were weight-stable, so fibre, nuts, beans, soy foods and better oils can all be added without making your body weight the project.

The last row is that whole stack: sterols, viscous fibre, soy protein and nuts, together. Chiavaroli 2018 pooled seven trial comparisons, 439 people, 3 weeks to 6 months, foods largely supplied, on top of an already low saturated fat diet. The result was 0.73 mmol/L, GRADE certainty high. Under the best conditions anybody has built, you land at 17 percent. That is the ceiling for food alone. Jenkins, whose portfolio work anchors this, discloses grants and in-kind food from most trade bodies whose products appear in the diet, openly, which is the right way.

Saturated fat, stated exactly

LDL falls reliably. Events fall modestly. Mortality does not move.

Hooper 2020, the Cochrane review, pooled 15 trials and 56,675 participants followed at least 24 months.14 Reducing saturated fat gave combined cardiovascular events at a risk ratio of 0.83, 95% CI 0.70 to 0.98, GRADE moderate. All-cause mortality was 0.96, CI 0.90 to 1.03, cardiovascular mortality 0.95, CI 0.80 to 1.12. Cochrane's own word for the lipid changes is small. Mozaffarian 2010, swapping saturated for polyunsaturated fat, found coronary heart disease at 0.81, CI 0.70 to 0.95.15

So: swapping butter for olive or rapeseed oil lowers LDL, shows up as modestly fewer events, and has never been shown to make anyone live longer. Butter is mostly saturated fat, which is chemistry, not a verdict on you. Astrup 2020 argues the case against it is weaker than guidelines imply.16 Ference 2017, across two million participants, concludes LDL causally causes atherosclerotic disease.17 Neither is fringe. This is a different-fat approach and not a low-fat one, because the finding is about replacement.

Eggs, genuinely contested

Berger 2015 found dietary cholesterol raised LDL by 6.7 mg/dL across 14 trials, 95% CI 1.7 to 11.7, flattening above about 900 mg a day.18 Vincent 2019 puts it at 1.9 to 4.6 mg/dL per extra 100 mg.19 One large egg is about 186 mg. Both analyses were egg-industry funded. Zhong 2019, pooling six American cohorts and 29,615 people with no egg money involved, found the cardiovascular association fell from a hazard ratio of 1.06 to 0.99 once dietary cholesterol was adjusted for.20 Eggs are a vehicle, not an exposure. Response varies and part of that is genetic, but there is no agreed definition of a high responder and no clinical test, so any prevalence figure is invented.

What does not lower cholesterol, and what can hurt you

Apple cider vinegar. Hadi 2021 pooled nine small heterogeneous trials: no significant effect on serum LDL.21

Garlic. Gardner 2007 ran the best trial in the field: 192 adults with LDL of 130 to 190 mg/dL, raw garlic against powdered, aged extract and placebo, six days a week for six months, NIH funded.22 Six month LDL changes were plus 0.4, plus 3.2, plus 0.2, and minus 3.9 mg/dL for placebo. Weaker meta-analyses disagree; back the well-run one.

Coconut oil. Neelakantan 2020, in Circulation, found LDL 10.47 mg/dL higher than with non-tropical vegetable oils, 95% CI 3.01 to 17.94.23 It beats butter, as Khaw 2018 showed,24 and is still not a heart-healthy fat.

Detox protocols. Klein and Kiat 2015 wrote that no randomised controlled trials have assessed the effectiveness of commercial detox diets in humans.25 Not weak evidence. No controlled evidence at all.

Red yeast rice, the one that can hurt

Monacolin K, the active compound, is chemically identical to lovastatin. EFSA said so in EFSA Journal 2018;16(8):5368, and could not identify an intake of monacolins from red yeast rice that does not give rise to concerns, with severe reactions at doses as low as 3 mg a day. Gerards 2015 found LDL 1.02 mmol/L below placebo, no different from statin therapy, with safety assessment poor in most trials.26

The dose problem decides it. Cohen 2017 analysed 28 American brands.27 Monacolin K was undetectable in two and varied more than sixtyfold across the other 26. At the manufacturers' own recommended servings, daily intake ranged more than 120-fold, from 0.09 to 10.94 mg. Gordon 2010 tested twelve identically labelled products and found four with elevated citrinin, a mycotoxin that damages kidneys.28 Commission Regulation (EU) 2022/860 now caps monacolins under 3 mg per daily portion.

It carries the muscle and liver risks of lovastatin, the same interactions and contraindications, including pregnancy and breastfeeding, and must never be taken alongside a prescribed statin. A 2024 contamination incident in Japan involving beni-koji products and puberulic acid, from a Penicillium mould, was a manufacturing failure rather than monacolin pharmacology, and the case numbers reported then were still under investigation, so I will not repeat them.

Does any food unclog arteries

No, and no controlled trial has shown food alone reversing plaque. Ornish 1998 is the study everyone cites, and it randomised 48 patients to a 10 percent fat vegetarian diet plus exercise, stress management, smoking cessation and group support.29 No food can be credited with any part of that. DISCO-CT, Henzel 2021, added a DASH diet to medication and found no significant difference in total plaque burden, p = 0.851.30 Food lowers the LDL that drives plaque. Nothing stronger has been shown.

Where food stops and a doctor starts with high cholesterol

Familial hypercholesterolemia

Between 1 in 200 and 1 in 250 people carry it and most do not know. The threshold that raises the question is an untreated LDL above 4.9 mmol/L (190 mg/dL), or total cholesterol above roughly 7.5. Suspicion strengthens with early heart attack in the family, men under 55 or women under 65, thickened tendons at the heel or knuckles, or a grey ring at the cornea before 45. It is a receptor defect, not a willpower problem, and not a dietary one. Food shifts LDL by 10 to 20 percent; an untreated 6.5 mmol/L needs more than 50. An article that implies food can close that gap delays a diagnosis by years.

Take these to a doctor, not a recipe

• LDL at or above 4.9 mmol/L (190 mg/dL), or total cholesterol at or above 7.5
• Triglycerides at or above 5.6 mmol/L (500 mg/dL)
• Heart attack or stroke in a parent or sibling before 55 in men, 65 in women
• Chest pressure, breathlessness on exertion, or jaw, arm or back discomfort
• Diabetes, kidney disease, or a previous cardiac event
• A number that jumped with no change in how you eat, worth a thyroid check among other secondary causes
• Visible cholesterol deposits around the eyes or in the tendons
• Already taking a statin and thinking about stopping

About statins

People do stop prescribed medication after reading articles like this one, so let me be careful. Food and medication answer different questions and are not rivals. Food moves LDL by roughly 10 to 20 percent, a moderate intensity statin by 30 to 50, through different mechanisms, and their effects add. Eating well on a statin can mean a target reached at a lower dose, and it improves what a statin does not touch: triglycerides, blood pressure, energy.

I am not going to recommend a statin either, because I have no basis for that. The prescribing decision belongs to you and your doctor, and my job starts after it. Giral 2019 followed a French cohort turning 75 with no cardiovascular history and found discontinuation associated with more cardiovascular admissions.31 That study is observational and sicker people stop medication. Raise it with a prescriber rather than settling it alone.

A number like yours is usually worth a conversation with a GP, and that is as far as anybody writing on a website should go.

Questions I get most often