Elevated homocysteine: what the value means and what lowers it

Many people know this pattern. A value is flagged on the lab report. Next to it: homocysteine.

In the evening you search for it. The first result talks about heart attack, the second about stroke, the third about dementia. And the pack of B vitamins is already in your shopping basket.

I understand this reflex: a value is too high, so it should come down. But with homocysteine, exactly this logic has hit its limit in large trials. In randomised trials with more than 70,000 participants combined, B vitamins were tested for lowering the value, and in the large individual trials it fell in some cases by more than a quarter. The heart attack rate did not fall with it.

The value is not useless. It simply asks a question that needs a different answer than a tablet.

What homocysteine is and why it is produced at all

Imagine a railway shunting yard. Trains come in, are recoupled and leave again. As long as the points are clear, nobody notices anything.

Homocysteine is one of those trains. It is not a poison from outside but an intermediate product that is constantly made in your cells, from methionine, an amino acid from dietary protein.

So there are two exits. One needs folate and B12, the other B6. If a set of points jams, the trains back up. The lab measures how many trains are standing in the yard, not which set of points is jammed.

On top of that, the value can rise when kidney function declines, and also with an underactive thyroid, and some medications can interfere with folate or B12 balance. Homocysteine is therefore a functional summary marker. It pulls many things together and points to nothing in particular.

The other tasks of the B vitamins are described in B vitamins: what each one does.

And now you know why an elevated value does not automatically mean vitamin deficiency, but first of all a search.

Elevated homocysteine: the causes, from common to overlooked

Many guides jump straight to vitamins when the value is elevated. That is only part of the list. Not every cause applies to you, but each one is worth considering.

A deficiency of B12, folate or B6

This is the classic cause. If B12 is missing, methionine synthase can stall. If folate is missing, the methyl group can be missing. A B12 deficiency can have many roots: a diet without animal foods, impaired absorption, older age, certain medications. If you eat a vegan diet, you need a reliable source of B12, more on this in Vegan diet: opportunities and limits.

Homocysteine can rise with B12 deficiency, but it is not a specific B12 test. How a deficiency is assessed in stages, with holotranscobalamin and methylmalonic acid, is covered in Measuring vitamin B12 deficiency correctly.

The kidneys: common and often overlooked

Homocysteine is closely tied to kidney function, and mildly reduced kidney function causes no symptoms for a long time.

The thyroid

A second, often overlooked possible cause can be an underactive thyroid. A team led by Catargi studied 40 people with hypothyroidism, 14 with autoimmune thyroiditis and 26 after thyroid cancer. Fasting homocysteine was 13.0 ± 7.5 µmol/l compared with 8.5 ± 2.6 µmol/l in controls, and TSH was the strongest predictor, independent of age, folate, B12 and creatinine. With hormone replacement the fasting value fell significantly, without returning fully to the normal range (Catargi 1999). In the subclinical form, the effect is small: a meta-analysis of twelve observational studies found on average 1.16 µmol/l more (Zhang 2020).

Which thyroid values say something, and when, is covered in Thyroid values: which ones really count. Ongoing treatment with thyroid hormones is not changed because of homocysteine.

Medications: four examples, honestly weighted

Important: None of these medications is stopped, reduced or replaced because of a homocysteine value, and stopping abruptly carries its own risks. If you suspect an effect, discuss it with the doctor who prescribed the medication. On metformin, the HOME trial argues for B12 checks, not against the medication. On acid blockers: Understanding heartburn and acid blockers.

Age, smoking, alcohol and coffee

Part of the value has to do with things no lab can see. In the Hordaland study of 16,176 people from western Norway, the geometric mean at age 40 to 42 was 10.8 µmol/l in men and 9.1 µmol/l in women, and at age 65 to 67 it was 12.3 and 11.0 µmol/l. It rose markedly with the number of cigarettes smoked, and more physical activity went along with lower values (Nygård 1995). These are population averages, not reference ranges.

With alcohol, even moderate amounts can shift the value. A team led by Gibson had 78 healthy men drink 24 grams of alcohol a day for two weeks each, as red wine or as vodka. After the red wine, homocysteine was 5 percent higher (P=0.03), after the vodka 3 percent higher (not significant), and serum B12 and folate fell significantly (Gibson 2008). What alcohol can change in the gut is covered in Alcohol and the gut.

More on coffee and hormones: Coffee, cortisol and adenosine. And stress? I have not found a robust human study on stress as an independent cause.

Genetics: MTHFR, honestly assessed

Hardly any lab topic is as charged online as the MTHFR variant C677T. People who carry it twice, the TT genotype, make a less stable form of the enzyme. The crucial question is how much this matters in everyday life.

For the heart, the data are even more sobering. In large unpublished genotype datasets with 48,175 cases and 67,961 controls, the odds ratio for coronary heart disease with TT compared with CC was 1.02, whereas the published studies had shown 1.15. The authors see this as publication bias or methodological problems and conclude that a lifelong moderately elevated homocysteine concentration has little or no effect on coronary heart disease (Clarke 2012). The American genetics society ACMG advises against testing for the variant in a routine work-up for a tendency to thrombosis (Hickey 2013). The debate about active folate forms is covered in Folate instead of folic acid. This must be kept separate from rare inherited disorders such as homocystinuria with very high values, which belong in a specialised metabolic work-up.

And now you know why an elevated homocysteine value is not followed first by a tablet, but by a list of questions.

Homocysteine and heart attack: the paradox, told in full

This chapter is the heart of the article, a story in three acts. It explains why cardiology and neurology are cautious today, and that caution has good reasons.

Act one: the hope

Around the year 2000 it looked like a breakthrough. People with higher homocysteine had more heart attacks and strokes, and the value could be lowered with cheap vitamins.

In the same year, a team led by Wald concluded from genetic and prospective studies that there was a causal relationship and calculated that 3 µmol/l less homocysteine could mean 16 percent less coronary heart disease and 24 percent fewer strokes (Wald 2002). That was the expectation: lowering should protect.

Act two: the large trials

So it was tested, in tens of thousands of people, over years, randomised and mostly against placebo. All doses in the table are study data and not recommendations.

Read the last column slowly. The value fell, in some cases by more than a quarter. None of these trials showed protection against heart attack. In VISP, a lower baseline value predicted fewer events, but stronger lowering changed nothing: the paradox in miniature.

Act three: the balance

According to Holmes and colleagues, up to 2011 no trial had tested lowering exclusively in a region with low folate supply. Whether the benefit depends on folate status remains open.

How I read the paradox

More homocysteine, more heart attacks, yet lowering it did not reduce heart attacks in the large trials. Both can be true at the same time. One plausible interpretation is that in heart attack, homocysteine is more of a marker than a main cause. People with reduced kidney function, who smoke, are older or move little often have higher homocysteine and at the same time a higher cardiac risk, which may depend on these factors themselves. The genetic data from Clarke 2012 fit this picture. For stroke, a small benefit could exist, most likely with low folate status, and here two serious camps disagree.

How I think about cardiovascular risk factors overall is covered in Cholesterol: what the science shows.

Kidneys, diabetes, stent: where high doses did not help or showed harm

If homocysteine is highest in kidney disease, people with kidney disease should benefit most from lowering it. That was the logic. In the HOST trial led by Jamison, 2056 people with advanced or end-stage kidney failure and homocysteine of 15 µmol/l or more received 40 mg folic acid, 100 mg pyridoxine and 2 mg cyanocobalamin daily, or placebo; these are study data. The value fell by 25.8 percent, there were 448 deaths compared with 436, hazard ratio 1.04 (Jamison 2007). In 4110 kidney transplant recipients in the FAVORIT trial, the cardiovascular endpoint also remained unchanged with a high-dose compared with a low-dose vitamin combination, hazard ratio 0.99 (Bostom 2011).

And now you know why the honest answer is not: lowering protects. But rather: the value is a pointer, and when it comes to protection, it is worth looking at what it points to.

Homocysteine and the brain: what VITACOG showed, and for whom

Perhaps it is not your heart that worries you, but forgetting. The name that no longer comes to mind. The parent with dementia.

There is a frequently cited study from Oxford here, VITACOG. It is interesting and deserves to be read with all its conditions.

Two further analyses sharpen the picture. Benefits in cognitive tests appeared mainly with a baseline value above the median of 11.3 µmol/l (de Jager 2012). And in brain regions affected early in Alzheimer's disease, shrinkage was up to sevenfold lower, again only with homocysteine above the median (Douaud 2013). That is a regional peak value.

The limitation that is often missing

The counterweight: a team led by Clarke pooled eleven randomised trials with around 22,000 people who had not been specifically selected for memory problems. Homocysteine fell by 26 to 28 percent, no significant effect on cognition emerged, and an effect of more than one month of cognitive ageing per year of treatment could be ruled out (Clarke 2014).

An international expert group led by Smith and Refsum nevertheless considers elevated homocysteine a modifiable risk factor for cognitive decline and dementia in older age (Smith 2018). Several authors of this consensus were also responsible for VITACOG. This is a serious expert opinion, not a guideline.

VITACOG measured omega-3 in plasma, not the omega-3 index. How the omega-3 index works as a blood value is covered in Measuring the omega-3 index, and why EPA and DHA are not the same as plant-based ALA in ALA, EPA and DHA. In the analysis by Oulhaj, DHA alone significantly strengthened the effect, EPA less so.

And now you know why the question is not whether B vitamins are good for the brain, but for whom and under which conditions.

Pregnancy, bones, mood: three further fields

In three further fields the pattern is similar: an association in observational data, and when it comes to interventions, the evidence gets thinner.

Pregnancy and wanting to conceive

A team led by Vollset linked the homocysteine values of 5883 Norwegian women, measured at age 40 to 42, with 14,492 pregnancies from the birth registry. In the top compared with the bottom quarter, the odds ratio for preterm birth was 1.38 and for very low birth weight 2.01. For pre-eclampsia it was 1.32, but the confidence interval included 1. With values above 15 µmol/l, placental abruption was more frequent, odds ratio 3.13 (Vollset 2000). The methodological weakness: measurements were often taken years to decades after the pregnancies. Cause and effect therefore cannot be separated.

Independent of homocysteine, there is a recommendation here with high certainty of a substantial net benefit. The US Preventive Services Task Force recommends that all persons planning or capable of pregnancy take a daily supplement containing 0.4 to 0.8 mg of folic acid to prevent neural tube defects, as a grade A recommendation (USPSTF 2023). This is a US guideline, and nothing in this article is a reason to deviate from this preventive measure. The debate about folate versus folic acid is covered in Folate instead of folic acid, and the decision belongs with your gynaecological care.

Bones

Two large cohort papers in the New England Journal of Medicine showed the same pattern in 2004. In 2406 people aged 55 and over, the relative risk of osteoporotic fractures per standard deviation of homocysteine was 1.4, independent of bone density (van Meurs 2004). In the Framingham study of 1999 older people, the risk of hip fracture in the top compared with the bottom quarter was almost four times as high in men and 1.9 times as high in women (McLean 2004).

A note, because it is often missing online: a widely cited 2005 JAMA study on folate, mecobalamin and hip fractures after stroke, which often circulates as evidence of fracture protection, was retracted (PMID: 15741530). It is not used as evidence here and is not counted.

Mood and depression

First, the most important point: depression is a serious illness and not a passing low mood. It cannot be traced back to a lab value. Nobody who develops depression is to blame for it, neither through their diet nor through their vitamin intake. Looking at homocysteine comes as a complement to psychiatric or psychotherapeutic treatment, never in its place.

There is an association nonetheless. In 3752 men aged 70 and over, the odds of depression rose by 4 percent per µmol/l of homocysteine, and in the accompanying meta-analysis the odds ratio for high homocysteine was 1.70 (Almeida 2008). A later meta-analysis found on average 2.53 µmol/l more homocysteine in depression, with the diagnostic instrument being one source of differences between studies (Moradi 2021).

An antidepressant is neither stopped nor reduced because of a homocysteine value. Stopping abruptly carries its own risks, and any change belongs in the hands of the doctor who prescribed it. If you have suicidal thoughts, you can reach the Telefonseelsorge on 0800 111 0 111 and 0800 111 0 222, and in acute danger call 112. These are German numbers. Outside Germany, please use your local crisis line and emergency number. How I see depression holistically, without questioning established treatment, is covered in Treating depression holistically.

And now you know why an elevated value in these fields raises a question, but does not deliver a ready-made answer.

Measuring homocysteine: fasting, cooled, lab specific

Imagine your value is suddenly much higher at the follow-up test. Has your metabolism changed that much in three weeks? Sometimes. But often the difference lies in the blood draw itself.

Why there is no universally valid normal value

The range of 5 to 15 µmol/l describes study populations and is not a normal value for you. Because exactly the things that shift the value differ between laboratories and populations:

• The measurement method. The British haematology guideline considers fixed cut-offs impossible because methods differ, and calls for local reference ranges (Devalia 2014).
• Age and sex. The value rises with age and is on average higher in men (Nygård 1995).
• The folate supply of the population. It partly determines how strongly genetics shifts the value (Holmes 2011).
• Sample handling. Time to centrifugation and temperature change the result (Nygård 1999).

An international expert group led by Refsum has published recommendations on methods, sampling, reference intervals and areas of use for homocysteine measurement (Refsum 2004). Standardised sampling and lab specific interpretation are therefore consensus.

That is why this article deliberately contains no table of normal values and no target value. The reference range that applies to you is on your lab report. What a value means emerges in a conversation with your doctor, together with your other findings. Deriving your own supplement intake from a single number is exactly the step the studies in this article warn against.

When the value is elevated, doctors often look at kidney function, thyroid, B12 status and folate. This is not a list for ordering tests yourself, but a direction. How B12 status is determined in stages is covered in Measuring vitamin B12 deficiency correctly, and what whole blood compared with serum can show for micronutrients in Micronutrient analysis.

And now you know why the question of the right value always begins with the question of how it was measured.

Lowering homocysteine: what the studies show, and what carries risks

What lowers the value? The short answer: whatever raised it. The long answer contains a part that rarely appears in advertising copy, the risks.

The cause comes first

If reduced kidney function is behind it, the kidneys are the issue, not the pack of vitamins. With an underactive thyroid, its treatment went along with a lower fasting value in the study by Catargi. If a medication affects B12 or folate balance, the prescribing doctor decides on checks or accompanying treatment. And if smoking, a lot of alcohol or a litre of coffee a day are involved, that is a starting point that can change more than a lab number. This is the consequence of the paradox: lowering only the marker leaves the cause untouched.

B vitamins for a confirmed deficiency

A confirmed deficiency of B12 or folate deserves treatment in its own right. How strongly B vitamins can then lower the value is shown by a dose response analysis, whose doses are study data and not a recommendation.

In the large trials, the reduction was measurable after weeks to months, in HOST for example after three months at 25.8 percent (Jamison 2007). When a follow-up test makes sense is decided by your doctor. On the form: in a randomised trial with 167 healthy people, a low dose of active folate, L-5-methyltetrahydrofolate, lowered homocysteine by 14.6 percent compared with placebo after 24 weeks, and the equimolar amount of folic acid by 9.3 percent (Venn 2003). The whole debate is covered in Folate instead of folic acid.

Why high-dose intake without a reason carries risks

B vitamins are water soluble, so any excess is simply excreted, people often say. The studies paint a different picture.

The house rule I derive from this: B vitamins in a targeted way for a confirmed deficiency and with medical supervision, not across the board and in high doses to lower a number. Why supplements make the most sense when they make up for a proven deficit is covered in When supplements make sense.

This does not refer to ongoing treatment. If you are being treated by a doctor for a confirmed B12 or folate deficiency, do not stop this treatment on your own because of these studies, since untreated B12 deficiency can damage nerves permanently. Folic acid supplementation when planning a pregnancy and during pregnancy also rests on its own recommendation basis (USPSTF 2023). The risk signals described here are no reason to deviate from it, and the decision belongs with your gynaecological care.

And diet? Folate is abundant in leafy greens, legumes, cabbage and liver. B12 is found almost exclusively in animal foods, in meat, eggs, offal and dairy products, more on this in Offal and bone broth. This is a direction from biochemistry. Studies testing a dietary change via homocysteine against hard endpoints were not available to this research.

How I interpret an elevated homocysteine value

And now you know why lowering homocysteine does not begin with a pack of tablets, but with a question.

Frequently asked questions about homocysteine

Where homocysteine connects to other topics