Thiamine

What it is

Thiamine is a water-soluble B vitamin that the body cannot make and must take in from food or supplements. Inside cells it is converted to thiamine diphosphate, which acts as a cofactor for enzymes that release energy from carbohydrates and amino acids. As a medicine it is sold as thiamine hydrochloride or mononitrate tablets, as an injection for hospital use, and as benfotiamine, a fat-soluble derivative sold as a supplement. The injection is a prescription drug licensed for treating thiamine deficiency and beriberi.

What the research says

The settled part of the evidence is narrow: thiamine corrects thiamine deficiency, and deficiency states such as beriberi and Wernicke's encephalopathy respond quickly once it is given. Beyond that the randomised evidence is thin or negative. A Cochrane review found only two randomised trials of thiamine for Wernicke-Korsakoff syndrome and concluded the evidence cannot tell clinicians what dose, route or duration to use. Meta-analyses of randomised trials in chronic heart failure and in sepsis found no benefit on the outcomes that matter. The benfotiamine trials in diabetic nerve and kidney disease are short and mixed, with the kidney trials null.

Evidence grade: Limited evidence.

How it works

Drug class: Water-soluble B vitamin given as replacement therapy; the active form thiamine diphosphate is a cofactor for several decarboxylase and transketolase enzymes

Thiamine is converted in the body to thiamine diphosphate, which is the cofactor several enzymes need to run the reactions that turn carbohydrate into usable energy, including the pyruvate dehydrogenase complex, the alpha-ketoglutarate dehydrogenase complex and transketolase in the pentose phosphate pathway. When thiamine runs out these enzymes stall, less ATP is made, lactate builds up, and the tissues with the highest energy demand - brain, heart, nerves - are hit first. (Source 1)

What it is used for

• The injection's label records that it is effective for thiamine deficiency and beriberi of either the dry or the wet type. The supporting evidence is clinical observation rather than controlled trials: reviews describe rapid, sometimes dramatic recovery after thiamine is given, but no randomised placebo-controlled trial of thiamine for beriberi was found. Evidence: limited. (Source 2)
• Thiamine has been standard treatment for more than 50 years but the randomised evidence is almost absent. The Cochrane review found two eligible trials, only one with usable data, and concluded the evidence is insufficient to say what dose, frequency, route or duration to use. Guideline doses of 200 mg to 600 mg a day intravenously rest on expert opinion, not trials. Evidence: limited. (Source 3)
• A meta-analysis of nine randomised trials in 520 patients found no significant effect of intravenous thiamine on short-term mortality, with some improvement in organ-dysfunction measures. The authors describe the total number of patients as relatively small and four of the nine trials as high risk of bias. Evidence: not-supported. (Source 4)
• A meta-analysis of seven randomised double-blind trials in 274 patients found no significant effect of thiamine on ejection fraction, left ventricular end-diastolic volume, six-minute walk distance, NT-proBNP or NYHA class. It did correct thiamine deficiency. Evidence: not-supported. (Source 5)
• The BENDIP trial found a significant difference in neuropathy symptom score between groups in the per-protocol population at six weeks but not in the intention-to-treat population, and no difference in total symptom score. A critical review concludes benfotiamine's clinical evidence is limited to short-duration symptom-based studies with no large long-term trials. Evidence: limited. (Source 6)
• A randomised placebo-controlled trial of benfotiamine 900 mg/day for 12 weeks in people with type 2 diabetes and albuminuria (n = 39 benfotiamine, n = 43 placebo) improved thiamine status but did not reduce urinary albumin excretion or the kidney injury marker KIM-1. Evidence: not-supported. (Source 7)
• The NIH Office of Dietary Supplements reports that the authors of a 2001 Cochrane review of three small double-blind trials of oral thiamine against placebo in Alzheimer's type dementia could draw no conclusions, because the trials were small and too poorly reported to combine. We could not reach the Cochrane review itself, so this is an agency account of it rather than systematic-review evidence. Evidence: unknown. (Source 8)

Interactions

• Furosemide and other loop diuretics (pharmacokinetic study): Loop diuretics increase urine output and research has linked furosemide to falls in thiamine concentration, sometimes into the deficient range, through urinary loss of thiamine. Whether giving thiamine prevents this has not been settled in clinical studies. (Source 9)
• Fluorouracil (5-FU) chemotherapy (case reports): Case reports describe beriberi or Wernicke's encephalopathy appearing during fluorouracil treatment, possibly because the drug speeds thiamine metabolism and blocks formation of the active form. (Source 10)
• Magnesium (case reports): Magnesium is needed to convert thiamine to its active form, so low magnesium can blunt the response to thiamine. Case reports describe patients whose deficiency symptoms persisted on therapeutic thiamine until low magnesium was corrected. (Source 11)
• Alcohol (pharmacokinetic study): Ethanol reduces how much thiamine the gut absorbs, lowers liver stores and interferes with converting thiamine to its active form; up to 80% of people with chronic alcoholism are reported to develop deficiency. (Source 12)
• Fermented fish, betel nut and fermented tea leaves (theoretical): These foods contain antithiamine compounds and thiaminases that reduce thiamine absorption and availability. In a refugee population, biochemical deficiency persisted despite supplements and thiamine-rich food rations, and the review attributes this to widespread consumption of these foods. (Source 13)
• Cooking water and heat (theoretical): Thiamine dissolves in water and is reduced by heat, so a meaningful amount is lost when cooking water is thrown out, and bread has 20%–30% less thiamin than its raw ingredients. (Source 14)
• Other medicines generally (label): The NIH position is that thiamine itself is not known to interact with medicines; the documented direction is the other way round, with certain medicines lowering thiamine levels. (Source 15)

Stopping it

• There is no described withdrawal syndrome from thiamine. Treatment length is judged by clinical response: a case series of 50 adults with non-alcoholic Wernicke's encephalopathy was treated with 300-600 mg intravenously twice daily for 5-10 days then oral maintenance, and the review says treatment should continue until signs and symptoms stop improving. (Source 16)
• The label's position on stopping treatment for beriberi is that after up to two weeks of injections an oral multivitamin containing 5 to 10 mg thiamine daily for one month is used to saturate body tissues, rather than stopping abruptly. (Source 17)
• If the cause of deficiency persists - continued heavy drinking, a restricted diet, malabsorption, parenteral nutrition without thiamine - the deficiency can be expected to return, because these are the conditions under which deficiency arises. (Source 18)

What goes wrong

Life-threatening anaphylaxis and deaths have been reported after intravenous or intramuscular thiamine, especially with repeated injections. (Source 19)

• Official position, Certainty not rated.
• Size: not quantified in the label.
• Who: People given parenteral thiamine.
• How long: risk is highest after repeated administration.
• Result: No rate is given. Reported reactions include collapse and death, warmth, pruritus, urticaria, weakness, sweating, nausea, restlessness, tightness of the throat, angioneurotic edema, cyanosis, pulmonary edema and gastrointestinal haemorrhage.
• Funding: not applicable.

An occasional individual may develop a hypersensitivity or life-threatening anaphylactic reaction to thiamine, especially after repeated injection. Collapse and death have been reported.

Reported anaphylaxis to a parenteral thiamine product in UK practice was rare in absolute terms: 10 reactions against a sales-derived denominator of more than five million patient-days. (Source 20)

• Expert review, not systematic, Low certainty.
• Size: 10 anaphylactic reactions over 5,431,235-6,651,947 patient-days of treatment.
• Who: Patients given Pabrinex, a parenteral thiamine-containing product, in the UK.
• How long: 14,880-16,080 years of treatment exposure.
• Result: 10 anaphylactic reactions from between 5,431,235-6,651,947 patient-days of treatment.
• Funding: not stated; the analysis used manufacturer sales and adverse incident data.

Limit of this finding: This is not an incidence rate, and the paper says so in the lines quoted: the number of doses actually given is unknown and sales data are only an approximation. The two ranges printed are also not consistent with each other - 6,651,947 patient-days is about 18,200 years of treatment, not the 16,080 years printed beside it, while the lower figure of 5,431,235 days does work out at 14,880 years. Treat these as an order of magnitude from voluntary adverse-incident reports set against sales figures, not as a measured rate.

available data indicate that there have been 10 anaphylactic reactions to Pabrinex from between 5,431,235-6,651,947 patient-days (14,880-16,080 years) of treatment.

Untreated thiamine deficiency can kill through heart failure and can leave permanent memory and cognitive damage. (Source 21)

• Expert review, not systematic, Low certainty.
• Size: not applicable.
• Who: People with thiamine deficiency disorders.
• How long: not applicable.
• Result: No rate given. The review states that untreated thiamine deficiency disorders can rapidly lead to death from heart failure and that acute Wernicke's encephalopathy can progress to chronic Wernicke-Korsakoff syndrome.
• Funding: not stated.

If not promptly treated, TDDs can rapidly lead to death from heart failure and irreversible neurologic implications.

Without treatment up to a fifth of people with Wernicke's encephalopathy die, and parenteral thiamine does not restore about a quarter of those who reach the chronic Korsakoff stage. (Source 22)

• Official position, Certainty not rated.
• Size: not stated.
• Who: People with Wernicke-Korsakoff syndrome, mostly with chronic alcoholism.
• How long: not applicable.
• Result: Without treatment, up to 20% of people with Wernicke's encephalopathy die. At the chronic stage, parenteral thiamin treatment does not lead to recovery in about one-quarter of patients.
• Funding: not applicable.

Limit of this finding: One sentence on the NIH page this is quoted from is clinically wrong, and nothing we assert here repeats it: the page says Wernicke's encephalopathy 'is usually characterized by peripheral neuropathy'. It is not. Wernicke's encephalopathy is classically recognised by eye-movement problems, unsteady walking and confusion; peripheral neuropathy - numbness and weakness in the hands and feet - belongs to beriberi, the other main consequence of thiamine deficiency. The death and non-recovery figures quoted here are unaffected by that error.

Without treatment, up to 20% of people with Wernicke’s encephalopathy die; those who survive develop Korsakoff’s psychosis

What the evidence supports

Thiamine deficiency in high-income countries is usually driven by alcohol misuse, restricted or heavily processed diets, bariatric surgery, renal disease, eating disorders and parenteral nutrition rather than by ordinary diet. (Source 18)

• Expert review, not systematic, Low certainty.
• Size: not applicable.
• Who: People in resource-rich and resource-limited settings.
• How long: not applicable.
• Result: No effect estimate. The review lists polished white rice or cassava staples in resource-limited settings, and alcoholism, processed-food diets, renal disease, eating disorders, bariatric surgery and parenteral nutrition dependence in resource-rich countries.
• Funding: not stated.

In resource‐rich countries with fortified foods and diverse diets, thiamine deficiency is more often seen in individuals with medical comorbidities, such as alcoholism, diets restricted to heavily processed foods, renal disease, eating disorders, bariatric surgery, and dependence on parenteral nutrition.

No tolerable upper intake level has been set for thiamine because reports of harm from high oral intakes are lacking, though the Food and Nutrition Board noted excess intake could still have adverse effects. (Source 23)

• Official position, Certainty not rated.
• Size: not applicable.
• Who: General population.
• How long: not applicable.
• Result: No UL set. High intakes are defined in the passage as 50 mg/day or more from food or supplements; absorption declines rapidly above 5 mg.
• Funding: not applicable.

Because of the lack of reports of adverse effects from high thiamin intakes (50 mg/day or more) from food or supplements, the FNB did not establish ULs for thiamin ⁷.

Oral thiamine is poorly absorbed above a few mg because uptake is partly saturable, which is the stated reason parenteral dosing is used for neurological deficiency. (Source 24)

• Expert review, not systematic, Low certainty.
• Size: not applicable.
• Who: Healthy adults and lactating women.
• How long: single doses.
• Result: Absorption is active and saturable below 1 micromol/L and passive above it. Single oral doses greater than 2.5-5 mg have been shown to go largely unabsorbed, though high blood concentrations were later achieved with oral doses up to 1500 mg.
• Funding: not stated.

In human studies, single oral doses of thiamine greater than 2.5–5 mg have been shown to go largely unabsorbed,, although more recently, high blood thiamine concentrations were achieved in healthy adults with doses of oral thiamine hydrochloride up to 1500 milligrams.

What the evidence does not support

The Cochrane reviewers concluded the randomised evidence cannot tell clinicians what dose, route or duration of thiamine to use for Wernicke-Korsakoff syndrome. (Source 25)

• Systematic review, Very low certainty.
• Size: 2 randomised trials.
• Who: People with alcohol misuse.
• How long: not applicable.
• Result: No pooled effect estimate was possible; the review reports insufficient evidence to guide dose, frequency, route or duration.
• Funding: not stated.

Evidence from randomised controlled clinical trials is insufficient to guide clinicians in determining the dose, frequency, route or duration of thiamine treatment for prophylaxis against or treatment of WKS due to alcohol abuse.

Intravenous thiamine did not reduce short-term mortality in sepsis and septic shock across nine randomised trials. (Source 4)

• Meta-analysis, Low certainty.
• Size: 520 patients across 9 randomised trials.
• Who: Adults with sepsis or septic shock.
• How long: short-term outcomes, mostly within hospital stay.
• Result: Short-time mortality OR = 0.78, p = 0.18; 24 h lactate change MD = 0.09, p = 0.36. Renal replacement therapy OR = 0.26, p = 0.0007; 24 h lactate MD = -0.35, p = 0.004; 24 h SOFA score change MD = 1.15, p < 0.00001.
• Funding: not stated.

No statistical significance was detected in short-time mortality (OR = 0.78, p = 0.18) and 24 h lactate change (MD = 0.09, p = 0.36) between the two groups.

Thiamine supplementation had no significant effect on any cardiac outcome in randomised trials in chronic heart failure. (Source 5)

• Meta-analysis, Low certainty.
• Size: 274 patients across 7 randomised double-blind controlled trials.
• Who: Adults with chronic heart failure.
• How long: not stated in the abstract.
• Result: Left ventricular ejection fraction WMD = 1.653%, 95% CI -1.098 to 4.405, p = 0.239; left ventricular end-diastolic volume WMD = -6.831 mL, p = 0.493; 6-minute walking test WMD = 16.526 m, p = 0.542; NT-proBNP WMD = 258.150 pg/mL, p = 0.306; NYHA class WMD = -0.223, p = 0.434.
• Funding: not stated.

Limit of this finding: In the quoted passage 'I2' is the heterogeneity statistic I-squared, which the journal prints with a superscript 2. It is not a measurement of 2 and not a second result. It says how much the pooled trials disagreed with each other: 'I2 = 0.0%' means they agreed closely and 'I2 = 87.1%' means they disagreed a great deal, and where it is that high a pooled average is a weak summary of the trials.

The results of the meta-analysis pooling these studies did not reveal any significant effect of thiamine treatment compared with placebo on left ventricular ejection fraction (WMD = 1.653%, 95% CI: -1.098 to 4.405, p = 0.239, I2 = 61.8%)

Benfotiamine 900 mg a day for 12 weeks improved thiamine status but did not reduce albumin excretion or a kidney injury marker in type 2 diabetes with nephropathy. (Source 7)

• Randomized trial, Moderate certainty.
• Size: benfotiamine (n = 39) or placebo (n = 43)
• Who: People with type 2 diabetes and urinary albumin excretion equivalent to 15-300 mg/24 h already on an ACE inhibitor or angiotensin receptor blocker.
• How long: 12 weeks.
• Result: Thiamine status improved versus placebo (P < 0.001). No significant decrease in 24-h urinary albumin excretion or 24-h KIM-1 excretion.
• Funding: not stated.

Compared with placebo, benfotiamine treatment resulted in significant improvement of thiamine status (P < 0.001). Benfotiamine treatment did not significantly decrease 24-h UAE or 24-h KIM-1 excretion.

The NIH Office of Dietary Supplements reports that the authors of a 2001 Cochrane review could draw no conclusions from the three small trials of oral thiamine in Alzheimer's type dementia. (Source 8)

• Official position, Very low certainty.
• Size: 3 double-blind randomised trials, each randomising fewer than 20 patients.
• Who: Patients with Alzheimer's type dementia.
• How long: not stated.
• Result: No pooled estimate possible. The trials compared 3 g/day oral thiamin to placebo; two crossover studies had no washout period.
• Funding: not applicable.

Limit of this finding: This is a government fact sheet describing a Cochrane review, not the review itself. We could not reach the 2001 Cochrane review, so this is recorded as an agency position - the design field says 'position' for that reason - and it should be read as the NIH summarising Cochrane, not as systematic-review evidence we checked.

The review authors stated that it was not possible to draw any conclusions from these three studies because they were small and the publications describing them did not provide enough detail to combine these data in a meta-analysis.

The same review's conclusion is that benfotiamine is biologically plausible but that its routine use outside documented thiamine deficiency cannot be recommended on current evidence. (Source 26)

• Expert review, not systematic, Low certainty.
• Size: not applicable; the review's concluding section.
• Who: People with diabetic peripheral neuropathy.
• How long: not applicable.
• Result: No effect estimate. The review calls for long-term randomised trials with structural and biomarker-based endpoints.
• Funding: not stated.

Limit of this finding: This is the review authors' recommendation, printed under its own 'Conclusions' heading. In the version of this passage we first recorded, that heading was deleted and the sentence ran on from the results narrative. It is an expert judgement from a narrative review, not a trial result.

Benfotiamine is biologically plausible but requires further validation in long-term randomized trials with structural and biomarker-based endpoints. Outside of documented thiamine deficiency, its routine use cannot be recommended based on current evidence.

Where the evidence is mixed

Randomised evidence for thiamine in Wernicke-Korsakoff syndrome is almost non-existent; the one analysable trial randomised 107 people and showed no dose-response pattern. (Source 3)

• Systematic review, Very low certainty.
• Size: 2 trials identified; 1 analysable trial with 107 participants.
• Who: People who abuse alcohol, with or at risk of Wernicke-Korsakoff syndrome.
• How long: 2 days of treatment in the analysable trial.
• Result: Lowest dose (5 mg/day) versus 200 mg/day on a delayed alternation test: mean difference -17.90, 95% CI -35.4 to -0.40, P = 0.04. No significant differences for the other dose comparisons.
• Funding: not stated.

A significant difference favoured 200 mg/day compared with the 5-mg/day dose in determining the number of trials needed to meet inclusion criteria on a delayed alternation test (mean difference (MD) -17.90, 95% confidence interval (CI) -35.4 to -0.40, P = 0.04).

Four of the nine sepsis trials were judged high risk of bias and the pooled sample was small enough that the authors warn about statistical power. (Source 4)

• Meta-analysis, Very low certainty.
• Size: 520 patients across 9 randomised trials.
• Who: Adults with sepsis or septic shock.
• How long: not stated.
• Result: Two articles low risk, three some concerns, four high risk of bias. In the thiamine-deficient subgroup of 2 studies and 51 patients, exploratory mortality OR = 0.18, p = 0.01.
• Funding: not stated.

Regarding bias risk assessment of studies, two articles were classified as low risk, three were categorized as raising some concerns, and four were identified as high risk.

In the BENDIP trial benfotiamine improved the neuropathy symptom score in the per-protocol analysis but the intention-to-treat result was not significant and the total symptom score did not differ. (Source 6)

• Randomized trial, Low certainty.
• Size: 165 patients randomised; 133/124 patients were analysed in the ITT/PP analysis: Benfotiamine 600 mg per day (n=47/43), benfotiamine 300 mg per day (n=45/42) or placebo (n=41/39)
• Who: Adults with symmetrical distal diabetic polyneuropathy.
• How long: 6 weeks.
• Result: Neuropathy Symptom Score differed between groups in the per-protocol population (p=0.033); in the intention-to-treat population the improvement was p=0.055. Total Symptom Score showed no significant differences at 6 weeks.
• Funding: not stated in the abstract; benfotiamine is a proprietary product and the trial was a phase III company-style study.

Limit of this finding: The trial's own primary endpoint missed significance in the analysis that counts everyone who was randomised: p = 0.055 in the intention-to-treat population, against the 0.05 threshold the trial itself used. The significant result (p = 0.033) comes from the per-protocol population, which drops people who did not complete treatment as planned and so tends to flatter a treatment. On a six-week trial of this size that is a negative primary result with a positive secondary analysis, not a demonstrated benefit.

In the ITT (intention to treat) population, the improvement of NSS was slightly above significance (p=0.055). The TSS (Total Symptom Score) showed no significant differences after 6 weeks of treatment.

A critical review concluded benfotiamine has a strong biochemical rationale but its clinical evidence in diabetic neuropathy is limited to short symptom-based studies with no large long-term trials. (Source 27)

• Expert review, not systematic, Low certainty.
• Size: not stated; structured narrative review of PubMed/MEDLINE.
• Who: People with diabetic peripheral neuropathy.
• How long: not applicable.
• Result: No pooled estimate. The review reports that benfotiamine's clinical evidence is confined to short-duration, symptom-based studies, with no large-scale, long-term trials published.
• Funding: not stated.

Benfotiamine has a strong biochemical rationale (transketolase activation, diversion of glycolytic intermediates from damaging pathways), but clinical evidence remains limited to short-duration, symptom-based studies, with no large-scale, long-term trials published.

The same authors' judgement - not a measurement - is that the risk of anaphylaxis with parenteral thiamine is low and that giving it is worth that risk in Wernicke-Korsakoff syndrome. (Source 28)

• Expert review, not systematic, Low certainty.
• Size: not applicable; the authors' concluding judgement.
• Who: Patients given parenteral thiamine for suspected Wernicke-Korsakoff syndrome in the UK.
• How long: not applicable.
• Result: No effect estimate. This is the authors' interpretation of the reaction counts reported in the Results section of the same abstract.
• Funding: not stated; the analysis used manufacturer sales and adverse incident data.

Limit of this finding: This is an opinion three authors formed from voluntary adverse-incident reports set against sales figures, printed under a separate 'Conclusion' heading. In the version of this passage we first recorded it ran straight on from the reaction counts with that heading deleted, which made a judgement look like a finding. It is reported here as their conclusion, which is what it is.

It is reasonable to assume that the risk of anaphylaxis is low, and lower than for many other drugs. The risk-benefit ratio for administration is favourable given the potential severity of brain damage in Wernicke-Korsakoff (WK) syndrome.

Where the research disagrees

What dose and route of thiamine to use for suspected Wernicke's encephalopathy

• Cochrane Dementia and Cognitive Improvement Group (2013), systematic review of 2 randomised trials: Evidence from randomised controlled clinical trials is insufficient to guide clinicians in determining the dose, frequency, route or duration of thiamine treatment for prophylaxis against or treatment of WKS due to alcohol abuse. (Source 25)
• European Federation of Neurological Societies guideline authors, as summarised by NIH ODS (2023), expert guideline (a position, not trial evidence): They recommend 200 mg thiamin, preferably intravenously, three times daily (total of 600 mg/day) until the signs and symptoms stop, along with a balanced diet. (Source 29)

How much

• Reference intake: Thiamine doses in medical use are set by the prescriber, not by a reference intake. For nutrition, the Food and Nutrition Board set Dietary Reference Intakes for thiamin, and the FDA Daily Value used on labels is 1.2 mg for adults and children aged 4 years and older. (Source 30)
• Upper limit: No tolerable upper intake level exists. The Food and Nutrition Board did not establish a UL for thiamin because reports of adverse effects from high thiamin intakes (50 mg/day or more) from food or supplements are lacking, while noting that excessive intakes could still have adverse effects. (Source 23)
• Studied: The only analysable randomised trial in Wernicke-Korsakoff syndrome gave five intramuscular doses - 5, 20, 50, 100 and 200 mg a day - for two days. (Source 31)
• Studied: The BENDIP trial gave benfotiamine 600 mg per day or 300 mg per day against placebo for 6 weeks. (Source 32)
• Studied: The diabetic nephropathy trial gave benfotiamine 900 mg/day for 12 weeks. (Source 33)
• Studied: The label's position on dosing: for beriberi, 10 to 20 mg intramuscularly three times daily for as long as two weeks, then an oral multivitamin containing 5 to 10 mg thiamine daily for one month; for Wernicke-Korsakoff syndrome, an initial 100 mg intravenously followed by 50 to 100 mg intramuscularly daily. (Source 17)

A common belief, and what the research shows

The belief: Thiamine is just a vitamin, so injections are harmless and the dose does not matter.

What the research shows: The injection label records that "Serious hypersensitivity/anaphylactic reactions can occur, especially after repeated administration. Deaths have resulted from IV or IM administration of thiamine (see ADVERSE REACTIONS)." An incidence analysis puts the risk low in absolute terms - "there have been 10 anaphylactic reactions to Pabrinex from between 5,431,235-6,651,947 patient-days (14,880-16,080 years) of treatment." - but it is not zero. The dose also matters in the opposite direction: the Cochrane review found the randomised evidence cannot say what dose to use, so the doses in guidelines are expert opinion.

Questions and answers

What is it?

Thiamine, the vitamin also known as B-one, is a water-soluble vitamin the body cannot make. Once absorbed it is converted to thiamine diphosphate, which is the working form. It is sold as a food nutrient, as a supplement, and as a prescription injection for treating deficiency. (Source 1)

What does it do in the body?

Thiamine diphosphate is the cofactor several enzymes need to release energy from carbohydrate and amino acids, including pyruvate dehydrogenase, alpha-ketoglutarate dehydrogenase and transketolase. Without it those enzymes stall, less ATP is made, lactate rises and cells can die. Thiamine is carried in red blood cells to the tissues with the highest energy demand, which is why the brain, heart, nerves, liver and pancreas are affected first. (Source 1)

Is it good or bad for you?

It depends entirely on context. In deficiency it is necessary and the response can be fast. In people who are not deficient, randomised trials have not shown benefit: meta-analyses in chronic heart failure and in sepsis found no effect on the outcomes measured. No upper intake level has been set because harm from high oral intakes has not been reported, but the Food and Nutrition Board still noted that excessive intake could have adverse effects, and injected thiamine carries a rare risk of anaphylaxis. (Source 23)

How do you get more of it?

Thiamine comes from whole grains, meat and fish, and in the United States and many other countries from breads, cereals and infant formulas that are fortified with it. Supplements usually contain thiamin mononitrate or thiamin hydrochloride; benfotiamine is a fat-soluble derivative converted to thiamine in the body. Absorption from a single oral dose is limited above a few mg, which is why hospitals use injections when deficiency affects the nervous system. (Source 34)

If it is harmful, what reduces it?

This does not really apply. Thiamine is not stored in large amounts and the body passes what it does not need into the urine, so there is no recognised need or method to reduce it. No upper intake level has been set. (Source 23)

Why might someone be low in it or missing it?

In lower-income settings deficiency usually comes from a monotonous diet based on polished white rice or cassava. In higher-income countries it is usually the company it keeps: alcohol dependence, diets limited to heavily processed food, kidney disease, eating disorders, bariatric surgery, dependence on parenteral nutrition, and critical illness. Loop diuretics and fluorouracil chemotherapy are documented as lowering thiamine levels. (Source 18)

Which whole foods contain it or feed it?

Whole grains, meat and fish naturally contain thiamine, and pork is a major source. Breads and cereals are the most common sources in the US diet, partly because they are fortified. Dairy products and most fruits contain little. Heating and discarding cooking water both reduce the amount that reaches you, and unenriched white rice has about a tenth as much as unenriched brown rice. (Source 34)

What happens if you do not have it?

Deficiency causes beriberi - mainly peripheral neuropathy and wasting, and in rare cases congestive heart failure that leads to oedema in the lower limbs and occasionally death. In developed countries beriberi is rare but still occurs. Giving supplemental thiamine, often parenterally, quickly cures beriberi. The other classic consequence is Wernicke-Korsakoff syndrome, covered under the other findings. (Source 35)

How can you test for it?

Two laboratory markers exist: thiamine diphosphate measured in whole blood, typically 70 to 180 nmol/L in healthy people, and the erythrocyte transketolase activity coefficient, where a ratio above 1.25 is taken to suggest deficiency. Neither is settled: there is no universal consensus on cut-offs, reference ranges come from generally healthy Western populations, and people with and without clinical signs can both show low thiamine status, which makes the results hard to interpret. (Source 36)

References

1. Annals of the New York Academy of Sciences. Thiamine deficiency disorders: a clinical perspective. 2021. PMID 33305487, DOI 10.1111/nyas.14536. Read the source
2. DailyMed / US FDA structured product label. THIAMINE HYDROCHLORIDE injection, solution [Sagent Pharmaceuticals] - FDA prescribing information. 2026. Read the source
3. Cochrane Database of Systematic Reviews. Thiamine for prevention and treatment of Wernicke-Korsakoff Syndrome in people who abuse alcohol — abstract, Main results. 2013. PMID 23818100, DOI 10.1002/14651858.CD004033.pub3. Read the source
4. Clinics (Sao Paulo). Efficacy of thiamine (vitamin B1) in sepsis and septic shock: A meta-analysis of randomized controlled trials — abstract, Results. 2026. PMID 41863966, DOI 10.1016/j.clinsp.2026.100901. Read the source
5. Clinical Cardiology. Role of Thiamine Supplementation in the Treatment of Chronic Heart Failure: An Updated Meta-Analysis of Randomized Controlled Trials — abstract, Results. 2024. PMID 38940395, DOI 10.1002/clc.24309. Read the source
6. Experimental and Clinical Endocrinology & Diabetes. Benfotiamine in diabetic polyneuropathy (BENDIP): results of a randomised, double blind, placebo-controlled clinical study — abstract, Results. 2008. PMID 18473286, DOI 10.1055/s-2008-1065351. Read the source
7. Diabetes Care. A double-blind, randomized, placebo-controlled clinical trial on benfotiamine treatment in patients with diabetic nephropathy — abstract, Results. 2010. PMID 20413516, DOI 10.2337/dc09-2241. Read the source
8. NIH Office of Dietary Supplements. Thiamin - Health Professional Fact Sheet. 2023. Read the source
9. NIH Office of Dietary Supplements. Thiamin - Health Professional Fact Sheet. 2023. Read the source
10. NIH Office of Dietary Supplements. Thiamin - Health Professional Fact Sheet. 2023. Read the source
11. Annals of the New York Academy of Sciences. Thiamine deficiency disorders: a clinical perspective. 2021. PMID 33305487, DOI 10.1111/nyas.14536. Read the source
12. NIH Office of Dietary Supplements. Thiamin - Health Professional Fact Sheet. 2023. Read the source
13. Annals of the New York Academy of Sciences. Thiamine deficiency disorders: a clinical perspective. 2021. PMID 33305487, DOI 10.1111/nyas.14536. Read the source
14. NIH Office of Dietary Supplements. Thiamin - Health Professional Fact Sheet. 2023. Read the source
15. NIH Office of Dietary Supplements. Thiamin - Health Professional Fact Sheet. 2023. Read the source
16. Annals of the New York Academy of Sciences. Thiamine deficiency disorders: a clinical perspective. 2021. PMID 33305487, DOI 10.1111/nyas.14536. Read the source
17. DailyMed / US FDA structured product label. THIAMINE HYDROCHLORIDE injection, solution [Sagent Pharmaceuticals] - FDA prescribing information. 2026. Read the source
18. Annals of the New York Academy of Sciences. Thiamine deficiency disorders: a clinical perspective. 2021. PMID 33305487, DOI 10.1111/nyas.14536. Read the source
19. DailyMed / US FDA structured product label. THIAMINE HYDROCHLORIDE injection, solution [Sagent Pharmaceuticals] - FDA prescribing information — label, ADVERSE REACTIONS section (LOINC 34084-4). 2026. Read the source
20. Alcohol and Alcoholism. Incidence of Adverse Reactions to Parenteral Thiamine in the Treatment of Wernicke's Encephalopathy, and Recommendations — abstract, Results. 2019. PMID 31565743, DOI 10.1093/alcalc/agy091. Read the source
21. Annals of the New York Academy of Sciences. Thiamine deficiency disorders: a clinical perspective. 2021. PMID 33305487, DOI 10.1111/nyas.14536. Read the source
22. NIH Office of Dietary Supplements. Thiamin - Health Professional Fact Sheet. 2023. Read the source
23. NIH Office of Dietary Supplements. Thiamin - Health Professional Fact Sheet. 2023. Read the source
24. Annals of the New York Academy of Sciences. Thiamine deficiency disorders: a clinical perspective. 2021. PMID 33305487, DOI 10.1111/nyas.14536. Read the source
25. Cochrane Database of Systematic Reviews. Thiamine for prevention and treatment of Wernicke-Korsakoff Syndrome in people who abuse alcohol — abstract, Authors' conclusions. 2013. PMID 23818100, DOI 10.1002/14651858.CD004033.pub3. Read the source
26. Nutrients. Alpha-Lipoic Acid and Benfotiamine in Diabetic Peripheral Neuropathy: A Critical Review of Mechanistic Rationale and Clinical Evidence Within a Nutritional Therapeutic Framework — abstract, Conclusions. 2026. PMID 42196997, DOI 10.3390/nu18101538. Read the source
27. Nutrients. Alpha-Lipoic Acid and Benfotiamine in Diabetic Peripheral Neuropathy: A Critical Review of Mechanistic Rationale and Clinical Evidence Within a Nutritional Therapeutic Framework — abstract, Results. 2026. PMID 42196997, DOI 10.3390/nu18101538. Read the source
28. Alcohol and Alcoholism. Incidence of Adverse Reactions to Parenteral Thiamine in the Treatment of Wernicke's Encephalopathy, and Recommendations — abstract, Conclusion. 2019. PMID 31565743, DOI 10.1093/alcalc/agy091. Read the source
29. NIH Office of Dietary Supplements. Thiamin - Health Professional Fact Sheet. 2023. Read the source
30. NIH Office of Dietary Supplements. Thiamin - Health Professional Fact Sheet. 2023. Read the source
31. Annals of the New York Academy of Sciences. Thiamine deficiency disorders: a clinical perspective. 2021. PMID 33305487, DOI 10.1111/nyas.14536. Read the source
32. Experimental and Clinical Endocrinology & Diabetes. Benfotiamine in diabetic polyneuropathy (BENDIP): results of a randomised, double blind, placebo-controlled clinical study — abstract, Methods. 2008. PMID 18473286, DOI 10.1055/s-2008-1065351. Read the source
33. Diabetes Care. A double-blind, randomized, placebo-controlled clinical trial on benfotiamine treatment in patients with diabetic nephropathy — abstract, Research design and methods. 2010. PMID 20413516, DOI 10.2337/dc09-2241. Read the source
34. NIH Office of Dietary Supplements. Thiamin - Health Professional Fact Sheet. 2023. Read the source
35. NIH Office of Dietary Supplements. Thiamin - Health Professional Fact Sheet. 2023. Read the source
36. Annals of the New York Academy of Sciences. Thiamine deficiency disorders: a clinical perspective. 2021. PMID 33305487, DOI 10.1111/nyas.14536. Read the source