Thiamine prophylaxis before refeeding — giving thiamine ahead of calorie reintroduction in any at-risk patient, before deficiency symptoms appear
Thiamine (vitamin B1) is a water-soluble vitamin with essentially no meaningful long-term storage depot in the body. Total body stores are on the order of 25–30 mg, held mostly in skeletal muscle, liver, heart, and brain — enough to cover normal turnover for roughly 2–3 weeks of zero intake. Any patient with prolonged inadequate oral intake, tube feeding interruption, or malabsorption enters a state of progressive thiamine insufficiency long before refeeding is ever considered, independent of alcohol use.
Unlike fat-soluble vitamins (A, D, E, K), which can be stored in adipose tissue and liver for months, thiamine is water-soluble and cleared renally within hours of absorption. The body maintains only a small circulating and tissue-bound pool, replenished by daily dietary intake of roughly 1.0–1.2 mg.
When intake falls — from anorexia, prolonged NPO status, restrictive eating, chronic vomiting, malabsorption, or simple food insecurity — there is no large reservoir to draw from. Functional depletion of thiamine-dependent enzyme activity can measurably begin within 1–3 weeks, well before any overt neurological or cardiac symptom appears.
Critically, this depletion process has nothing intrinsically to do with alcohol. Alcohol use disorder is a well-known and classic risk factor because it combines poor intake, impaired absorption, and impaired hepatic activation of thiamine — but the underlying vulnerability (a tiny buffer, rapid turnover, dependence on daily intake) applies to every human being, and to every cause of prolonged malnutrition.
Because thiamine has essentially no storage reserve, "how long has intake been inadequate" matters more than "why" intake became inadequate — the depletion clock runs the same way whether the cause is alcohol use, an eating disorder, cancer, or a prolonged hospital NPO order.
Mild-to-moderate thiamine depletion frequently produces vague or absent symptoms: fatigue, mild irritability, subtle cognitive slowing, or nothing detectable at all on routine exam. Laboratory confirmation (erythrocyte transketolase activity, whole blood thiamine) is rarely fast enough or available enough to guide real-time bedside decisions.
This silence is precisely what makes the depletion dangerous: a patient can arrive at the point of nutritional intervention — a hospital admission, a feeding plan, an infusion of dextrose-containing fluids — looking clinically unremarkable from a thiamine standpoint, while their functional reserve is already critically low. The deficiency only becomes clinically obvious once a large new metabolic demand (see Stage 2) is placed on top of it.
Thiamine pyrophosphate (TPP), the active cofactor form of thiamine, is required for at least three central enzymes of carbohydrate metabolism: pyruvate dehydrogenase, α-ketoglutarate dehydrogenase, and transketolase (pentose phosphate pathway). When a malnourished patient begins receiving carbohydrate calories — enteral formula, parenteral dextrose, or simply a resumed diet — flux through all three enzymes rises sharply, and TPP is consumed faster than an already-depleted reserve can supply it.
Pyruvate dehydrogenase (PDH) converts pyruvate to acetyl-CoA, the gateway step linking glycolysis to the citric acid cycle. Without adequate TPP, PDH activity falls, pyruvate backs up, and cells shift toward anaerobic lactate production — worsening tissue oxygenation and acid-base status precisely when a patient is trying to be nutritionally rescued.
α-ketoglutarate dehydrogenase performs an analogous TPP-dependent step further along the citric acid cycle, and transketolase links glycolytic intermediates to the pentose phosphate pathway, supporting NADPH production and nucleotide synthesis — both essential for the anabolic rebuilding that refeeding is meant to achieve.
Fat and protein calories place comparatively little demand on these specific TPP-dependent steps. It is the carbohydrate component of a feeding plan — intravenous dextrose, enteral formula carbohydrate content, or the carbohydrate fraction of a resumed oral diet — that acutely escalates thiamine consumption.
The core mismatch: carbohydrate refeeding increases thiamine demand precisely at the moment a malnourished patient has the least thiamine to spare — a supply-demand gap that opens within hours of the first calories, not days.
Neurons rely almost entirely on aerobic glucose oxidation for ATP and have essentially no capacity to switch to alternative fuels quickly. Brain regions with the highest metabolic turnover — thalamus, mammillary bodies, periaqueductal gray, cerebellar vermis — are also the regions classically damaged in acute thiamine deficiency, because they are the least able to tolerate any interruption in TPP-dependent oxidative metabolism.
This is why the consequence of an uncovered carbohydrate load is not generic fatigue but a specific, structurally localized injury pattern — the same regions repeatedly implicated regardless of whether the underlying cause of malnutrition was alcohol-related or not.
The core prevention strategy is sequencing: thiamine is administered before, or together with, the very first calories of a refeeding plan — not held in reserve as a treatment to be started once neurological or cardiac symptoms appear. This pre-emptive approach exists because acute thiamine-deficiency injury to the central nervous system can be difficult, and sometimes impossible, to fully reverse once it has become established.
Thiamine repletion, once frank neurological signs (confusion, ataxia, eye-movement abnormalities) have developed, still forms the backbone of emergency treatment for established deficiency — but a meaningful fraction of patients are left with permanent deficits even after treatment, particularly memory impairment consistent with Korsakoff-type injury. That asymmetry — cheap and safe to prevent, difficult and unreliable to fully reverse — is the entire rationale for prophylactic rather than reactive dosing.
Prophylactic administration does not require waiting for a diagnostic trigger, a specific lab value, or a symptom checklist. The trigger is the clinical situation itself: a patient with a plausible history of prolonged inadequate intake who is about to receive, or has just started, carbohydrate-containing nutrition.
The guiding clinical logic is sequencing, not diagnosis: give thiamine first, then feed — because thiamine is inexpensive and essentially free of meaningful risk, while missing the window carries a real chance of irreversible harm.
Thiamine is inexpensive, widely available, and carries a very low risk profile at prophylactic doses, whether given orally, intramuscularly, or intravenously. This favorable risk-benefit balance is precisely why prophylactic dosing is recommended broadly for any patient judged to be at meaningful risk, rather than being reserved only for those who already show signs of deficiency — the downside of unnecessary prophylaxis is minimal, while the downside of a missed opportunity is potentially permanent.
Alcohol use disorder is the classically taught risk factor for thiamine deficiency, and it remains an important one — but it is only one pathway to the same underlying vulnerability: prolonged inadequate intake with minimal body reserve. Eating disorders, cancer cachexia, prolonged NPO status, hyperemesis gravidarum, bariatric surgery complications, chronic gastrointestinal illness, and general food insecurity all produce comparable depletion and warrant the same prophylactic consideration before refeeding.
The population that benefits from prophylactic thiamine is defined by exposure — duration and severity of inadequate intake — not by any single diagnosis. Practical categories include:
• Eating disorders: anorexia nervosa, severe restrictive intake, or purging behaviors producing prolonged caloric and micronutrient deficits • Oncology patients: cancer cachexia, chemotherapy-related anorexia or mucositis limiting intake for weeks • Prolonged NPO status: extended fasting around major surgery, critical illness, or bowel rest for GI pathology • Hyperemesis gravidarum: severe, prolonged vomiting in pregnancy limiting intake and absorption • Bariatric surgery complications: malabsorption or prolonged poor intake post-procedure • Chronic GI illness: inflammatory bowel disease, short bowel syndrome, chronic pancreatitis • General malnutrition: food insecurity, elderly patients with chronic poor intake, prolonged unsupplemented IV fluids
In every one of these groups, the pathophysiology in Stages 1–2 is identical: reserves are depleted, and reintroducing calories — especially carbohydrate — creates the same acute supply-demand mismatch.
A screening approach limited to "does this patient drink heavily" will systematically miss a large share of patients who are nonetheless at genuine risk. Broader risk-based screening instead asks about duration of reduced or absent intake, recent weight loss, chronicity of illness, and any history of malabsorption — regardless of whether alcohol is part of the picture.
This reframing does not diminish the importance of thiamine repletion in alcohol-related presentations; it simply recognizes that the preventive principle generalizes, and that limiting prophylaxis to one subgroup leaves other equally vulnerable patients unprotected.
Prophylactic thiamine is not a single pre-feeding dose that ends once the first calories are tolerated. Because carbohydrate-driven metabolic demand remains elevated for as long as nutritional rehabilitation continues, supplementation is typically continued for several days into the refeeding period — covering the full window during which thiamine-dependent enzyme flux stays above baseline.
A single dose given immediately before the first feed addresses the acute onset of demand, but carbohydrate metabolism remains elevated throughout the early refeeding period as caloric intake is advanced toward a target goal — not just at the first bite or first infusion. If supplementation stops too early, a patient can still develop a delayed deficiency state days into what looked like a successful nutritional recovery.
Continued dosing over several days is therefore treated as part of the same preventive strategy as the initial dose, not a separate treatment decision. The exact duration is individualized to the severity of baseline risk and the pace of caloric advancement, but the underlying principle is coverage across the whole period of elevated demand, not just its opening moment.
Prophylaxis is a course, not a single event: the pre-feeding dose opens the window of protection, and continued dosing over the following days keeps it open until carbohydrate metabolism — and thiamine demand — has stabilized.
As oral, enteral, or parenteral intake advances toward an adequate, sustained caloric target and the patient demonstrates tolerance without biochemical or clinical evidence of refeeding complications, the acute prophylactic phase gives way to standard maintenance micronutrient practice. The decision to stop concentrated prophylactic dosing rests on clinical judgment — accounting for how severe the original depletion risk was, how quickly calories were advanced, and whether any early warning signs of refeeding syndrome emerged along the way — rather than a fixed calendar cutoff applied uniformly to every patient.