Nutrition in Cirrhosis: Protein, Sarcopenia and the Meal Most Patients Skip
Key takeaways
- Sarcopenia — loss of skeletal muscle — affects roughly 37.5% of people with cirrhosis worldwide and roughly doubles the risk of death, independent of Child-Pugh and MELD.
- Protein restriction in hepatic encephalopathy is obsolete. A randomised trial showed a normal-protein diet is safe during an episode of encephalopathy, and restricting protein only increased protein breakdown.
- EASL and ESPEN both set the target at about 1.2–1.5 g of protein per kg of body weight per day, alongside roughly 35 kcal/kg/day — far more than most Indian cirrhosis patients actually eat.
- In our own randomised trial in alcohol-related cirrhosis, pushing intake higher still did not improve 12-month survival — but crossing above 25 kcal/kg and 0.8 g protein/kg per day did, whichever arm the patient was in.
- A cirrhotic liver enters starvation metabolism after roughly 12 hours without food. A late-evening snack added about 2 kg of lean tissue over 12 months in a randomised trial; the same calories given in the daytime added none.
- Skeletal muscle cut-offs for sarcopenia derived from Western populations do not transfer to Indian bodies. Indian cut-offs, derived from 275 healthy Indian organ donors, are 36.54 cm²/m² in men and 30.21 cm²/m² in women.
Adults with cirrhosis need roughly 35 kcal/kg and 1.2–1.5 g of protein per kg of body weight each day, spread across four to six meals, with a carbohydrate-and-protein snack at bedtime so the overnight fast never exceeds about 12 hours. Protein should not be restricted, even during hepatic encephalopathy. Sarcopenia is present in about 37.5% of patients with cirrhosis and roughly doubles mortality risk.
— Dr Chetan Kalal, DM Hepatology, Gleneagles Hospital Mumbai
A man of 46 with alcohol-related cirrhosis comes to the clinic weighing 58 kg. His albumin is 2.6, he has moderate ascites, and somebody along the way told him that protein is bad for the liver. So he eats rice and a thin dal at lunch, skips dinner because the ascites makes him feel full, and has nothing between 8 p.m. and 9 the next morning. On paper he is being treated for decompensated cirrhosis. In practice he is being starved, and the starvation is doing more damage month to month than anything else on his prescription.
This is the most common and the most correctable problem I see in cirrhosis care in India. Nutrition is not adjunctive here. It is a treatment with a measurable effect on ascites, on encephalopathy, on transplant candidacy and on survival — and it is the one part of the plan that consistently gets left to a photocopied diet sheet. My own research has been in this area for over a decade, first at ILBS and now at Gleneagles, and the findings have changed how I counsel patients in ways that are worth setting out plainly.
Why a cirrhotic liver starves you overnight
A healthy person who skips dinner draws on liver glycogen through the night and wakes up fine. A cirrhotic liver holds very little glycogen. Once that small store runs out — which happens after roughly 12 hours of fasting, sometimes less — the body switches to gluconeogenesis and starts breaking down skeletal muscle for amino acids. An overnight fast in cirrhosis is therefore metabolically closer to a two- or three-day fast in a healthy adult.
Run that every night for a year and the arithmetic is brutal. Muscle is not just strength; in cirrhosis it is also the main site of ammonia detoxification outside the liver. Lose muscle and you lose ammonia-handling capacity, which is one reason sarcopenic patients decompensate into hepatic encephalopathy more readily and recover from it more slowly.
Layered on top are the mechanical problems. Ascites presses on the stomach and produces early satiety. Salt restriction makes food taste of nothing, so patients eat less of it. Zinc deficiency blunts taste further. Recurrent hospital admissions mean repeated periods of nil-by-mouth for procedures. Each is minor on its own. Together they produce a patient who is losing half a kilo of lean tissue a month while everyone focuses on the diuretic dose.
Sarcopenia: what the muscle loss actually costs
The data here are unusually clear for a nutritional question. A 2022 meta-analysis in the Journal of Hepatology pooled 22 cohort studies and 6,965 patients with cirrhosis. Sarcopenia was present in 37.5% overall, and was commoner in men, in alcohol-associated liver disease, and in Child-Pugh C disease. After adjustment, sarcopenia carried an increased risk of death with a hazard ratio of 2.30 (95% CI 2.01–2.63). The association held in every subgroup tested, and held after excluding patients with hepatocellular carcinoma.
A hazard ratio above two, robust across sex, aetiology and disease severity, puts muscle mass in the same prognostic league as the variables we build our scoring systems out of. MELD does not measure it. Child-Pugh does not measure it. Which means two patients with identical MELD scores can have materially different one-year outlooks, and the difference sits in the psoas muscle on a CT slice that was taken for another reason entirely.
Why Western sarcopenia cut-offs mislead in Indian patients
Here is where most international guidance stops being directly usable, and where our own work matters. Sarcopenia on CT is defined by the skeletal muscle index at the third lumbar vertebra — the cross-sectional muscle area normalised for height. But the threshold below which you call it sarcopenia has to come from a healthy reference population, and the reference populations used in most published cut-offs were Western.
Our group measured single-slice L3 CT images in 275 healthy Indian adults — living organ donors, mean age 32, mean BMI 24.2 — and derived cut-offs of 36.54 cm²/m² for men and 30.21 cm²/m² for women. Applied to 148 Indian patients with alcohol-related cirrhosis, the prevalence of sarcopenia was 12.8%. That is a very different number from the 37.5% global pooled figure, and the two are not a like-for-like comparison: different cohorts, different case mix, different thresholds. But the direction of the problem is unambiguous. Indian adults carry less skeletal muscle at baseline than the Western populations from which the standard cut-offs were derived, and applying a foreign threshold to an Indian patient will systematically misclassify them.
That study also produced a finding that catches people out. Compensated patients had more body fat than healthy controls and comparable muscle. Decompensation is where both muscle and fat fall away together. So a comfortably built patient with Child A cirrhosis and a normal BMI is not reassuring evidence of good nutrition — it may simply be adiposity masking a muscle compartment that has not yet started to go. Sarcopenic obesity is real, common in our metabolic fatty liver population, and invisible on a weighing scale.
How much to eat — the floor matters more than the ceiling
EASL’s 2019 clinical practice guidelines on nutrition in chronic liver disease and the ESPEN practical guideline both land in the same place: around 35 kcal/kg/day of energy and 1.2–1.5 g/kg/day of protein for patients with cirrhosis, calculated on dry or ideal body weight where ascites makes the scale unreliable. For a 60 kg man that is roughly 2,100 kcal and 72–90 g of protein a day — three eggs, two bowls of dal, a serving of paneer or fish, curd, and it is still a stretch.
The obvious next question is whether going higher does more good. We tested it. In a randomised controlled trial of malnourished patients with alcohol-related cirrhosis (NCT02140294), we assigned 104 patients either to 35–40 kcal and 1.2 g protein/kg/day by diet alone, or to 40–45 kcal and 1.5 g/kg/day by diet plus a polymeric oral supplement, for three months, with follow-up at 3 and 12 months.
The aggressive arm did better on the things nutrition should improve. Nutritional status by Royal Free Hospital Subjective Global Assessment improved in 33.3% versus 14.2%. Dry body weight and mid-upper arm circumference both rose. Ascites resolved in 53.3% of the intervention group against 44.4% of controls. But median 12-month survival was no different between the arms.
The result that changed my practice was the one that cut across both groups. Irrespective of which arm a patient was randomised to, an energy intake above 25 kcal/kg/day and a protein intake above 0.8 g/kg/day significantly improved 12-month survival. The threshold that mattered was the floor, not the ceiling.
That reframes the clinic conversation entirely. Chasing 1.5 g/kg in a patient who is nauseated and anorexic tends to end in a supplement tin gathering dust in a cupboard and a family that feels it has failed. Getting a patient reliably above 25 kcal and 0.8 g/kg every single day, including the bad days, is achievable, and on our data it is where the survival signal lives. Aim for the guideline target — but never let the perfect target become the reason nothing gets eaten at all.
The protein myth in hepatic encephalopathy
The belief that protein triggers encephalopathy and should be restricted dates to uncontrolled observations from the 1950s. It survives in India today in ward instructions, in discharge summaries, and in what families tell each other. It is wrong, and it has been shown to be wrong in a controlled trial.
Córdoba and colleagues randomised 30 cirrhotic patients admitted with an episode of encephalopathy to either a low-protein diet with progressive increments or a normal-protein diet for 14 days, on top of standard treatment. The course of the encephalopathy was no different between the groups. Protein synthesis was the same in both. What differed was protein breakdown — higher in the low-protein group. In other words, restricting protein did nothing for the encephalopathy and actively accelerated the muscle loss that makes future encephalopathy more likely.
Both EASL and ESPEN now advise against protein restriction in hepatic encephalopathy. If a patient genuinely cannot tolerate dietary protein — a rare situation, and one that needs a specialist’s judgement rather than a blanket rule — vegetable and dairy protein sources are better tolerated than red meat, and branched-chain amino acid supplementation is discussed in the guidelines as an option in selected patients. None of that is a reason to start a malnourished cirrhotic on a low-protein diet. Treat the precipitant of the encephalopathy: infection, constipation, a bleed, dehydration from over-diuresis. Do not treat the dal.
Meal timing: the snack that adds two kilograms of muscle
If the overnight fast is the problem, shortening it is the fix, and this has been tested directly. Plank and colleagues randomised 103 patients with cirrhosis to receive 710 kcal of supplementary nutrition either during the day, between 9 a.m. and 7 p.m., or at night, between 9 p.m. and 7 a.m., and measured total body protein by neutron activation analysis over 12 months.
Total daily energy and protein intake rose equally in both groups. Total body protein rose only in the night-time group — by 0.38 kg at three months and 0.53 kg at twelve, equivalent to roughly 2 kg of lean tissue. The daytime group showed no significant change. Same calories, same protein, different clock, entirely different outcome.
The practical translation is simple enough to write on a prescription. Four to six small meals a day rather than two or three large ones, because early satiety from ascites defeats large meals. Never more than about 12 hours between the last food of the night and the first of the morning. And a carbohydrate-plus-protein snack at bedtime, every night, treated with the same seriousness as a tablet.
What this looks like on an Indian plate
Guideline numbers are useless to a patient who eats roti and sabzi. Translating them is most of the work.
Protein, for the majority of Indian patients who are vegetarian, comes from dal and other pulses, rajma and chana, paneer, curd and milk, soya, and eggs where acceptable. Dairy and vegetable protein are particularly well suited here, since they are the sources best tolerated in patients with a history of encephalopathy. Non-vegetarians should be eating fish and chicken, not avoiding them out of some idea that meat burdens the liver.
The bedtime snack should carry both carbohydrate and protein: a glass of milk with a banana, curd with poha, a paneer sandwich, sattu in milk, or a prescribed oral nutritional supplement if food is not going down. It does not need to be elaborate. It needs to be every night.
Salt restriction and nutrition are frequently set against each other, and they should not be. Salt restriction for ascites limits sodium, not calories or protein. Food cooked with cumin, black pepper, lemon, garlic, asafoetida, tamarind and fresh coriander stays interesting without added salt. Where appetite is failing, a low-salt diet that the patient will not eat is worse than a slightly more liberal one that they will — and correcting zinc deficiency, which blunts taste, is worth attention in its own right.
Two things to stop. Complete abstinence from alcohol is non-negotiable in alcohol-related liver disease and remains the single most powerful intervention available. And crash dieting or intermittent fasting, whatever the metabolic benefits in the general population, is actively dangerous in established cirrhosis — the extended fasting window is precisely the mechanism destroying muscle. If weight loss is genuinely indicated for a patient with obesity and cirrhosis, it has to be supervised, gradual, and protein-replete.
| Element | Target | Evidence / source |
|---|---|---|
| Energy | ~35 kcal/kg/day (dry or ideal weight) | EASL 2019; ESPEN 2020 |
| Protein | 1.2–1.5 g/kg/day | EASL 2019; ESPEN 2020 |
| Minimum that changed survival | >25 kcal/kg and >0.8 g protein/kg/day | Kalal et al., RCT, alcohol-related cirrhosis, 2022 |
| Protein in hepatic encephalopathy | Do not restrict | Córdoba RCT 2004; EASL 2019; ESPEN 2020 |
| Maximum fasting interval | ~12 hours; late-evening snack nightly | Plank RCT 2008 (+0.53 kg total body protein at 12 months) |
| Meal pattern | 4–6 small meals daily | EASL 2019; ESPEN 2020 |
| Sarcopenia cut-off (L3-SMI, Indian) | <36.54 cm²/m² men; <30.21 cm²/m² women | Benjamin, Shasthry, Kalal et al., Liver Int 2017 |
| Alcohol | Complete abstinence | EASL 2019; ESPEN 2020 |
Nothing on that table is a prescription. Energy and protein targets are calculated per patient, adjusted for ascites, renal function and diabetes, and they change as the disease changes. Work them out with your hepatologist and a clinical dietitian, not from a table on the internet.
When nutrition becomes a transplant question
Sarcopenia is not only a prognostic marker; it is a modifiable one, and the window in which to modify it is before liver transplantation, not after. A patient who arrives at surgery with depleted muscle has a harder post-operative course, and the association between pre-transplant muscle mass and post-transplant outcome is consistent enough across the literature that muscle assessment now forms part of a serious pre-transplant work-up.
The practical implication for anyone being evaluated: nutritional optimisation in the months before listing is real work with real returns, not a formality to be ticked off. If a CT abdomen has already been done for another indication, the L3 muscle measurement can often be extracted from it retrospectively at no extra cost or radiation — which is the cheapest prognostic test available in chronic liver disease and one of the least used.
Frequently asked questions
Is protein really safe if I have had hepatic encephalopathy?
Yes. A randomised trial of patients admitted with an episode of encephalopathy found a normal-protein diet was as safe as a restricted one, with no difference in the course of the encephalopathy — and the restricted group broke down more of their own body protein. EASL and ESPEN both advise against protein restriction. Encephalopathy is treated by finding and correcting its precipitant, not by removing dal from the plate.
How much protein should I be eating with cirrhosis?
The guideline target is 1.2–1.5 g per kg of body weight per day, with roughly 35 kcal/kg/day of energy, calculated on dry weight if you have ascites. Our own trial data suggest the more important line is the minimum: staying above 0.8 g/kg/day of protein and 25 kcal/kg/day was associated with better 12-month survival. Your exact numbers should be set by your hepatologist and dietitian, since diabetes, kidney function and fluid retention all change the calculation.
Why do I need to eat before bed?
A cirrhotic liver holds very little glycogen, so it runs out of stored fuel after roughly 12 hours without food and starts breaking down muscle instead. A randomised trial found that giving supplementary nutrition at night — rather than the same amount during the day — increased total body protein by about half a kilogram over 12 months, equivalent to roughly 2 kg of lean tissue. The daytime group gained nothing. A carbohydrate-and-protein snack at bedtime is one of the few free wins in cirrhosis care.
Can I lose weight if I have fatty liver and cirrhosis?
Sometimes, but never by fasting or crash dieting. Extended fasting windows are exactly the mechanism that destroys muscle in cirrhosis, and sarcopenic obesity — low muscle hidden under normal or high body fat — is common and easily missed on a weighing scale. Any weight reduction in established cirrhosis must be gradual, protein-replete and medically supervised.
Will a protein powder or supplement help?
Oral nutritional supplements have a genuine role when food alone cannot reach the target, and in our trial the supplemented arm improved nutritional status and resolved ascites more often. They are a bridge, not a substitute for meals, and the choice of product matters in cirrhosis — several commercial supplements carry sodium loads that are unsuitable with ascites. Ask before buying one off a shelf.
My weight has not changed — does that mean my nutrition is fine?
No, and this is the commonest false reassurance in the clinic. Fluid retention from ascites and oedema can hold weight steady or push it up while lean tissue is being lost underneath. Mid-upper arm circumference, handgrip strength and, where a CT is available, the L3 skeletal muscle index give a far truer picture than the scale.
This article is general education, not medical advice. Nutrition targets in cirrhosis are individual and change with the stage of the disease, kidney function, diabetes and fluid status. Do not start, stop or alter any diet, supplement or medication on the basis of what you have read here — discuss it with your hepatologist or a qualified clinical dietitian first.
Book a consultation with Dr Chetan Kalal at Gleneagles Hospital, Mumbai — call +91 83294 76669 or use the appointment form on this site. International and NRI patients may request a teleconsultation.
References
- European Association for the Study of the Liver. EASL Clinical Practice Guidelines on nutrition in chronic liver disease. J Hepatol. 2019;70(1):172–193. PMID 30144956. doi:10.1016/j.jhep.2018.06.024
- Bischoff SC, Bernal W, Dasarathy S, Merli M, Plank LD, Schütz T, Plauth M. ESPEN practical guideline: Clinical nutrition in liver disease. Clin Nutr. 2020;39(12):3533–3562. PMID 33213977. doi:10.1016/j.clnu.2020.09.001
- Kalal C, Benjamin J, Shasthry V, Kumar G, Sharma MK, Joshi YK, Sarin SK. Effect of long-term aggressive nutrition therapy on survival in patients with alcohol-related cirrhosis: a randomized controlled trial. Indian J Gastroenterol. 2022;41(1):52–62. PMID 35235198. doi:10.1007/s12664-021-01187-3
- Benjamin J, Shasthry V, Kalal CR, Anand L, Bhardwaj A, Pandit V, et al. Characterization of body composition and definition of sarcopenia in patients with alcoholic cirrhosis: a computed tomography based study. Liver Int. 2017;37(11):1668–1674. PMID 29065258. doi:10.1111/liv.13509 (indexed in PubMed under the surname “Kaal” owing to a journal typographical error)
- Tantai X, Liu Y, Yeo YH, Praktiknjo M, Mauro E, Hamaguchi Y, et al. Effect of sarcopenia on survival in patients with cirrhosis: a meta-analysis. J Hepatol. 2022;76(3):588–599. PMID 34785325. doi:10.1016/j.jhep.2021.11.006
- Plank LD, Gane EJ, Peng S, Muthu C, Mathur S, Gillanders L, et al. Nocturnal nutritional supplementation improves total body protein status of patients with liver cirrhosis: a randomized 12-month trial. Hepatology. 2008;48(2):557–566. PMID 18627001. doi:10.1002/hep.22367
- Córdoba J, López-Hellín J, Planas M, Sabín P, Sanpedro F, Castro F, et al. Normal protein diet for episodic hepatic encephalopathy: results of a randomized study. J Hepatol. 2004;41(1):38–43. PMID 15246205. doi:10.1016/j.jhep.2004.03.023
