Gram Research analysis of a 2026 international study of 9,068 critically ill overweight or obese patients found that very low protein intake, below about 0.45 grams per kilogram of adjusted body weight per day, was linked to a higher risk of dying in intensive care. Protein intake did not appear to affect how soon patients left the ICU alive.
According to research reviewed by Gram, a large international study found that critically ill patients who are overweight or obese may face a higher risk of dying in the intensive care unit (ICU) when they receive very little protein in the early days of care. The study followed 9,068 patients from 785 ICUs around the world during their first four days of treatment. Patients whose protein intake was very low, below about 0.45 grams per kilogram of adjusted body weight per day, had a higher rate of dying in the ICU. Protein intake did not appear to affect how soon patients left the ICU alive. Adjusted body weight is a calculated weight that leaves out much of the extra body fat, and it is used to plan protein doses for heavier patients. Because the study is observational, it shows an association rather than proof that more protein saves lives. Care teams should still tailor nutrition to each patient.
Key Statistics
A 2026 cohort study of 9,068 critically ill patients with overweight or obesity found that protein intake below about 0.45 grams per kilogram of adjusted body weight per day was linked to a higher rate of dying in the intensive care unit.
In the same study, the strongest link appeared at zero protein intake, where the estimated risk of dying in the ICU was 38% higher (adjusted hazard ratio 1.41, 95% confidence interval 1.17 to 1.69).
The Quick Take
- What they studied: Researchers wanted to know whether the amount of protein heavier patients receive in the first few days of intensive care is linked to their chances of surviving the ICU stay and leaving it alive.
- Who participated: The analysis included 9,068 adults who had been in intensive care for at least 96 hours and had a body mass index of 25 or higher, meaning they were overweight or had obesity. They came from 785 ICUs across many countries.
- Key finding: Very low protein intake, below about 0.45 grams per kilogram of adjusted body weight per day, was linked to a higher rate of dying in the ICU. The strongest link appeared at zero protein, where the estimated risk of dying was 38% higher (hazard ratio 1.41, 95% confidence interval 1.17 to 1.69). Protein intake was not clearly linked to how quickly survivors left the ICU.
- What it means for you: If you or a loved one is in intensive care, the findings suggest that getting enough protein early on may matter, especially for heavier patients. Protein needs vary by person, so decisions should be made by the care team. This research does not give home dieting advice.
The Research Details
This study was a secondary analysis of a larger international survey. The survey was a snapshot that recorded how ICUs around the world feed their patients. Researchers pulled out the records of patients who had stayed in the ICU for at least four days and who were overweight or had obesity. They then looked at how much protein each patient received on days one through four, averaged over those days.
To make the comparisons fair, the team measured protein in grams per kilogram of adjusted body weight. Adjusted body weight is ideal body weight plus 25% of the gap between a person’s actual weight and their ideal weight. This approach is the one recommended by the European Society for Clinical Nutrition and Metabolism (ESPEN) for critically ill patients with extra weight.
The researchers then tracked what happened to each patient. Some died in the ICU, some left the ICU alive, and some were still in the unit at the end of follow-up. Because a patient who dies cannot later leave alive, the statistical method used accounts for these competing outcomes. This allowed the team to estimate how protein intake related to each outcome separately.
Many patients in intensive care are too sick to eat normally, so their protein is often given through feeding tubes or intravenous lines. Heavier patients pose a puzzle: using their actual weight could lead to very large protein doses, while using a very low weight could lead to too little. Using adjusted body weight is one way to set a sensible target.
Because the study followed patients over time and tracked both death and discharge, it could show how protein intake relates to outcomes rather than only describing what hospitals usually do. The competing risks method is designed for this kind of situation, where one outcome can prevent another from happening.
The large number of patients and the global reach also make the findings more useful for thinking about routine practice in many different settings.
The study has several strengths. It is large, with 9,068 patients from 785 ICUs in many countries, which makes the results more stable and broadly applicable. It used a statistical method built for situations where death can end the chance of leaving the ICU, and it measured protein against a clinical guideline standard.
There are important limits to keep in mind. It is observational, so it can show that low protein intake and higher death rates occur together, but it cannot prove that protein itself caused the difference. Sicker patients may have received less protein for reasons that also affect survival. The study also relied on records from a survey, so some information may be incomplete or imprecise. The sample size listed in the metadata for this article is 68, but the abstract reports 9,068 patients analyzed, and the larger figure is used here. Readers should check the full paper for the final count.
What the Results Show
The main result was a clear pattern around very low protein intake. When patients received less than about 0.45 grams of protein per kilogram of adjusted body weight per day during their first four days in the ICU, their rate of dying in the unit was higher. This threshold is a rough cutoff that the statistical model identified, not a precise magic number for every patient.
The strongest association appeared at the lowest end of the scale, at zero protein intake. At that level, the estimated risk of in-ICU death was 38% higher (adjusted hazard ratio 1.41, 95% confidence interval 1.17 to 1.69). The confidence interval means the true effect is likely somewhere in that range, and because it does not include 1, the link is unlikely to be due to chance alone. Because few patients likely received exactly zero protein, this figure is best read as the direction and size of the trend rather than a prediction for a typical patient.
Above the 0.45 grams threshold, the study found no clear evidence that more protein further changed the risk of dying in the ICU. In other words, once patients were getting a reasonable amount, extra protein did not show a clear additional benefit for survival in this data.
For patients leaving the ICU alive, the researchers found no evidence that early protein intake changed how soon they were discharged. This suggests that protein’s main link in this study was with survival rather than with speed of recovery.
The study looked at protein intake as an average over the first four days rather than on a single day. This approach captures a pattern of feeding rather than one meal or one bag of formula, which may be a more realistic view of what patients actually received.
Although the findings focus on heavier patients, the adjusted body weight method was used to standardize results across different body sizes. This means that a single protein number could be compared fairly between people with very different actual weights.
The results also suggest that the ICU discharge question and the death question should be looked at separately. Protein intake was tied to death risk but not to how quickly survivors left, which indicates that the two outcomes may be driven by different factors.
Guidelines from ESPEN already recommend using adjusted body weight to guide protein doses in critically ill patients with obesity, but until now it was unclear whether this approach had any link to patient outcomes. This study is one of the first large analyses to test that question directly.
Previous research on nutrition in critical illness has often found that not getting enough nutrition, including protein, during the early phase of intensive care is linked to worse outcomes. This new analysis adds to that picture by focusing on heavier patients, a group that is common in ICUs but has been less studied on this specific question.
At the same time, earlier trials of high versus lower feeding in critical illness have produced mixed results, so this study should be seen as one piece of a larger and still-developing body of evidence.
The biggest limitation is that the study is observational. Researchers compared what patients received with what happened to them, but they could not randomly assign patients to different protein levels. Doctors may have given more protein to patients who seemed stronger, or less to those who were too unstable to tolerate feeding, which could distort the results.
The data came from a survey of nutritional practice rather than a study designed specifically to test protein doses. This means some details, such as exactly how much protein was delivered versus ordered, may not have been captured perfectly.
The strongest result was at zero protein intake, which is an extreme point. Few patients may have received zero protein, so the estimate at that point is less precise than the overall pattern. The study also focused only on the first four days of ICU care, so it does not tell us what happens with protein intake later in the stay.
Finally, the study included only patients who had a body mass index of 25 or higher, so the findings may not apply to patients of normal weight or to those treated outside the ICU setting.
The Bottom Line
Research shows that very low protein intake in the early days of intensive care is linked to a higher chance of dying in the ICU for patients with overweight or obesity. Confidence in this association is moderate because the study is large and well-analyzed, but it is observational, so it cannot prove cause and effect. Care teams should continue following guidelines that use adjusted body weight to plan protein doses and should avoid allowing protein intake to fall very low when patients are unable to eat. Confidence that adding more protein beyond a reasonable amount improves survival is low based on this study alone.
Intensive care doctors, nurses, dietitians, and hospital nutrition teams should pay close attention, since they make decisions about feeding critically ill patients. Families of people in the ICU may also find it useful to ask whether protein needs are being met. People who are healthy and eating normally at home should not change their diet based on this study, because it applies only to critically ill patients in intensive care.
In the research, the key window was the first four days after ICU admission, so protein intake during this early period appeared to matter most in the data. Any effect on survival is measured during the ICU stay itself rather than over months or years. Because the study is observational, benefits cannot be predicted for any single patient, and the actual effect of changing feeding practices would need to be confirmed in controlled trials.
Frequently Asked Questions
How much protein do overweight patients need in intensive care?
Doses are often planned using adjusted body weight, a calculated weight that leaves out much of the extra body fat. A 2026 study of 9,068 patients linked protein intake below about 0.45 grams per kilogram of adjusted body weight per day to higher in-ICU death rates. Exact targets should come from the care team.
Does protein intake affect how quickly ICU patients leave the unit alive?
Protein intake did not appear to be associated with how soon survivors were discharged from intensive care, according to research reviewed by Gram. The clearest signal in the study concerned death, not discharge timing. Because the study is observational, it shows an association rather than proof of cause.
What is adjusted body weight and why do ICU teams use it?
Adjusted body weight equals ideal body weight plus 25% of the difference between actual and ideal weight. ESPEN recommends it for critically ill patients with overweight or obesity. It helps set protein targets without basing them on excess body fat, which can otherwise lead to unnecessarily large doses.
Should I change how much protein I eat at home based on this study?
This research applies to critically ill patients in intensive care, not healthy adults eating at home. Protein needs in everyday life depend on age, activity level, and health conditions. Anyone considering a diet change should speak with a doctor or registered dietitian first.
Want to Apply This Research?
- For users who log meals, track daily protein grams and compare them with a personalized target set by a doctor or dietitian. A useful measurable goal is to log protein intake every day and note whether it stays above a level the care provider recommends. If the app allows it, enter body weight and use a calculator for adjusted body weight to see the per-kilogram value.
- Users who are managing their own nutrition at home can build a simple habit of including a protein source at each meal, such as eggs, beans, fish, poultry, or dairy. This habit should be guided by a healthcare professional, especially for anyone with kidney disease or other conditions that change protein needs. For anyone in the hospital, the app can prompt users to ask the care team how their protein goals are being met.
- Over the long term, users can track weekly averages of protein intake, body weight, and energy levels, and bring these records to regular check-ups. Reviewing trends each month helps spot gaps before they become a problem. Any change in weight, appetite, or illness should prompt a conversation with a clinician about updating protein targets.
This article summarizes an observational research study and is for general information only. It is not medical advice and does not replace guidance from a doctor, critical care team, or registered dietitian. Protein needs in intensive care are individual and must be set by qualified clinicians. Anyone with questions about a hospitalized loved one’s nutrition should speak directly with the care team.
This research translation is published by Gram Research, the science division of Gram, an AI-powered nutrition tracking app.