Impact of Nutritional Therapy during Intensive Care Unit Admission on Post-Intensive Care Syndrome in Patients with COVID-19

COVID-19 causes symptoms of post-COVID-19 conditions, such as cough and respiratory distress, fatigue, headache, palpitations, chest pain, arthralgia, physical impairment, depression, and sleep disturbance, even after the acute phase. These symptoms have long-term physical, mental, and social health effects on patients [1] and are referred to as post-COVID-19 conditions. Post-COVID-19 conditions have been conservatively estimated to affect approximately 10% of infected patients and have become a social issue with many unresolved factors, such as pathophysiology, treatment methods, and risk factors [2]. World Health Organization defines post-COVID-19 condition as the continuation or development of new symptoms 3 months after the initial SARS-CoV-2 infection, with these symptoms lasting for at least 2 months with no other explanation.

Physical, cognitive, and mental disorders that occur after treatment in an intensive care unit (ICU) are referred to as post-intensive care syndrome (PICS) [3], and persistent inflammation and mechanical ventilation in critically ill patients lead to skeletal muscle mass loss and many functional impairments [4]. Respiratory complications and a prolonged ICU stay in COVID-19 patients cause malnutrition and the loss of skeletal muscle mass, which, in turn, reduce quality of life (QOL) and lead to physical dysfunction after ICU discharge [5], both of which increase the risk of developing PICS [5‒7]. PICS after COVID-19 and post-COVID-19 conditions significantly overlap and may exacerbate each other [6].

The importance of nutritional therapy has been emphasized for patients with severe COVID-19, and although there is currently no evidence in COVID-19 patients, the same nutritional therapy provided for severely ill patients is recommended [7, 8]. In critically ill COVID-19 patients, energy intake needs to be initiated at approximately 30% with a target of 15–20 kcal/kg/day and increased gradually to 80–100% by day 4, with protein intake reaching at least 1.2–1.3 g/kg/day by days 3–5 [7, 8]. In a previous study [9], COVID-19 patients in the ICU received nutritional therapy with a total energy intake of 1,000–2,000 kcal/day, either by tube or oral intake, and more than 50% of ICU patients received less than 1.2 g/kg/day of protein. Therefore, protein intake is insufficient for COVID-19 patients in the ICU.

An adequate nutritional assessment and treatment have been shown to reduce complications and improve clinical outcomes in patients, including those admitted to the ICU, those with chronic illness, and the elderly [10‒12]. However, few clinical studies have examined the relationship between an adequate nutritional supply in acute nutritional therapy and PICS. The EAT-ICU trial [13] showed that physical dysfunction did not improve after 6 months. However, according to Nakamura’s high protein versus medium protein diet [14], it may contribute to direct outcomes of nutritional therapy such as muscle mass [14]. Theoretically, nutritional shortages during the acute phase need to be avoided because they may lead to an energy debt, which may ultimately result in the development of PICS and delayed immune recovery; however, further studies are warranted to confirm this. In the present study, we hypothesized that energy and protein intakes during the acute phase in the ICU would be involved in the development of PICS and post-COVID-19 condition and we examined the effect of acute nutrition on the long-term outcome of PICS and post-COVID-19 condition in critically ill COVID-19 patients admitted to the ICU.

The present study was a secondary analysis of the COVID-19 Recovery Study II (CORES II). The CORES II study investigated post-illness symptoms, complications, and physical, mental, and social statuses after COVID-19 by following up patients hospitalized with COVID-19 after their discharge from the hospital. The present study is a sub-study within a multi-institutional study of 20 institutions and a post hoc analysis of data from 9 of 20 medical institutional. This study was approved by the Ethics Committee on (Approval No. NCGM-S-004471) and was supported by MHLW Research on Emerging and Re-emerging Infectious Diseases and Immunization (Program Grant Number JPMH21HA2011). The study data were collected and managed using Research Electronic Data Capture (REDCap) secure, web-based data capture application hosted at JCRAC data center of National Center for Global Health and Medicine [15]. Online supplementary Table 1 (for all online suppl. material, see https://doi.org/10.1159/000542298) shows basic data and the symptoms of the post-COVID-19 condition investigated in this study. Fatigue and breathlessness were rated on the scale shown in online supplementary Table 2. In the present study, post-COVID-19 condition was defined as the presence of symptoms at the 1-year time point. The EuroQol5-dimensions 5-level (EQ5D5L) [16] and Hospital Anxiety Depression Scale (HADS) [17] were also investigated.

A self-administered questionnaire survey was performed approximately 1 year after the diagnosis of COVID-19. Patients ≥20 years who were diagnosed with COVID-19 at a medical institution, admitted and discharged from 20 medical institutions including the National Center for Global Health and Medicine between April and September 2021 for the treatment of COVID-19 were included. The present study is an analysis of data from 9 of 20 medical institutional. After explaining participation in the study orally, in paper form, subjects had given their written or electromagnetic consent to participate in the study. A burden reduction fee was paid to subjects, which was paid directly to subjects gratuity worth 1,000 yen.

Data were collected separately for enteral and parenteral nutrition. Regarding parenteral nutrition, the calculation did not include energy from parenteral nutrition <5% glucose or propofol. Total energy (kcal/kg/week) and protein (g/kg/week) intakes per body weight were calculated during the first week after ICU admission. In subjects with early discharge within 1 week of ICU admission, from post-discharge to day 7 of discharge, energy and protein values at discharge were used to supplement. Linearly complemented values were used if there was a missing value in the period before discharge.

The primary outcome was a decrease in QOL assessed by the EQ5D5L 1 year after the diagnosis of COVID-19. The EQ5D5L was calculated using the conversion table in the development of the scoring method for the Japanese version of the EQ5D5L [18]. A decrease in QOL was defined as EQ5D5L <0.8 [19]. The secondary outcome was defined as physical dysfunction, mental dysfunction, the Malnutrition Universal Screening Tool (MUST), and post-COVID-19 conditions. Patients were defined as having physical dysfunction if they had symptoms of breathlessness evaluated by Modified Medical Research Council Dyspnea Scale or if their daily life was limited to light housework or office work (defined as a score ≥2 for breathlessness or fatigue in the items listed in online suppl. Table 2). Mental dysfunction was defined as a score ≥8 on HADS for anxiety or depression [20]. MUST was calculated as follows: body mass index (BMI) >20 was scored as 0 points, 18.5–20 as 1 point, and <18.5 as 2 points. Weight loss was calculated from a patient’s weight on admission and 1 year later, with a score of 0 points for <5%, 1 point for 5–10%, and 2 points for >10%. An insufficient nutritional intake was calculated as 2 points if the patient had anorexia. One point was considered an intermediate risk and ≥2 points as a high risk.

Total energy (kcal/kg/week) and protein (g/kg/week) intakes during the first week after ICU admission were used to divide patients into two groups based on median values and comparisons were performed between the groups. If the null hypothesis was not rejected by the Shapiro-Wilk test, results were expressed as the mean ± SD and compared using Welch’s t test. If the null hypothesis was rejected, the Mann-Whitney U test was used for comparisons. Nonparametric-paired values were expressed as medians (interquartile ranges) and compared using the Wilcoxon signed-rank sum test. Regarding categorical variables, the percentage of patients in each category was calculated and compared using the χ² test. A multivariable regression analysis adjusted for age, sex, BMI, SOFA, the ECMO status, NHF/NPPV status, mental illness, and ventilatory duration was performed to examine the relationships between energy and protein intakes during the first 7 days after ICU admission and physical dysfunction, impaired QOL, mental dysfunction, post-COVID-19 conditions, and MUST. Results with p value <0.05 are indicated by * and were significantly different.

Figure 1 shows the study outline. Tables 1 and 2 show the relationships between patient backgrounds and each endpoint when patients were divided into two groups based on median total energy and protein intakes. Median energy (kcal/kg/week) and protein (g/kg/week) intakes during the 7 days of ICU admission were 65.1 kcal/kg/week and 3.3 g/kg/week, respectively. On day 4 of ICU admission, energy intake exceeded 20 kcal/kg/day in 26 patients (11.8%), and on day 7 of ICU admission, energy intake exceeded 20 kcal/kg/day in 49 patients (22.3%). On day 5 of ICU admission, protein intake exceeded 1.2 g/kg/day in 16 patients (7.3%), and on day 7 of ICU admission, protein intake exceeded 1.2 g/kg/day in 20 patients (9.1%). Furthermore, the incidences of physical dysfunction, mental dysfunction, and decreased QOL were slightly higher in the group with lower energy and protein intakes. BMI was lower in the group with higher energy and protein intakes (energy: 26.3 vs. 29.0, p value <0.01, protein: 26.1 vs. 29.2, p value <0.01), while the ratio of mental illness was higher (protein: 8.6% vs. 1%, p value <0.01) and the duration of mechanical ventilation was longer (energy: 8 days vs. 5 days, p value <0.01, protein: 9 days vs. 4 days, p value <0.01). On the other hand, no significant differences were observed in severity scores, such as SOFA.

Online supplementary Table 3 shows the energy and protein intake for each day. Overall, nutrient intake tends to be low.

Online supplementary Figure 1 shows the relationship between the presence or absence of symptoms of post-COVID-19 conditions when patients were divided into two groups according to median energy (kcal/kg/week) and protein (g/kg/week) intakes during 7 days in the ICU. The group with higher energy and protein intakes had a lower incidence of post-COVID-19 conditions.

Figure 2 shows the outcomes of the multivariable analysis of total energy (kcal/kg/week) and protein (g/kg/week) intakes and each symptom of post-COVID-19 conditions. Total energy (kcal/kg/week) and protein (g/kg/week) intakes were associated with a decreased risk of the following symptoms: malaise (energy: unit odds ratio 0.984 [0.975–0.994], p value <0.01, protein: unit odds ratio 0.738 [0.607–0.898], p value <0.01), palpitations (energy: unit odds ratio 0.979 [0.966–0.994], p value <0.01, protein: unit odds ratio 0.697 [0 0.527–0.923], p value = 0.01), sleep disturbance (energy: unit odds ratio 0.989 [0.979–0.999], p value = 0.04, protein: unit odds ratio 0.804 [0.652–0.99], p = 0.04), arthralgia (energy: unit odds ratio 0.984 [0.974–0.997], p value = 0.02, protein: unit odds ratio 0.617 [0.458–0.833], p value <0.01), myalgia (energy odds ratio 0.983 [0.968–0.999], p value = 0.04, protein: unit odds ratio 0.678 [0.491–0.938], p value = 0.02), and muscle weakness (energy: unit odds ratio 0.985 [0.974–0.996], p value <0.01, protein: unit odds ratio 0.711 [0.567–0.892], p value <0.01). Protein intake was also associated with a decreased risk of ED (unit odds ratio 0.68 [0.464–0.995], p value = 0.04) and olfactory impairment (unit odds ratio 0.741 [0.555–0.989], p value = 0.04).

Table 3 shows the results of a logistic regression analysis of total energy (kcal/kg/week) and protein (g/kg/week) intakes with physical dysfunction, mental dysfunction, decreased QOL, and MUST as objective variables. A univariate regression analysis showed that neither physical nor mental dysfunction was associated with total energy (kcal/kg/week) or protein (g/kg/day) intake. Decreased QOL was associated with total energy (kcal/kg/week) intake (unit odds ratio 0.947 [0.898–0.999], p = 0.04). Moderate or higher MUST was also associated with total energy (kcal/kg/week) and protein (g/kg/week) intakes (energy: unit odds ratio 0.990 [0.981–0.998], p value = 0.03, protein: unit Odds ratio 0.821 [0.699–0.970], p value = 0.02). A multivariable regression analysis adjusted for age, sex, BMI, SOFA, ventilation duration, a history of psychiatric disease, and the presence of ECMO was performed. The primary outcome, decreased QOL defined by EQ5D5L <0.8, was associated with a lower risk of total energy intake (unit odds ratio 0.985 [0.975–0.994], p < 0.01) and total protein intake (unit odds ratio 0.720 [0.593–0.875], p < 0.01). Regarding the incidence of moderate or higher MUST, total energy intake (unit odds ratio 0.988 [0.978–0.999], p value = 0.03) and total protein intake (unit odds ratio 0.776 [0.623–0.967], p value = 0.02) were associated with a risk reduction. A relationship was also observed between total energy intake and physical dysfunction (unit odds ratio 0.990 [0.983–0.998], p = 0.02) but not mental dysfunction.

We investigated the potential effects of acute nutritional therapy for critically ill COVID-19 patients in the ICU and the long-term outcomes of various post-COVID-19 conditions and PICS 1 year after its diagnosis. The results obtained showed that insufficient total energy and protein intakes during the first week of ICU admission in COVID-19 patients were associated with decreased QOL, malaise, arthralgia, myalgia, palpitations, sleep disturbance, muscle weakness, and moderate or higher MUST 1 year after discharge. Total energy intake insufficient was also associated with physical dysfunction.

Early nutritional interventions in the ICU are known to decrease the rate of infectious complications [21]. In critically ill patients requiring admission to the ICU, recommended energy and protein intakes are 20–25 kcal/kg/day and 1.2–1.3 g/kg/day, respectively [22]. In the present study, on day 7 of ICU admission, energy intake exceeded the guideline recommendation of 20 kcal/kg/day in 49 patients (22.3%), and protein intake exceeded the guideline recommendation of 1.2 g/kg/day in only 20 patients (9.1%). In many cases, energy and protein intakes were lower than the recommended guidelines [22]. A survey of the implementation of nutritional therapy in relation to ICU in Japan [23] also found that the amount of energy and protein administered in ICU was low, and the amount of protein was very low. It is thought that the prognosis will be improved with higher nutrition doses under such a low nutrition provision [24, 25]. In particular, it is thought that more protein is needed under low-energy supply [26], so the supply of protein becomes more important. This may also be partly attributed to energy from propofol or infusions of less than 5% dextrose not being included in this study. This lower nutrition intake may have been associated with decreased QOL, symptoms of post-COVID-19 conditions, and malnutrition at 1 year. To the best of our knowledge, previous studies that examined the relationship between nutrition and long-term outcomes in severe COVID-19 patients reported no association with PICS, the symptoms of post-COVID-19 conditions, or malnutrition at 1 year. The present results emphasize the importance of acute nutritional therapy.

Energy intake has been associated with physical dysfunction 1 year after the diagnosis of COVID-19. Malnutrition is one of the factors contributing to the loss of skeletal muscle mass and function and may result in negative health outcomes, such as muscle weakness [27]. Malnutrition is also very common, particularly in critically ill COVID-19 patients who are admitted to the ICU, with 67% of critically ill patients being malnourished [28, 29]. Hypermetabolism may be a factor contributing to skeletal muscle loss in patients with COVID-19, with resting energy expenditure increasing in severe COVID-19 patients from the start of ventilation until the third week and being maintained until the seventh week [30]. Therefore, a nutritional shortage during the first week of severe COVID-19 may result in a greater energy debt and contribute to the development of PICS. An acute nutritional shortage leads to muscle weakness, reduced physical activity [31], decreased energy expenditure [32], and appetite loss [33], which may have triggered a negative spiral of further nutritional shortage, resulting in malnutrition 1 year later and adversely affecting physical function.

The group with higher energy and protein intakes also had a lower incidence of the symptoms of post-COVID-19 conditions. The symptoms of post-COVID-19 conditions in the present study that were associated with energy and protein intakes were malaise, arthralgia, myalgia, muscle weakness, palpitations, sleep disturbance, ED, and anosmia. The persistence of these symptoms may have been mediated by a decreased skeletal muscle mass, energy shortage, and trace element shortage [34, 35]. Previous studies [36, 37] demonstrated that nutritional shortage mainly affected the physical measures of QOL, albeit in other diseases. In the present study, nutritional shortage in the first week may have contributed to the symptoms of post-COVID-19 conditions and PICS and had a negative impact on QOL.

The strength of this study is that it is a multicenter study, with patients followed for 1 year to assess the impact on long-term QoL, PICS, and post-COVID-19 conditions. Few clinical studies have examined the relationship between nutritional supply and PICS in the acute phase of nutritional therapy. Energy and protein intakes during the first week after ICU admission are calculated and their impact on long-term health status is analyzed. Assessing specific nutritional intakes and identifying their impact provides practical nutrition management guidelines. The specific impact of nutritional intake on QoL and physical and mental functioning is shown, highlighting the importance of nutritional therapy. In particular, this study demonstrates that inadequate energy and protein intakes are associated with decreased QoL and increased various physical symptoms. We were able to demonstrate the relationship between each symptom of post-COVID-19 condition and acute nutritional therapy. There are several limitations that need to be addressed. Nutritional intakes were lower than the guidelines. Therefore, it is necessary to reevaluate the relationships observed in this study after increasing energy and protein intakes. Furthermore, physical dysfunction was assessed using indices, including breathlessness and fatigue, based on a questionnaire and not with common indices. Another limitation is that nutritional doses were only evaluated up to the seventh day in the ICU, which may have led to an underestimation of energy and protein levels in patients discharged early. EQ5D5L has not been evaluated prior to COVID-19 infection, so treatment for COVID-19 may have affected EQ5D5L.

Insufficient total energy and protein intakes during the first week after ICU admission in COVID-19 patients were associated with decreased QOL 1 year after the diagnosis of COVID-19. These acute nutrition shortages were also associated with an increased risk of developing PICS, post-COVID-19 conditions and malnutrition after 1 year, which may contribute to decreased QOL. To avoid the occurrence of PICS and other conditions, it might be prudent to try to ensure adequate nutritional therapy during the acute phase of the disease. However, the analysis of this study is an observational study, and future RCTs would be needed.

We thank the participants for their time and cooperation and the research office at the National Center for Global Health and Medicine, the CORESⅡ study group for the use of data and MHLW.

This study protocol was reviewed and approved by the Institutional Review Boards of the National Center for Global Health and Medicine, Approval No. NCGM-S-004471. This study protocol adhered to the Declaration of Helsinki’s ethical criteria. This study was approved by the Ethics Committees of all participating institutions. Written informed consent was obtained from all participants.

The authors have no conflicts of interest to declare.

This work was supported by MHLW, Research on Emerging and Re-emerging Infectious Diseases and Immunization (Program Grant No. JPMH21HA2011). The funder had no role in the design, data collection, data analysis, and reporting of this study.

Hiroyasu Iso, Mariko Hosozawa, Yoko Muto, Miyuki Hori, and Arisa Iba: conceptualization of the study. Kensuke Nakamura: methodology of the study. Shinya Suganuma: formal analysis of the study. Hiroyasu Iso, Mariko Hosozawa, Yoko Muto, Miyuki Hori, Arisa Iba, Shinya Suganuma, Kensuke Nakamura, Keiichiro Kawabata, and Muneaki Hemmi: investigation of the study. Hiroyasu Iso, Mariko Hosozawa, Yoko Muto, Miyuki Hori, Arisa Iba, Kensuke Nakamura, Hideaki Kato, Akira Kawauchi, Junji Hatakeyama, Shigeki Fujitani, Tomohiro Asahi, Taku Oshima, Kohei Ota, and Hiroshi Kamijo: resources of the study. Mariko Hosozawa, Yoko Muto, Miyuki Hori, and Arisa Iba: data curation of the study. Shinya Suganuma and Kensuke Nakamura: writing – original draft of the manuscript. Hiroyasu Iso, Mariko Hosozawa, Yoko Muto, Miyuki Hori, Arisa Iba, Shinya Suganuma, Kensuke Nakamura, Hideaki Kato, Keiichiro Kawabata, Muneaki Hemmi, Akira Kawauchi, Junji Hatakeyama, Shigeki Fujitani, Tomohiro Asahi, Taku Oshima, Kohei Ota, and Hiroshi Kamijo: writing – review and editing of the manuscript. Shinya Suganuma and Kensuke Nakamura: visualization of the manuscript. Hiroyasu Iso and Kensuke Nakamura: supervision of the study. Mariko Hosozawa and Yoko Muto: project administration of the study. Hiroyasu Iso: funding acquisition of the study. All authors have read and agreed to the published version of the manuscript.

The data that support the findings of this study are not publicly available due to their containing information that could compromise the privacy of research participants but are available from the CORES II Research office ([email protected]) upon reasonable request.