The COVID-19 pandemic has brought about immeasurable adverse health effects across the world. Up to 30% of the SARS-CoV-2 infected patients present with high nutritional risk; if such patients have protein deficiencies or electrolyte imbalance, their ability to survive the infection is further reduced [6]. There are several reasons for COVID-19 to cause disorders of nutrition, digestion, and absorption. First, the SARS-CoV-2 virus can invade the human body by binding to the human angiotensin-converting enzyme 2 (ACE-2) receptor in hepatocytes derived from bile duct epithelial cells, thus causing liver tissue injury by the up-regulation of ACE-2 expression [5]. Second, ACE-2 receptors also express in ileum and colon enterocytes. The compensatory proliferation of enterocytes caused by the binding of SARS-CoV-2 with ACE-2 can also disrupt the normal intestinal flora [4]. Finally, inflammatory reactions caused by SARS-CoV-2 could induce dysbiosis, causing GI disturbance, especially diarrhea [3]. Data from several sources have identified the association between nutritional risk and mortality in severely or critically ill COVID-19 patients. Xiaobo Zhao et al. reported that critical COVID-19 patients had a significantly higher proportion of NRS scores, leading to a higher risk of mortality and longer hospital stay [17]. Furthermore, a report conducted by Tao Li confirmed the prevalence of malnutrition in elderly patients with COVID-19 in Wuhan, China [18]. The nutritional status of our study participants was evaluated by a clinical nutritionist, and the average NRS-2002 score was three, which meant a high risk of malnutrition in the need of nutritional supports [19].
The obvious efficacy of providing additional nutritional support to reduce the severity of COVID-19 has been described by several recent reviews [20]. For instance, experts from ASPEN extrapolated the results of previous publications and recommended starting nutrition within 24 – 36 h after the admission for COVID-19 critically ill patients in the ICU [11]. The experts from the European Society for Parenteral and Enteral Nutrition (ESPEN) suggest that nutritional support could be started within 1–2 day for COVID-19 patients undergoing mechanical ventilation therapy [11]. Besides, non-critically ill patients should also receive nutritional supplements as early as possible, particularly, increasing protein and calorie intake by oral nutritional supplements [21]. Besides the necessary carbohydrates, lipids, amino acid supplements, vitamins, and minerals were also reported to reduce the risk of COVID-19 infections and deaths [22]. For example, vitamin C was found to reduce the susceptibility of lower respiratory tract infections, and vitamin D could induce cathelicidins and defensins to inhibit the viral replication rate [23].
Enteral nutrition (EN) and Parenteral Nutrition (PN) are two routine ways for nutritional support to the patients. Most experts have preferred EN over PN [10]. Theoretical benefits of EN include the preservation of the integral mucosal architecture, gut-associated lymphoid tissue, and hepatic immune function. In addition, EN could also reduce inflammatory response during the nutrition input and prevent antigenic leak from the gut [24]. A practical nutritional guideline for COVID-19 patients, published by Ronan Thibault et al. states that EN delivered within 48 h following ICU admission should be the first line plan, while PN should be prescribed if EN is contraindicated or insufficient [8]. Hence, EN support was applied for the most severe COVID-19 cases in our hospital. Participants in EEN group, received EN therapy within two days, and gradually changed to the oral nutrition. While for patients with digestive symptoms or high risk of aspiration, PN support was firstly provided within 24–48 h, and gradually shifted to the EN support. So, PN treatment was applied for all participants in LEN group within two days, to guarantee their necessary nutritional supply. Although previous observations have shown that nutritional supplements administered at an early stage are important for enhancing host resistance to virus infection, there is still no consensus on the reasonably appropriate initiation time of EN [25]. Besides, whether there is an association between the initiation of energy intake and the incidence of RS is still elusive.
RS is described as a range of metabolic and electrolyte alterations that occur because of the reintroduction or increased provision of calories after a period of decreased or no caloric intake [12]. The participants in our study were severe cases, and all had a reduction of energy intake with high nutritional risks on admission, therefore, EN application could induce the incidence of RS. Since RS often occurs within five days after the calorie intake, the related electrolyte levels, such as serum phosphorus, potassium, and magnesium, were monitored three days after EN to identify the incidence as well as the degree of RS in our study. Hypophosphatemia is often considered the hallmark of this syndrome. A rapid egress of phosphorus ions from intravascular to intracellular space after the EN initiation, results in hypophosphatemia [26]. The intracellular phosphate plays a key role in energy production and transfer and participates in the synthesis of adenosine triphosphate (ATP). Increased consumption of phosphate due to enhanced production of phosphorylated intermediates during refeeding leads to a reduced generation of ATP and 2,3-diphosphoglycerate, and impairs cardiac and respiratory functions in turn [27]. Besides hypophosphatemia, hypokalemia and hypomagnesemia are also of equal importance. Decreased serum potassium causes an imbalance in the electrochemical membrane potential and deteriorates the transmission of bio-electrical impulses as well as nerve impulses, resulting in abnormal cardiac rhythm and release of neurotransmitters [28]. Hypokalemia could also cause periodic paralysis in the neuromuscular junction [29]. Serum magnesium is essential for activating the sodium/potassium ATP-pump (Na + - K + - ATPase), and magnesium deficiency could aggravate potassium loss in cardiomyocytes, reflected by changes in the electrocardiogram (ECG) wherein PR and QT intervals are prolonged and QRS complex is widened [30].
In our study, the serum levels of electrolytes on admission were similar between two groups, and they were all under a normal range. Therefore, basic laboratory examinations were comparable. Three days after EN treatment, the serum levels of potassium, sodium, phosphorus, and magnesium were significantly lower in the LEN group than in the EEN group. Moreover, each of these electrolytes was below a normal range in the latter group, which suggests that later EN initiation might have an adverse effect on body homeostasis. We also observed a significantly higher incidence as well as more serious effects of RS in the LEN group. We chose two days as the cut-off to distinguish the early and late EN initiation because the intestinal dysbacteriosis always occurs within 48 h without normal gut nutrients support. Later initiation of EN further impairs the integral mucosal architecture and increases the risk of bacterial translocation [31]. Besides, later EN support is a risk factor to induce the incidence of RS. Interestingly, AUC for EN was 0.653, which is favorable for RS prediction. Additionally, logistical regression analysis also demonstrates that the hazard risk of RS is more than 4 times in the LEN group compared to the EEN group.
Two pivotal processes occur when EN is initiated. To begin with, increased insulin secretion drives glucose into cells, which results in an increased uptake of serum phosphorus, potassium, magnesium, calcium, and thiamine. Then, the cell membrane Na + - K + - ATPase starts actively, pumping sodium out of the cell and taking potassium into the cell. As a result, the uptake of serum phosphorus and magnesium further increased [32]. In view of the above two effects, early nutritional support is beneficial for regulating the body homeostasis to better tolerate changes in the internal environment in the process of nutrition input, and reduce the RS incidence.
The relationship between EN initiation and mortality in other respiratory diseases has been widely investigated. Chun Miao et al. (2017) investigated the effect of the standardized treatment of early EN on the prognosis of patients with acute respiratory distress syndrome (ARDS) and found that early EN support could improve the nutritional status, decrease the blood glucose fluctuations, and further benefit the 28-day mortality [33]. Subsequently, Yan Guo et al. (2018) showed a consistent result wherein early, and reasonable application of EN could reduce the incidence of infection, improve lung function, and reduce the duration of mechanical ventilation as well as the length of ICU stay of the ARDS patients [34]. In contrast, Sarah J. Peterson et al. (2017) suggested that early higher calorie exposure was associated with higher subsequent hazard of mortality for patients with ARDS, while provision after day 8 decreased the hazard [35]. An important finding in our research was that early initiation of EN for severe COVID-19 patients could significantly decrease the hospital stay length, as well as ICU stay length, without affecting the 6-month overall survival.
Finally, to evaluate the safety of early EN application, we evaluated the incidence of airway complications and GI intolerance. According to an earlier report, nearly twenty percent of critically ill patients, who received EN therapy during the period of noninvasive positive pressure ventilation, developed at least one adverse event, including the development of pneumonia, progressive respiratory failure, and so on [35]. Mariko Kogo et al. also collected data from subjects with acute respiratory failure and found that EN was a risk factor for airway complications, with a significantly longer duration of median noninvasive ventilation.17 For the other aspect, efforts to provide early EN are thwarted by GI intolerance, such as diarrhea, nausea, vomiting, GI bleeding, and so on [36]. Fortunately, the incidence of airway complications and GI intolerance were not different between the two groups in our study, possibly because of excellent care and regular monitoring in our hospital, and indicates that it is safe to start EN early for severe COVID-19 patients with high nutritional risk.
This study had several potential limitations. First, our analysis was based on a retrospective study with a relatively small sample; hence there may be a selection bias to limit the generalizability of our results. To verify our results, a large prospective study, over an extended period of time, is needed. Second, the subjects included in our study were under strict restrictions; severe COVID-19 patients had NRS scores ≥ 3. For other patients with mild symptoms and a risk of malnutrition, whether the early use of EN will decrease the RS incidence is still elusive. Last but not least, we analyzed only 6-month overall survival for all the participants, the cut-off point was too short, and longer survival statistics are eagerly awaited for further conclusion.