Effects of Nutritional Protocol Changes on Hemodialysis Adequacy and Patient Health During the COVID-19 Pandemic

Introduction

Nutritional practices during hemodialysis remain a controversial issue. Nutrition during hemodialysis is associated with a rapid postprandial decline in blood pressure and an increased risk of symptomatic hypotension [1–5]. Notably, a decrease in dialysis adequacy is another undesirable consequence [5–7]. Dry weight provides important insights into a patient’s nutritional status. In particular, a reduction in dry weight is a significant indicator of malnutrition in hemodialysis patients [8]. In an observational study conducted in Japan, the suspension of cafeteria services during the pandemic was associated with a decrease in dry weight [9].

Intradialytic hypotension (IDH) causes organ ischemia and is associated with accelerated loss of kidney function [10], decreased dialysis adequacy [5–7], increased vascular thrombosis [11], cerebrovascular events [11], cardiovascular mortality [11,12], and increased risk of all-cause mortality [11–13]. Recently, a randomized crossover study by Fotiadou et al [5] involving 26 patients demonstrated that the application of standard intradialytic meals with high-protein or low-protein content was associated with increased intradialytic blood pressure variability and reduced dialysis adequacy. Therefore, to reduce the frequency of IDH, the 2007 European Best Practice Guidelines recommend avoiding food intake during and immediately before dialysis [14]. However, preventing nutrition during dialysis may lead to skipping 3 meals a week in the standard hemodialysis regimen (4 hours each, 3 days per week) and protein-energy wasting (PEW), which is characterized by a decrease in body protein and energy storage in hemodialysis patients, who are already known to suffer from nutritional disorders [15,16]. Notably, PEW is associated with several adverse outcomes, including cardiovascular events and all-cause mortality [17]. Intradialytic nutrition may be seen as an opportunity to reduce the risk of malnutrition, and positive benefits might be achieved with regard to PEW and quality of life [18–21]. Accordingly, the 2018 consensus statement of the International Society of Renal Nutrition and Metabolism encouraged patients to ingest nutrition during dialysis [22].

The COVID-19 pandemic caused lifestyle changes in hemodialysis patients [23]. COVID-19 is transmitted through droplets and aerosols, and hemodialysis patients are at risk of severe COVID-19 [24,25]. Therefore, during the pandemic, to prevent transmission between patients and to ensure compliance with social distancing protocols, many centers implemented strict nutritional restrictions during dialysis. Even during periods when restrictions were relaxed or completely lifted, nutritional support was far from being at pre-pandemic levels [17,26].

There are limited reports evaluating nutritional changes during dialysis in the context of the COVID-19 pandemic. Moreover, many centers continue to impose restrictions on nutritional support practices due to financial concerns, even though pandemic-related restrictions are no longer in place. Before the COVID-19 pandemic, our center routinely provided nutritional support during dialysis. However, during the peak of the pandemic, we strictly discontinued this practice to comply with social distancing measures and prevent transmission among patients during meals.

Therefore, this retrospective study of patients undergoing dialysis during the COVID-19 pandemic feeding restrictions aimed to evaluate the effects of different feeding protocols on blood pressure, dialysis adequacy, and patient dry weight during dialysis.

Material and Methods

ETHICS:

This study was approved by the local ethics committee in our institution (2023/01-28) and conducted in accordance with the 1975 Declaration of Helsinki and its later amendments. All patients enrolled in the study provided informed consent.

STUDY DESIGN AND PARTICIPANTS:

This retrospective study was conducted between September 2019 and October 2020 in patients receiving hemodialysis treatment at Van Yuzuncu Yil University Hospital to examine the effect of nutritional habits on dialysis patients. The patients were examined separately in 2 periods: in the first period, when nutrition was routinely provided by the center during dialysis before the COVID-19 pandemic (September 2019–February 2020), and in the second period, when nutrition was prohibited during the pandemic (May 2020–October 2020). In these 2 consecutive 6-month periods with and without nutrition, the monthly visits of the patients were evaluated and the average of the data was recorded.

Patients who received hemodialysis treatment for at least 6 months were included in our study. The hemodialysis treatment procedure showed similar characteristics; in detail, dialysis duration (4 hours in 3 dialysis sessions per week), dialysis filter (high-flux membranes), dialysis solution (sodium bicarbonate), solution temperature (37°C), and dialysate flow rate (500 ml/min) were similar. However, dry weight, ultrafiltration amount, and pump speeds were different. Hemodialysis treatment was applied to the patients in morning and afternoon sessions, and the patient’s session plan was not changed during the study period. A standard 500 kcal meal prepared with the support of a nutritionist consisting of white cheese, olives, bread, eggs, and tea was given. The nutritional habits of patients between dialysis treatments could not be fully determined during either period, and were therefore not included in the analyses. Patients who died during the study period or lacked clinical and laboratory follow-up were omitted. COVID-19, other active infection, malignancy, rheumatological diseases, and physical and neurological conditions such as dementia that would hinder nutrition were determined as comorbid conditions that could cause poor life expectancy; hence these patients were also excluded.

METHODS:

Data on patients’ laboratory results, hemodialysis characteristics, drug use, and hemodialysis vascular access were compared across the feeding and no-feeding periods. In addition, other parameters reflecting the effects of intradialytic nutrition on hemodialysis adequacy and patient hemodynamics – such as dry weight, body mass index (BMI), predialytic systolic blood pressure (SBP), predialytic diastolic blood pressure (DBP), predialytic mean arterial pressure (MAP), urea reduction ratio (URR), single-pool Kt/V (spKt/V), ultrafiltration volume, ultrafiltration rate, number of intradialytic hypotension episodes per month, and number of intradialytic hypotension episodes over 6 months – were also compared.

DATA COLLECTION:

Data were collected by examining patient files and from the electronic database of the hospital. Demographic data [sex, age, height, weight, BMI], chronic kidney disease etiology (diabetes mellitus, hypertension, urologic causes, chronic glomerulonephritis, polycystic kidney disease), comorbid conditions (hypertension, diabetes mellitus, coronary heart disease, dyslipidemia, congestive heart disease), hemodialysis duration, dry weight, predialytic SBP, predialytic DBP, predialytic MAP, dialysate sodium, blood flow, dialysate flow, dialysate calcium, solution temperature, use of anti-hypertensive drugs [alpha-blockers, beta-blockers, angiotensin-converting enzyme inhibitors (ACEIs) or angiotensin receptor blockers (ARBs), calcium-channel-blockers, diuretics], and hemodialysis vascular access (arteriovenous fistula, arteriovenous graft, tunneled central venous catheter) were recorded. Laboratory data including urea, creatinine, potassium, calcium, uric acid, phosphate, albumin, C-reactive protein (CRP), parathyroid hormone (PTH), transferrin saturation, ferritin, total cholesterol, low-density lipoprotein (LDL)-cholesterol, high-density lipoprotein (HDL)-cholesterol, triglycerides, white blood cell (WBC) count, hemoglobin (Hb) level, and platelets were noted as well.

URR and spKt/V were used to evaluate dialysis adequacy. URR, spKt/V, ultrafiltration amount, ultrafiltration rate, and the number of IDH episodes occurring during hemodialysis sessions were retrieved from the patient files.

DEFINITIONS:

BMI was calculated using the formula (weight in kg)/(height in m)2. Patients’ laboratory tests were routinely measured once a month. Each patient was evaluated by the same nephrologist at every hemodialysis session. Dry weight was defined as the lowest tolerable weight at which the patient exhibited minimal symptoms or signs of hypovolemia or hypervolemia [27]. Dry weight determination was conducted using multiple methods, including laboratory tests, blood pressure measurement, volume assessment, chest X-ray findings, ultrasonographic measurement of the inferior vena cava diameter, and blood volume monitoring [28]. To determine dialysis adequacy, blood samples were taken from the patients using the slow flow method at the beginning and end of the middle dialysis sessions of the first week of each month, and plasma urea concentrations were determined [29]. URR was calculated with the following formula: URR = (pre-hemodialysis urea − post-hemodialysis urea)/pre-hemodialysis urea ×100. The spKt/V was calculated using the Daugirdas formula: spKt/V=−ln(R−0.008×t)+(4−3.5×R)×UF/W, where R is the ratio of pre- to post-hemodialysis concentration of blood urea nitrogen (BUN), t is dialysis session length (hours), UF is the volume removed in dialysis (L) and W is post-hemodialysis weight (kg) [30,31]. IDH was defined as at least a 20 mmHg decrease in SBP accompanied by clinical symptoms/events and the need for intervention [14].

STATISTICAL ANALYSIS:

Descriptive data were recorded as frequencies and percentages for categorical variables, while continuous variables were presented as medians (interquartile range [IQR]) or mean±standard deviation (SD), according to the distribution pattern). Comparisons for continuous variables between the 2 phases of our study were made with paired t-test or Wilcoxon signed-rank test. Statistical analyses were performed using SPSS for Windows (SPSS version 22.0, IBM Corp., Armonk, NY). All tests were 2-sided and a P-value of 0.05 or less was considered statistically significant.

Results

BASELINE CHARACTERISTICS OF THE STUDY POPULATION:

In total, 76 patients (58% male) were included in this study. The mean age of the patients was 58.9±15.6 years. Diabetes mellitus was the most common cause of chronic kidney disease (32 patients, 42.1%), which was followed by hypertension (12 patients, 15.8%) and urologic causes (9 patients, 11.8%). Hypertension, diabetes mellitus, and coronary heart disease were the most common comorbid conditions. The median hemodialysis duration was 85 (60–142.8) months. The baseline demographic and clinical characteristics of the patients are summarized in Table 1.

OUTCOMES DURING FEEDING AND NON-FEEDING PERIODS:

The laboratory and clinical outcomes of the patients in the feeding and no-feeding periods during dialysis are presented in Table 2. The dry weight and BMI of the patients were higher during the feeding period than during the no-feeding period, but this difference was not statistically significant (dry weight, feeding: 65.7±14.2 kg vs dry weight, no-feeding: 65.4±13.7 kg, P=0.62; and BMI feeding: 24.7±5.0 kg/m2 vs BMI no-feeding: 24.4±4.89 kg/m2, P=0.11). Predialytic SBP, DBP and MAP values of the patients were statistically significantly higher during the feeding period than during the no-feeding period [SBP feeding: 140 (130–140) mmHg vs SBP no-feeding: 130 (120–140) mmHg, P=0.004; DBP feeding: 80 (80-80) mmHg vs DBP no-feeding: 80 (80-80) mmHg, P=0.013; and MAP feeding: 100 (97–100) mmHg vs MAP no-feeding: 97 (93–100) mmHg, P=0.004]. Between the 2 periods, dialysate sodium, the antihypertensive treatment used by the patients, hemodialysis vascular access, predialytic urea, creatinine, potassium, uric acid, phosphate, CRP, transferrin saturation, total cholesterol, LDL-cholesterol, HDL-cholesterol, and Hb values were similar. During the feeding period, blood flow was significantly lower and dialysate calcium was higher than in the no-feeding period [blood flow, feeding: 350 (320–350) mL/min vs blood flow, no-feeding: 360 (350–380) mL/min, P=0.01, and dialysate calcium, feeding: 1.25 (1.25–1.5) mmol/L vs dialysate calcium, no-feeding: 1.25 (1.25–1.5) mmol/L, P=0.04, respectively]. Serum calcium, albumin, and ferritin levels were significantly higher in the no-feeding period than in the feeding period [calcium, no-feeding: 9.1 (8.8–9.7) mg/dL vs calcium, feeding: 8.8 (8.4–9.0) mg/dL, P=0.01; albumin, no-feeding: 4.0 (3.9–4.2) mg/dL vs albumin, feeding: 3.9 (3.8–4.1) mg/dL, P=0.002; and ferritin, no-feeding: 920.4±474.6 pg/L vs ferritin feeding: 686.3±408.9 pg/L, P=0.001]. Serum PTH, triglycerides, WBC count, and thrombocyte levels were statistically significantly higher in the feeding period than in the no-feeding period [PTH, feeding: 484 (253.3–911.8) ng/L vs PTH, no-feeding: 314.5 (141.3–536.8) ng/L, P=0.01; triglycerides, feeding: 182.5 (122.5–249.5) mg/dL vs triglycerides, no-feeding: 159.5 (107.5–196) mg/dL, P=0.001; WBC, feeding: 5.4±1.5 103/mm3 vs WBC, no-feeding: 4.8±1.6 103/mm3, P=0.005; and thrombocytes, feeding: 201 (157–241.7) 103/mm3 vs thrombocytes, no-feeding: 174 (136.5–211) 103/mm3, P=0.01, respectively].

DIALYSIS ADEQUACY AND IDH FREQUENCY DURING FEEDING AND NO-FEEDING PERIODS:

Dialysis adequacy between the 2 periods was compared. URR and spKt/V were statistically significantly higher in the no-feeding period than in the feeding period [URR, no-feeding: 72.5% (69–76 patients) vs URR, feeding: 71% (68–75 patients), P=0.01; and spKt/V, no-feeding: 1.59±0.23 vs spKt/V, feeding: 1.52±0.26, P=0.004]. Ultrafiltration amount (mL per hemodialysis session) and ultrafiltration rates (mL/h/kg) in hemodialysis sessions were statistically significantly higher in the no-feeding period than in the feeding period (ultrafiltration amount, no-feeding: 2930.6±777.3 vs ultrafiltration amount, feeding: 2701.6±825.2, P=0.01; and ultrafiltration rate, no-feeding: 11.4±3.0 vs ultrafiltration rate, feeding: 10.4±3.2, P=0.01). The number of IDH was statistically significantly higher in the feeding period than in the no-feeding period [feeding: 1.4±1.2 per month vs no-feeding: 0.81±0.86 per month, P=0.01; and feeding: 7 (3–12.2) per 6 months vs no-feeding: 3 (1–7) per 6 months, P=0.0].

Discussion

During the COVID-19 pandemic, behavioral changes in nutritional habits during dialysis were enforced. In the present study, the effect of these changes on dialysis adequacy and the frequency of IDH was investigated. This is the first study on the subject in Turkey. Our findings illuminate a distinctive trend: feeding during dialysis is associated with worse outcomes in dialysis adequacy, amount of ultrafiltration, ultrafiltration rate, and frequency of IDH compared with no-feeding. Our results are consistent with previous studies showing that feeding during dialysis causes a rapid postprandial blood pressure decrease, leading to an increase in the frequency of IDH [1–5], and is associated with a decrease in dialysis adequacy [5–7]. Our results underline another important point: lack of feeding during dialysis does not cause a significant decrease in dry weight, BMI, and serum urea levels compared with dialysis with feeding allowed. In this respect, our study differs from previous studies reporting that low energy intake on dialysis days due to neglect of nutrition during dialysis is associated with PEW [10–13,32], and a decrease in dry weight and serum urea levels [9].

The effect of intradialytic nutrition on deterioration of hemodynamic stability during dialysis has attracted the attention of many researchers. Some retrospective observational studies have reported that oral feeding during dialysis is not associated with an increased incidence of IDH, but these studies suffer from various limitations [33,34]. On the other hand, interventional studies show that nutrition during dialysis is associated with an increase in the frequency of IDH by causing a rapid postprandial decrease in blood pressure [1,2,4,5,35]. Despite its retrospective observational design, our study supports these interventional studies. An increase in the incidence of IDH may cause recirculation of the circulating blood volume towards the splanchnic area in the postprandial period, leading to dialysis failure [36,37]. In our study, predialytic SBP, DBP, and MAP were higher in the feeding period than in the no-feeding period. This may be associated with a lower amount of ultrafiltration and hypervolemia. Notably, in a recently published interventional, randomized crossover study by Fotiadou et al [5], 26 patients were randomized into 2 groups with a high- and low-protein meal in the first period. One week later, the patients were crossed. In that study, feeding during dialysis, regardless of high- or low-protein meal, was observed to be associated with higher intradialytic SBP variability and reduced adequacy of dialysis.

After March 11, 2020, when the first COVID-19 case was seen in our country, restrictive measures were implemented in dialysis facilities as in many areas. In April, the provision of meals in facilities was strictly stopped. These restrictions were widely maintained even after periods when the pandemic was severe. A survey study [17], which started with 55 dialysis centers, highlighted the dramatic decrease in the food provision of the centers during or after dialysis due to the COVID-19 pandemic. Food provision by the centers was 52% in 2018. However, with the pandemic, it was observed that it decreased to 33% in 2020 and to 27% in 2022. Notably, the study was conducted in Japan, where nutrition during dialysis is seen as an opportunity for addressing energy needs [3,38].

Discontinuation of food service in dialysis facilities may be a risk factor for the development of PEW, which is associated with death in hemodialysis patients, especially from cardiovascular causes [39,40]. The most common cause of PEW is inadequate protein and energy intake, and low energy intake by hemodialysis patients on dialysis days has already been known to occur [41]. Dry weight is one of the dynamic variables of dialysis and is a good indicator of PEW [42], and it may vary depending on changes in lean body mass and body fat [43]. Therefore, reduced dry weight is considered evidence of malnutrition in dialysis centers [8]. A very recent study [9], which included 204 elderly patients (≥65 years of age) receiving hemodialysis treatment, 79 of whom were cafeteria users, examined the relationship between the cessation of cafeteria service due to COVID-19 pandemic measures and the change in their dry weight. During the observation period, a significant decrease in dry weight was observed in cafeteria users, but not in non-cafeteria users. Multivariate logistic regression analysis revealed a close relationship between reducing dry weight and discontinuing cafeteria service. Fragility, an important characteristic of the elderly, might also have a role in these findings. Therefore, it is more difficult for older adults to adapt to changes in restrictions and eating behaviors associated with the COVID-19 pandemic compared with younger adults [44,45]. In a previous study, we observed that the mood profiles of hemodialysis patients were better than those of peritoneal dialysis patients during the social isolation period due to the COVID-19 epidemic [46]. Constant communication with the patient’s dialysis facility and continued access to health services may have prevented negative effects on eating habits. Another important issue is that non-dialysis nutrition is often neglected in intradialytic nutrition studies. Patients may have met their protein and energy needs through meals at home.

Our study has some limitations. First of all, due to the retrospective design, data were obtained from medical records. Food intake and IDH may not have been recorded while the patient was on dialysis. Since the study was conducted in a single center, the results may have been affected by the center’s treatment approach and characteristics of the patients at that center. Our study does not include a control group due to the mandatory restrictions imposed during the COVID-19 pandemic, which required strict adherence to social distancing measures. Another significant limitation is the lack of data on the patients’ nutritional habits between dialysis treatments. Moreover, despite a significant stress factor such as the COVID-19 pandemic, we did not have access to any information on the patient’s emotional state.

Conclusions

Our study explores the effects of discontinuing nutrition during dialysis in hemodialysis patients due to the COVID-19 pandemic. We observed that the absence of intradialytic nutrition was associated with a decrease in the frequency of IDH and an improvement in dialysis efficiency. Notably, patients did not show significant changes in their dry weight during this period, indicating that their health status remained relatively stable despite the nutritional restrictions. This research is among the first to examine the impact of these dietary changes on hemodialysis patients in Turkey during the pandemic. It suggests that further studies are warranted to investigate the effects of nutritional conditions outside of dialysis periods on patient outcomes.