European Geriatric Medicine (2021) 12:1201–1212

https://doi.org/10.1007/s41999-021-00520-5

RESEARCH PAPER

A comparison of two different refeeding protocols and its effect

on hand grip strength and refeeding syndrome: a randomized
controlled clinical trial
Sissel Urke Olsen1 · Karin Hesseberg2 · Anne‑Marie Aas3 · Are Hugo Pripp4 · Anette Hylen Ranhoff5,6 · Asta Bye7,8

Received: 12 March 2021 / Accepted: 19 May 2021 / Published online: 4 June 2021
© European Geriatric Medicine Society 2021, corrected publication 2022

Key Summary Points

Aim Does a more assertive compared to a cautious initial refeeding protocol improve patient handgrip strength?
Findings In this randomized clinical trial of 85 older adults whom received two different refeeding protocols, with tight
control of electrolytes, no significant difference in in by handgrip strength was found (p = 0.78).
Message A more assertive refeeding protocol providing 20 kcal/kg/day did not improve hand grip strength nutritional status
at 3 months compared with a cautious refeeding protocol (10 kcal/kg/day), neither did it show a higher incidence of mortality.

Abstract
Purpose Optimal refeeding protocols in older malnourished hospital patients remain unclear. We aimed to compare the
effect of two different refeeding protocols; an assertive and a cautious protocol, on HGS, mortality and refeeding syndrome
(RFS), in patients ≥ 65 years
Methods Patients admitted under medical or surgical category and at risk of RFS, were randomized to either an enteral
nutrition (EN) refeeding protocol of 20 kcal/kg/day, reaching energy goals within 3 days (intervention group), or a protocol
of 10 kcal/kg/day, reaching goals within 7 days (control group). Primary outcome was the difference in hand grip strength
(HGS) at 3 months follow-up, in an intention to treat analysis. RFS (phosphate < 0.65 mmol/L) during the hospital stay and
mortality rates at 3 months were secondary outcomes.
Results A total of 85 patients were enrolled, with mean (SD) age of 79.8(7.4) and 54.1% female, 41 in the intervention group
and 44 in the control group. HGS was similar at 3 months with mean change of 0.78 kg (95% CI − 2.52 to 3.36, p = 0.42).
Serum phosphate < 0.65 mmol/L was seen in 17.1% in the intervention group and 9.3% in the control group, p = 0.29. There

3
* Sissel Urke Olsen Faculty of Medicine, Institute of Clinical Medicine,
sisselurke@gmail.com University of Oslo, Oslo, Norway
4
Karin Hesseberg Faculty of Health Sciences, Oslo Metropolitan University
Karin.hesseberg@diakonsyk.no (OsloMet), Oslo, Norway
5
Anne‑Marie Aas Department of Clinical Science, University of Bergen,
a.m.aas@medisin.uio.no 7804 Bergen, Norway
6
Are Hugo Pripp Medical Department, Diakonhjemmet Hospital,
apripp@oslomet.no Bergen, Oslo, Norway
7
Anette Hylen Ranhoff Department of Nursing and Health Promotion, Faculty
Anette.Ranhoff@diakonsyk.no of Health Sciences, Oslo Metropolitan University, Oslo,
Norway
Asta Bye
8
abye@oslomet.no Regional Advisory Unit for Palliative Care, Department
of Oncology, Oslo University Hospital, Oslo, Norway
1
Department of Clinical Dietetics, Diakonhjemmet Hospital,
Oslo, Norway
2
Department of Physiotherapy, Diakonhjemmet Hospital,
Oslo, Norway

13
Vol.:(0123456789)
1202 European Geriatric Medicine (2021) 12:1201–1212

was no difference in mortality rates (39% vs 34.1%, p = 0.64). An indication of more respiratory distress was found in the
intervention group, 53.6% vs 30.2%, p = 0.029.
Conclusion A more assertive refeeding protocol providing 20 kcal/kg/day did not result in improved HGS measured 3 months
after discharge compared with a cautious refeeding (10 kcal/kg/day) protocol. No difference in incidence of mortality or
RFS was found.
Trial registration ClinicalTrials.gov Protocol Record 2017/FO148295, Registered: 21st of February, 2017.

Keywords Malnutrition · Enteral nutrition · Hand grip strength · Refeeding syndrome · Phosphate

Introduction high-quality indicators to demonstrate the effects of nutri-

tional interventions [21]. Hand grip strength (HGS) used as
Hospitalized older adults are vulnerable to malnutrition a marker for muscle strength, but also tested as a biomarker
with prevalence rates ranging from 4 to 40% [1]. Provid- for nutritional status [22–27]. Muscle strength reacts early
ing appropriate nutritional care during a hospital stay has to inadequate nutritional intake [24, 27] and a response
the potential to improve nutritional status and thus improve to nutritional interventions are often quantified earlier by
key quality indicators such as readmissions, length of stay changes in muscle strength rather than in body composition
(LOS), infections, mortality, and quality of life [2–4]. When [24]. This coincides with the fact that the 2018 consensus
treating malnutrition, enteral nutrition (EN) is one of many on diagnosing malnutrition, developed by Global Leadership
nutritional interventions, but it carries a risk of develop- Initiative on Malnutrition, recommend using HGS as part
ing refeeding syndrome (RFS) [5]. RFS is not fully under- of the diagnostic criteria for malnutrition along with weight
stood, but it can be characterized as severe electrolyte and loss and low BMI [28].
fluid shifts occurring when the body shifts from a catabolic Refeeding too cautiously can increase risk of malnutri-
state, using mainly fat and protein as an energy source, to tion, postpone recovery of illness and even lead to unneces-
an anabolic state when the metabolism regains access to sary loss of physical function [29]. The primary aim of this
carbohydrates [6–8]. This may result in fluid and electro- study was, therefore, to assess the effect of a refeeding pro-
lyte imbalance, especially hypophosphatemia, and neuro- tocol initially providing 20 kcal/kg/day on muscle strength
logic, pulmonary, cardiac, neuromuscular, and hematologi- at 3 months follow-up as an indicator, and compare it to the
cal complications [9, 10]. To prevent development of RFS, current, cautious NICE protocol (5–10 kcal/kg/day) in mal-
the National Institute of Health and Clinical Excellence nourished older hospital patients. The secondary outcomes
(NICE) published guidelines in 2006, [11], updated in 2017 were occurrence of RFS during hospital stay and 3 months
[12], recommending a cautious refeeding EN protocol of mortality. We hypothesized that the more assertive refeeding
5–10 kcal/kg/day for patients at risk for RFS [11]. How- protocol would improve muscle strength and reduce mortal-
ever, the supporting literature underpinning this recommen- ity, but potentially aggravate RFS.
dation was mostly based upon cohort studies, case series
and consensus expert opinions[7]. In recent years, treatment
protocols for patients with eating disorders have changed Methods
from using cautious approaches (as exemplified by the NICE
guidelines), to a more assertive approach, acknowledging Design and participants
the benefits of meeting estimated needs earlier in this group
[13–15]. This has resulted in shorter LOS and better weight This study was a parallel-group semi double-blinded ran-
gain without compromising medical safety, including the domized controlled trial (RCT) with two arms. The nurses
complication of RFS [16]. These findings seem contrary to administering the EN were not aware of the study details
studies in critically ill patients where energy restriction, slow whilst the assessor and patients were blinded to the group
titration rates and close monitoring of electrolytes, resulted allocation. Reporting follows the CONSORT 2010 statement
in less infections, renal therapy and gastrointestinal intoler- [30].
ance as well as improved overall survival when compared Older men and women ≥ 65 years were recruited between
with standard care ranging from > 50% to > 70–80% of goal March 2017 and August 2019, from both medical and surgi-
rate at day one [17–20]. cal acute care wards in a general hospital in Oslo, Norway.
Despite expectations that EN will contribute to improved Patients at high or moderate risk of RFS according to the
nutritional status, empirical evidence of optimal refeeding NICE screening criteria tool (Table 1), that agreed to be
strategies in older hospital patients is lacking. A challenge hospitalized during the entire intervention period of 7 days,
when it comes to conducting such studies, is the lack of were eligible. Patients included were clinically considered to

13
European Geriatric Medicine (2021) 12:1201–1212 1203

Table 1  NICE guideline screening criteria tool for detecting patients at risk for refeeding syndrome
Risk of developing refeeding syndrome
High risk Moderate risk

One or more of the following: BMI < 16 kg/m2 Two or more of the following: BMI < 18.5 kg/m2
Unintentional weight loss > 15% within the last 3–6 months Unintentional weight loss > 10% within the last 3–6 months
Little or no nutritional intake for > 10 days Little or no nutritional intake > 5 days
Low levels of potassium, phosphate or magnesium prior to feeding A history of alcohol abuse or drugs including insulin, chemotherapy,
antacids or diuretics

BMI, body mass index

be in need of EN to meet their energy requirements. Exclu- older adults with acute and chronic diseases. For patients
sion criteria were advanced and end-stage dementia (accord- with renal failure (glomerular filtration rate < 30) a lower
ing to the International classification of Diseases, ICD 10 target of protein (0.8 g/kg/day) was used [31]. Fluid require-
codes and an assessment of the responsible physician), those ment was set at 30 ml/kg/day as standard unless medical
already receiving EN or intravenous nutrition, not wishing therapy required otherwise. The DH group’s protocol started
to be inpatients for at least seven days, admission directly to at 20 kcal/kg/day, and increased to meet estimated needs
the intensive care unit, previous participation in the study, within 3 days. NICE group protocol started at either 5 kcal/
terminally ill patients or those with a short life expectancy kg/day (body mass index (BMI) < 14) or 10 kcal/kg/day
(weeks or few months). (BMI > 14), and increased to meet estimated needs within
seven days. Phosphate, magnesium, potassium, sodium,
Study procedures blood sugar, and C-reactive protein were monitored daily in
both groups. An electrolyte replacements protocol was used
Patients were recruited by the main investigator, clini- [32]. Electrolytes were supplemented when phosphate lev-
cal dietitians and medical doctors. The main investigator els were < 0.65 mmol/L, magnesium levels < 0.65 mmol/L
screened patient lists according to the inclusion and exclu- and potassium levels < 3.6 mmol/L, and were given orally
sion criteria and participated in pre-ward round meetings or intravenous according to level of deficiencies. [11] All
and ward rounds. Eligible patients or their proxy were then patients were prescribed Thiamine 100 mg before EN and
asked to participate and informed consent obtained. There- for 10 days, or until discharged from the hospital. In addi-
after, patients were randomized and enrolled to either (1) tion, one multivitamin/mineral tablet, and three tablets of
the intervention group receiving a more assertive refeeding vitamin B complex (Thiamine 1.4 mg, Riboflavin 1.7 mg,
protocol (Diakonhjemmet Hospital guidelines (DHgroup) or Vitamin B6 1.6 mg, Vitamin B12 2 µg, Folate, 400 µg,
(2) the control group receiving a cautious refeeding protocol Niacin 19 mg, Pantothenic acid 5 mg, Biotin 30 µg), were
(NICE guidelines (NICE group)) [11]. administered daily. The refeeding plan was copied into the
The sample size was calculated based on the primary end- patient’s electronic chart.
point HGS. Given a power of 80% and an alpha = 0.05, we Due to ethical reasons, patients were not denied oral
aimed at including 126 patients, allowing for 20% dropouts, intake, although oral nutritional supplements were discour-
to detect a difference of 1 kg (SD 1.8) between the groups aged for patients in the study, except from patients who had
at 3 months follow-up. discontinued the EN due to various reasons. Oral intake was
A computer-generated block randomization schema registered in the patients’ electronic medical chart by the
designed by a statistician at the Faculty of Health Science nurses, and the main investigator obtained the information
at Oslo Metropolitan University was used to allocate the and analysed the energy intake for each patient.
patients. Randomizations blocks ranged from four to ten.
The randomization list was not accessible to the main inves- Outcomes measures
tigator, nor was the block size. An external assistant pro-
vided the treatment allocation by phone. Outcome assessments were carried out by the main investi-
After randomization, a trained nurse inserted the feed- gator at baseline and by a blinded assessor on day four, day
ing tube, and a radiograph confirmed correct placement. A seven and at 3 months follow-up. The same assessor was
high protein EN formula was used (1.25 kcal/ml, 6.3 g of used for each patient whenever feasible.
protein/ml, and 80% water content) in both groups. Energy The primary outcome was change in muscle strength
requirement was estimated to be 30 kcal/kg/day and pro- measured as HGS at 3 months follow-up. It was measured
tein requirements at 1.2–1.5 g /kg/day as recommended for using a calibrated handheld digital dynamometer (JAMAR

13
1204 European Geriatric Medicine (2021) 12:1201–1212

­PLUS® Hand dynamometer, Patterson Medical, USA), considers the entire data set. In addition, it adjusted for base-
adjusted to appropriate grip width [33]. Standardised line values as part of the procedure [37, 38]. Chi-Square test
instructions and demonstration were given. Each patient for independence was used to analyze possible differences
completed two trials on each hand, with their arms in a 90 in RFS and mortality. No imputation of missing data was
degree angle, sitting in a chair, and the best result of domi- performed. The data was analysed following the intention
nant hand were analysed as the final HGS (value expressed to treat principle (ITT), and all randomized patients were
in kilograms) [34]. Due to underlying diseases some patients included, whether they completed the trial or not.
were confined to bed, but tests were performed in the same
position each time. Secondary outcome measures were Ethics
RFS during hospital stay and 3 months mortality. RFS was
defined as any occurrence of phosphate < 0.65 mmol/L dur- Informed consent was obtained from all patients or their
ing the 7 days of refeeding [18, 19, 35, 36]. proxy/next of kin. If the patient was not able to give written
Mortality is given as the number of deaths in each group consent due to cognitive decline or severe illness, a next of
counted from EN start until 3 months follow-up. Date of kin was asked to confirm and sign the consent. The study
death was extracted from the patients’ electronic medical protocol was approved by The Regional Committee for Med-
chart. ical and Health Research Ethics in South-Eastern Norway
After enrollment, the following information was collected (REC number 2015/676).
from each patient; age, gender, weight, height, primary diag-
nosis, comorbidity, number of disease diagnoses and living
arrangements and education. Height was measured on day Results
one either standing or using a laser gauge for patients not
able to stand. Patients were weighed in light clothing on From April 1st 2017 to August 29th 2019, 167 patients
either an electronic chair scale (SECA, Germany) or on a were screened, of which 85 signed the informed consent
standing electronic scale (Soehnle scale). BMI was calcu- and were included (Fig. 1). Compared to those included, the
lated based on height and weight. Blood analysis of phos- 82 not included patients were similar in age but had a higher
phate, magnesium, sodium, potassium, C-reactive protein percentage of females. The two main reasons for declin-
and glucose were ordered daily by the medical doctor or by ing participation were that EN seemed unpleasant (41.5%)
the main investigator. To evaluate possible adverse events of and refusing to be inpatients for 7 days (11%). Forty-one
EN during the 7 days of refeeding, the following information patients were randomly assigned to the DH group and 44 to
was retrospectively collected from the patients’ electronic the NICE group. Seven patients in DH group and 11 in the
records as noted by a nurse or medical doctor; diarrhea, con- NICE group did not tolerate the feeding tube and discon-
stipation, nausea, vomiting, aspiration, edema, admission to tinued the treatment during the first 1–2 days, but they still
intensive care unit, and respiratory distress. After recom- continued to be followed up in the study and included in the
mencing nutrition, the presence of one of dyspnea, oxygen intention to treat analysis. The average time from admission
saturation fall, or pulmonary crepitations were deemed to to study day one was mean 4.6 (3.43) days in the DH group
indicate respiratory distress. LOS was also recorded and it and 4.5 (4.3) days in the NICE group. Baseline character-
was defined as when patients were medical cleared and ready istics were similar in the groups (Table 2). Mean (SD) age
for discharged, regardless of EN protocol. was 79.8 (7.4) years, BMI 18.5 (3.4) kg/m2, and weight 55.8
(12.4) kg. Forty-six (54%) patients were female. Accord-
Statistical analysis ing to the NICE guideline screening tool (Table 1), 41.2%
were considered as high risk of RFS, and 58.8% as moderate
Statistical analyses were performed with the software IBM risk of RFS. The most frequent admission diagnoses were
SPSS v25. A p < 0.05 indicated statistical significance. Data pneumonia, chronic obstructive pulmonary disease (COPD),
are presented as the mean and standard deviation (SD) for cancer and hip fracture. All patients, except from one in the
continuous variables and as number and percentages for NICE group, had one or more comorbidities, with chronic
categorical variables. Differences between the intervention obstructive lung disease (COPD), atrial fibrillation, cancer
group and control group were assessed by t tests for con- and dementia as the most frequent diseases.
tinuous data, with non-parametric tests for ordinal data and
with the Chi-squared test for independence for categorical Dietary intake
data. Mixed models with a random intercept and adjusting
for the outcome variable at baseline were used for the main Mean energy intake (EN, oral and total intake) in the
statistical assessment of primary endpoint with repeated two groups from baseline to day seven is presented in
measurements. This model is robust for missing data and Table 3. The DH group had significant higher mean energy

13
European Geriatric Medicine (2021) 12:1201–1212 1205

Assessed for eligibility Not included (n=82)

(n=167) • Did not meet inclusion criteria
(n=2)
Enrollment • Did not want tube feeding
Baseline: (n=34)
• Did not want to be in a study
(n=12)
• Too sick (n=9)
• Did not wish to be inpatient for7
days (n=9)
• Eating full meals (n=7)
Randomized (n= 85) • Other (n=9)

Allocated to intervention DH (n=41) Allocation Allocated to control NICE (n = 44)

Received allocated intervention (n = 41) -Received allocated intervention
(n= 44)

Lost to follow-up on day 4 and 7 (n=8) Follow-Up hospital Lost to follow-up on day 4 and day 7 (n=7)
Day 4 Died (n=3) stay Day 4 Died (n=2)
Day 7 Died (n=6) Day 7 Died: (n=1)
Discharged before day 7 (n=2) Discharged before day 7 (n=4)

3 months follow-up Analysed *HGS at 3 months (n=18)

Analysed* HGS at 3 months (n=14)
Lost to follow-up (Too sick, moved, did not
Lost to follow-up (Too sick, moved, did not
want follow-up, could not reach) (n=7)
want follow-up, could not reach) (n=9)
Died between day 7 and 3-months follow-up
Died between day 7- and 3-months follow-up
(n=12)
(n=10)
Analyzed intenon to treat (n-44)
Analyzed intenon to treat (n-41)

Fig. 1  CONSORT flowchart for enrollment, allocation, follow-up and number of patients analyzed. HGS hand grip strength

provision from EN on day one (800 kcal versus 447 kcal, Changes in hand grip strength
p < 0.001) through day three (1041 kcal versus 640 kcal,
p < 0.001), than the NICE group. From day four the EN There was no significant difference between the interven-
provision between the groups was similar. No difference in tion and the control group at the 3 months follow-up for the
mean oral energy intake was seen during the same period. primary outcome HGS (0.78 kg, 95% CI − 2.52 to 3.36,
At day seven the DH group had a total mean (SD) energy p = 0.42) (Table 4). Although HGS increased in both groups,
intake of 27.5 (15.6) kcal/kg/day compared to 27.4 (14.2) the change from baseline to 3 months follow-up within the
kcal/kg/day in the NICE group (p = 0.90). On day three groups was not significant. During the follow-up period
and seven neither of the groups met their daily energy there was large attrition of patients able or willing to per-
requirements by EN alone as determined by the refeeding form the HGS test. At day seven, 31(30%) were missing,
protocol. However, the DH group met the estimated energy and at 3 months 53 (62%) were missing. Number of patients
need of 30 kcal/kg/day, on days five and six. All patients changing from under normal score of HGS (< 16 kg for
received thiamine during their hospital stay, with an aver- women, and < 27 kg for men) at day seven, to normal HGS at
age of 4.6 days in both groups. 3 months were three men and two women The most common

13
1206 European Geriatric Medicine (2021) 12:1201–1212

Table 2  Baseline characteristics Total (n = 85) DH group (n = 41) NICE group (n = 44)
of the intervention group (DH
group) and control group (NICE Age in years, mean (SD) 79.8 (7.4) 80.9 (7.9) 78.7 (6.8)
group) receiving two different
Weight in kg, mean (SD) 55.8 (12.4) 54.7 (13.4) 56.9 (11.4)
enteral nutrition refeeding a
protocols BMI, mean (SD) 18.5 (3.4) 18.4 (4.0) 18.7 (2.8)
Gender, female, n (%) 46 (54.1) 22 (53.7) 24 (54.5)
b
High risk for RFS, n (%) 35 (41.2) 19 (46.3) 16 (36.4)
c
Moderate risk for RFS, n (%) 50 (58.8) 22 (44.0) 28 (56.0)
d
Malnourished 85 (100) 41 (48.) 44 (51.8)
HGS day 1, mean (SD) 15.1 (9.6) 16.4 (7.9) 13.9 (11.2)
Living alone, n (%) 42 (49) 19 (46.3) 23 (52.3)
Living with partner, n (%) 34 (40) 19 (46.3) 15 (34.1)
Nursing home, n (%) 9 (10.6) 3 (7.3) 6 (13.6)
Education, lower, n (%) 39 (45.9) 17 (43.6) 22 (56.4)
Education, higher (college), n (%) 46 (54.1) 24 (52.1) 22 (48)
Admission diagnoses, n (%)
Pneumonia 23 (27.1) 13 (31.7) 10 (22.7)
COPD 11 (12.9) 5 (12.2) 6 (13.6)
Cancer 8 (9.4) 2 (4.9) 6 (13.6)
Hip fracture 8 (9.4) 4 (9.8) 4 (9.1)
Urinary tract infection 4 (4.7) 1 (2.4) 3 (6.8)
Malnutrition 5 (5.9) 3 (7.3) 2 (4.5)
Congestive heart failure 5 (5.9) 2 (4.9) 3 (6.8)
Liver disease 2 (2.4) 1 (2.4) 1 (2.3)
Dehydration 4 (4.7) 3 (7.3) 1 (2.3)
Others 15 (17.6) 7 (17.0) 8 (18.2)
Nr of comorbidities, mean (SD) 2.5 (1.2) 2.6 (1.1) 2.38 (1.24)
0, n (%) 1 (1.1) 0 1 (2.3)
1, n (%) 16 (18.8) 8 (19.4) 8 (18.2)
2, n (%) 34 (40.0) 14 (34.1) 20 (45.5)
3, n (%) 18 (21.1) 11 (26.8) 7 (15.9)
4, n (%) 10 (11.7) 5 (12.2) 5 (11.4)
5 and 6, n (%) 6 (7.0) 3 (7.3) 3 (6.8)

Data are number (n) and percent (%), or means and standard deviation (SD). BMI, body mass index;
COPD, chronic obstructive lung disease; HGS, hand grip strength; n, number of participants
a
BMI, Body mass index calculated as the weight in kilogram divided by the square of the height in meters
b
High risk for developing refeeding syndrome, RFS, according to the NICE criteria
c
Moderate risk for developing refeeding syndrome, RFS, according to the NICE criteria
d
ICD 10th E44, mild or moderate protein-caloric malnutrition, E43, severe protein-calorie malnutrition

reason for loss to follow-up testing was death, although other 13.6 days in the NICE group (p = 0.68). Weight change and
factors included: worsening of pain and underlying condi- the incidence of RFS were similar in the two groups during
tion, no longer able to follow instructions, did not want to follow-up (Table 4).
be tested or early discharge. Respiratory distress developed in 22 (53.6%) patients in
the DH group and 13 (30.2%) in the NICE group (p = 0.029).
Secondary outcomes Sixteen of these patients were diagnosed with pneumonia
subsequent to admission, 11 in the DS group and five in the
During the seven-day intervention period, six patients died NICE group. Overall, there was no significant difference
in the DH group and three in the NICE group. The overall between respiratory distress in the high-risk RFS group ver-
mortality at 3 months was 16 (39%) and 15 (34.1%), in the sus the moderate risk RFS group.
DH group and NICE group, respectively (p = 0.64). Mean Nineteen (22%) patients experienced other adverse events
(SD) LOS was 14 days (10.0), 14.5 days in the DH and after recommencing EN including; diarrhea, vomiting,

13
European Geriatric Medicine (2021) 12:1201–1212 1207

Table 3  Energy intake from a

DH group b
NICE group
enteral nutrition, oral energy
intake and total energy intake Day n= Mean (SD) n= Mean (SD) Mean difference 95% CI P value
in the DH group and the NICE
Enteral nutrition intake (kcal/day)
group
1 41 800 (436) 44 447 (245) 353 202 to 507 < 0.001
2 41 950 (562) 43 530 (369) 420 214 to 625 < 0.001
3 41 1041 (733) 42 640 (457) 401 137 to 665 0.003
4 38 1006 (728) 42 777 (558) 229 − 57 to 514 0.12
5 37 1085 (771) 41 872 (595) 213 − 96 to 522 0.17
6 35 1104 (827) 38 988 (676) 116 − 235 to 468 0.51
7 34 1114 (788) 37 1087 (782) 27 − 47 to 402 0.88
Oral energy intake (kcal/day)
1 41 433 (426) 44 448 (451) 14 − 175 to 204 0.9
2 41 511 (382) 43 548 (485) 36 − 153 to 225 0.7
3 40 426 (376) 43 614 (568) 187 − 19 to 399 0.076
4 38 518 (491) 44 567 (543) 49 − 180 to 278 0.67
5 37 504 (442) 39 497 (428) −6 − 205 to 192 0.14
6 36 456 (443) 38 434 (400) − 21 − 216 to 174 0.83
7 34 322(334) 36 409 (411) 87 − 2 to 11 0.18
Total energy intake (oral and enteral nutrition) per kg bodyweight (kcal/kg/day)
1 41 23.6 (11.6) 44 16,0 (8.5) 7.6 3.19 to 12.0 < 0.001
2 41 28.2 (12.2) 43 19.7 (10.6) 8.5 3.53 to 13.4 < 0.001
3 40 27.6 (14.6) 42 22.6 (11.1) 5.0 0.74 to 10.6 0.08
4 38 28.8(16.0) 42 24.5 (10.8) 4.2 1.60 to 11.3 0.14
5 37 30.8 (15.9) 40 25.1 (11.3) 5.6 3.1 to 0.54 0.07
6 36 29.8 (16.4) 38 26.2 (13.1) 3.7 3.54 to 11.1 0.31
7 34 27.5 (15.6) 37 27.4 (14.2) − 0.45 − 7.20 to 7.0 0.90
a
DH group, Diakonhjemmet Hospital; a more assertive initial refeeding rate (20 kcal/kg/day), meeting esti-
mated energy needs within 3 days
b
NICE group, National Institute of Health and Clinical Excellence: cautious initial refeeding rate (10 kcal/
kg/day), meeting estimated energy needs within 7 days

aspiration, constipation, edema, nausea, and admission to either the medical or the surgical wards. Terminal illness
the intensive care unit. These events were evenly divided in was excluded at inclusion but despite this, six patients
the two groups. in the DH group and three in the NICE group died dur-
ing the seven-day intervention period. An additional 22
patients died before 3-months follow-up. Considering the
Discussion high mortality, vulnerability and clinical conditions of the
included patients it may have been un-realistic to expect
A more assertive initial energy refeeding protocol, increas- that reaching estimated energy needs within 3 days instead
ing to meet estimated energy needs within 3 days of com- of after 7 days would lead to improved HGS 3 months
mencement, compared to a more cautious refeeding protocol after discharge. Earlier studies in patients with anorexia
increasing to meet needs within 7 days, did not improve nervosa have shown that 8 weeks of refeeding is necessary
HGS, nor did it increase the incidence of RFS or mortality. before restoration in muscle function can be seen [39].
However, there was more respiratory distress in the group Generally, HGS is used to evaluate muscle function and in
with the more assertive refeeding protocol. older patients to evaluate for sarcopenia [33]. Sarcopenia
This is, to our knowledge the first RCT in older mal- is defined as: “loss of skeletal muscle mass and strength
nourished patients, investigating a more assertive refeed- that occurs with age” and malnutrition is considered a
ing protocol versus a cautious refeeding protocol in line risk factor for this syndrome [40]. Sarcopenia was preva-
with the NICE guidelines. All patients included in this lent in the study population at baseline since the mean
study had one or more comorbidities in addition to malnu- HGS was 13.3 kg (SD 7.3) among women and 17.2 kg
trition and they were all admitted with acute diseases, to (SD 11) among men, all well below the cut off points to

13
1208 European Geriatric Medicine (2021) 12:1201–1212

Table 4  Analyzes of differences between groups, primary and secondary outcomes

e
Outcomes n DH group n NICE group Mean difference 95%CI P value
Mean (SD) Mean (SD)
a
Handgrip strength (kg)
Day 4 29 16.7 (8.1) 29 16.8 (9.55) 0.53 − 1.63 to 3.15 0.76
Day 7 26 18.5 (7.0) 28 17.2 (10.4) 0.45 − 1.5 to 3.4 0.95
3 months 14 21.6 (8.1) 18 17.8 (10.6) 0.78 − 2.52 to 3.36 0.42
a
Weight, (kg)
Day 4 20 54.3 (18.2) 22 53.4 (15.5) − 0.56 − 4.0 to 2.9 0.75
Day 7 21 56.6 (14.9) 26 56.0 (11.9) − 0.73 − 4.1 to 2.6 0.66
3 months 16 55.9 (12.6) 19 58.2 (12.9) − 1.2 − 4.75 to 2.4 0.51
b
Phosphate < 0.65 mmol/L
RFS, n (%) 41 7 (17.1) 43 4 (9.3) 0.29
b
Magnesium < 0.65 mmol/L 41 12 (29.3) 43 13 (30.2) 0.92
n(%)
b
Potassium ≤ 3.5 mmol/L 41 17 (41.5) 43 17 (39.5) 0.79
n(%)
Mortality, 3 months, n (%) 41 16 (39.0) 44 15 (34.1) 0.64
Study start to death, days, mean (SD) 41 24.1 (26.4) 44 32.5 (27.8) 8.4 28.3 to 11.5 0.40
c
LOS, days, mean (SD) 35 14.5 (8.7) 40 13.6 (10.8) 0.94 − 5.6 to 3.7 0.68
Respiratory distress, n (%) 41 22 (53.6) 43 13 (30.2) 0.029*
d
Other adverse events, n (%) 41 7 (17.0) 42 12 (28.5) 0.12

RFS, refeeding syndrome; LOS, length of stay; CI, confidence interval

Categorical data are analyzed by Chi-square test for independence with Pearson Chi-square, and independent t test for continuous data
a
Differences between groups at day 4, day 7 and at 3 months using linear mix mixed models for repeated measurements, adjusted for the baseline
values of the outcome variable
b
Phosphate or magnesium levels < 0.65 mmol/L, and potassium ≤ 3.5 mmol/L, during the 7-day EN protocol
c
Patients who died during hospital stay were excluded, LOS was reported when patients were medical cleared and ready for discharged, regard-
less of EN protocol
d
Diarrhea, vomiting, aspiration, constipation, edema, nausea, admission to the intensive care unit
e
Mean difference refers to intervention minus control

detect sarcopenia i.e. 16 kg in women and 27 kg in men have been included as part of the definition, but in a mul-
[41]. This emphasizes the vulnerability of the included timorbid older patient group, it would have been difficult
population. to differentiate between underlying disease symptoms and
No difference in the incidence of RFS was found between RFS symptoms [50]. When including multiple criteria for
the two groups during the seven-day hospitalization. How- RFS, there is a possibility for underreporting the incidence
ever, it is challenging to compare the occurrence of RFS of the syndrome, as demonstrated by Rio et al. [43]. In their
in the present study with other studies since various cut- study, RFS was defined as severe electrolyte shifts, acute
off levels of phosphate are used. In this study, phosphate circulatory fluid overload and disturbance to organ function
levels < 0.65 mmol/L were used to identify RFS, with the 72 h after recommencing nutrition. Two percent developed
result that 17.1% in the DH group and 9.3% in the NICE RFS according to this definition, whilst 9% developed serum
group (p = 0.29) were found to have RFS during refeeding. phosphate levels < 0.5 mmol/L in the same study.
This is in line with former non-randomized studies in older More respiratory distress was found in in the DH group
multimorbid patients with reported occurrences of RFS (22 patients versus 13 in the NICE group (p = 0.029) which
between 4 and 25%. [35, 36, 42–44]. Though, these studies could be a side effect of the more assertive initial refeed-
used different cut-off levels varying between < 0.33 mmol/L ing protocol [18, 19]. The incidence of respiratory distress
and < 0.77 mmol/L [35, 36, 42–44], which reflects the need between the two groups appears statistically and clinically
to develop a standardized definition and outcomes. Phos- significant and possible etiologies include aspiration or
phate is known to be the hallmark of refeeding syndrome, greater volume load, as a result of a higher initial infusion
and as in this study, often used as the sole diagnostic sign rate. However, the collection of this data was not standard-
of RFS [18, 45–49]. Additional clinical symptoms could ized since the information was gathered retrospectively from

13
European Geriatric Medicine (2021) 12:1201–1212 1209

chart notes by nurses and medical doctors. Whilst, due to the and their experience with EN. Noticeably, and a strength,
non-standardized nature of this retrospective observation, it we did included patients at very high risk of RFS. Second,
is not possible to conclude that the more assertive refeeding the study did not recruit a sufficient number of participants
protocol was associated with a higher incidence of respira- to reach its calculated power due to difficulty recruiting in
tory distress. In future studies it is important to collect data this vulnerable group. Third, due to the high morbidity and
on respiratory distress in a standardized and systematic way mortality in the study population, adherence to the EN pro-
to be able to determine whether our finding was connected to tocol decreased during the 7 days of the intervention. This
refeeding rates or a result of other factors. Of the 35 patients resulted in a high number of missing patients and data, with
who developed respiratory distress, 19 died (54.3%) before consequences for the result of the ITT analysis. However,
3 months follow-up. The high respiratory distress mortal- mixed model analyses for repeated measurements should
ity reflects the high morbidity of the study population; with adequately adjust for baseline values, and replaces miss-
about 40% suffering from three or more comorbidities at ing data as part of the procedure [37]. Additionally, there
inclusion and nearly 37% of patients having died at 3 months were difficulties adhering to the study intervention in several
follow-up. Taking this into consideration it is possible that patients due to dislocation of the tube by accident, pause in
some of the included patients had developed cachexia, a EN for medical reasons and patients simply decided to dis-
complex metabolic syndrome observed in chronic diseases continue the feeding. Noticeably, oral intake compensated
associated with systemic inflammation, weight loss and sar- for poor EN protocol adherence, but it led to provision of
copenia [40]. Cancer, COPD, chronic heart failure, as well more calories than the EN protocols had intended to deliver
as old age, are all known to be associated with the presence (20 and 10 kcal/kg/day).
of cachexia [51]. Enteral nutrition is regarded as an ineffec- Forth, whilst the energy intake delivered by EN can be
tive treatment in cachexia, since the weight loss associated measured with precision, the additional oral intake was less
is more a consequence of metabolic disorders than a result amenable to precise measurement although every effort
of low food intake [40]. Clinical guidelines, therefore rec- for accuracy was made, this may result in an unquantified
ommend that in case of cachexia, EN should not be used potential bias. However, measuring and calculating oral
routinely with short life expectancy [52, 53]. This makes it food intake, should be considered a strength to this study.
necessary to question if EN was the appropriate treatment Fifth, again due to the severe frailty and high mortality of the
for some of the malnourished patients included in this study. study population there was a significant attrition rate during
Based on the results from the present study, the question follow-up. This is unfortunately inevitable when conduct-
of which refeeding protocol for older malnourished patients ing research in such a vulnerable population and reflects
is most appropriate is still open. Recently, two different con- the need to optimize treatment protocols for such high-risk
sensus or papers regarding RFS have been published, one populations.
by the American Society for Parenteral and Enteral Nutri- The lack of high-quality nutrition trials in a malnour-
tion (ASPEN) in 2020 [5], and one consensus document ished older patient group, and conflicting findings from other
supported algorithm by Friedli et al. [54]. The two guide- groups at risk of RFS, necessitates research into the effects
lines deviate from each other when it comes to calorie and of different refeeding protocols. This study demonstrated no
titration rates. ASPEN [5] recommend initial refeeding rate difference in HGS, RFS or mortality in malnourished older
between 10 and 20 kcal/kg/day for the first 24 h, thereafter hospital patients at 3 months follow-up, when comparing
increasing by 33% of goal calories every 1–2 days meeting two refeeding protocols. One may ask if it is possible to
estimated calorie needs between days two and four. While follow NICE guidelines’ [11] recommendation providing 5
Friedli et al. [54], recommend initial calories, and titration or 10 kcal/kg/day without denying patients food? Demon-
rates according to risk criteria, very high risk, high risk strated in this study where the cautious EN protocol started
and low risk, ranging from 5 to 25 kcal/kg/day, with a slow at 10 kcal/kg/day, but due to oral food intake summed up to
increase to meet estimated calorie needs within 4–10 days. 16 kcal/kg/day.
These two guidelines are yet to be validated.
First, and a strength, our study was designed and carried
out as an effectiveness study (second run of the Cochrane Conclusion
hierarchy of evidence) [55], to answer the question: “in
practice does a more assertive initial refeeding versus a This study did not demonstrate superiority of assertive over
cautious refeeding protocol improve muscle strength in cautious refeeding protocols for older malnourished hospital
a population at risk of RFS according to NICE screening patients, and this may be due to the study limitations. How-
criteria?” Consequently, the study was subject to the daily ever, the more assertive refeeding protocol was not associ-
routines and frustrations of hospital life, and the ability of ated with increased harm, with the caveat that respiratory
follow-up on the protocol depended on the health personnel distress might be a side effect, and should be monitored

13
1210 European Geriatric Medicine (2021) 12:1201–1212

closely in future studies. This reveals that there is an urgent setting improves outcome and quality of life in malnourished
need to develop a unified consensus of a definition, diagnos- medical patients. Clin Nutr 30(2):194–201. https://​doi.​org/​10.​
1016/j.​clnu.​2010.​07.​021
tic criteria and management of RFS, based on clinical stud- 4. Schuetz P, Fehr R, Baechli V, Geiser M, Deiss M, Gomes F
ies. Only then, true incidence rates can be found, optimal et al (2019) Individualised nutritional support in medical
refeeding protocols developed for different diseases and age inpatients at nutritional risk: a randomised clinical trial. Lan-
demographics, and more knowledge regarding who are at cet 393(10188):2312–2321. https://​d oi.​o rg/​1 0.​1 016/​s 0140-​
6736(18)​32776-4
risk and will develop RFS. 5. da Silva JSV, Seres DS, Sabino K, Adams SC, Berdahl GJ, Citty
SW et al (2020) ASPEN consensus recommendations for refeed-
Acknowledgements The authors would like to thank Dr. William J ing syndrome. Nutr Clin Pract 35(2):178–195. https://​doi.​org/​
Morton for his valuable advice and proofreading of the manuscript. 10.​1002/​ncp.​10474
6. Stanga Z, Brunner A, Leuenberger M, Grimble RF, Shenkin A,
Funding This study is a grant, funded by the DAM foundation, through Allison SP et al (2008) Nutrition in clinical practice-the refeed-
Norwegian Health Association, Grant number: FO148295-2. ing syndrome: illustrative cases and guidelines for prevention
and treatment. Eur J Clin Nutr 62(6):687–694. https://​doi.​org/​
Availability of data and material The dataset used for the current study 10.​1038/​sj.​ejcn.​16028​54
would be available upon reasonable request from the corresponding 7. Mehanna HM, Moledina J, Travis J (2008) Refeeding syndrome:
author. what it is, and how to prevent and treat it. BMJ 336(7659):1495–
1498. https://​doi.​org/​10.​1136/​bmj.​a301
8. Friedli N, Stanga Z, Sobotka L, Culkin A, Kondrup J, Laviano
Declarations A et al (2017) Revisiting the refeeding syndrome: results of
a systematic review. Nutrition 35:151–160. https://​doi.​org/​10.​
Conflict of interest The authors declared no conflict of interest. 1016/j.​nut.​2016.​05.​016
9. Marinella MA (2003) The refeeding syndrome and hypophos-
Ethical approval The study protocol was approved by The Regional phatemia. Nutr Rev 61(9):320–323
Committee for Medical and Health Research Ethics in South-Eastern 10. Kraft MD, Btaiche IF, Sacks GS (2005) Review of the refeeding
Norway (REC number 2015/676). syndrome. Nutr Clin Pract 20(6):625–633. https://​doi.​org/​10.​
1177/​01154​26505​02000​6625
Informed consent Informed consent was obtained from all patients 11. National Collaborating Centre for Acute C (2006) National
or their proxy/next of kin. If the patient was not able to give written Institute for Health and Clinical Excellence: guidance. Nutrition
consent due to cognitive decline or severe illness, a next of kin was support for adults: oral nutrition support, enteral tube feeding
asked to confirm and sign the consent. and parenteral nutrition. London: National Collaborating Centre
for Acute Care (UK)
12. National Institute for Health and Care Excellence: Clinical
Statement of authorship Authors of manuscript: Sissel Urke Olsen Guidelines (2017) Nutrition support for adults: oral nutrition
(SUO), Anne-Marie Aas (AMA), Karin Hesseberg (KH), Are Hugo support, enteral tube feeding and parenteral nutrition. London:
Pripp (AHP), Anette Hylen Ranhoff (AHR) and Asta Bye (AB). State- National Institute for Health and Care Excellence (UK) Copy-
ment of authorship: SUO, AB, AHR, AMA equally contributed to right © NICE 2019.; 2017. Copyright © NICE 2019
the conception and design of the research; SUO, AB, AHR and AMA 13. Garber AK, Cheng J, Accurso EC, Adams SH, Buckelew SM,
contributed to the design of the research; SUO, AHP and KH to the Kapphahn CJ et al (2021) Short-term outcomes of the study of
acquisition and analysis of the data; SUO, KH, AHP, and AB contrib- refeeding to optimize inpatient gains for patients with anorexia
uted to the interpretation of the data; SUO, AB, KH, AHR and AMA nervosa: a multicenter randomized clinical trial. JAMA Pedi-
drafted the manuscript. All authors critically revised the manuscript, atr 175(1):19–27. https://​doi.​org/​10.​1001/​jamap​ediat​r ics.​2020.​
agree to be fully accountable for ensuring the integrity and accuracy 3359
of the work, and read and approved the final manuscript. 14. Garber AK, Sawyer SM, Golden NH, Guarda AS, Katzman
DK, Kohn MR et al (2016) A systematic review of approaches
to refeeding in patients with anorexia nervosa. Int J Eat Disord
49(3):293–310. https://​doi.​org/​10.​1002/​eat.​22482
15. Matthews K, Hill J, Jeffrey S, Patterson S, Davis A, Ward W
References et al (2018) A higher-calorie refeeding protocol does not increase
adverse outcomes in adult patients with eating disorders. J Acad
1. O’Shea E, Trawley S, Manning E, Barrett A, Browne V, Timmons Nutr Diet 118(8):1450–1463. https://d​ oi.o​ rg/1​ 0.1​ 016/j.j​ and.2​ 018.​
S (2017) Malnutrition in hospitalised older adults: a multicentre 01.​023
observational study of prevalence, associations and outcomes. 16. Rizzo SM, Douglas JW, Lawrence JC (2019) Enteral nutrition via
J Nutr Health Aging 21(7):830–836. https://​doi.​org/​10.​1007/​ nasogastric tube for refeeding patients with anorexia nervosa: a
s12603-​016-​0831-x systematic review. Nutr Clin Pract 34(3):359–370. https://d​ oi.o​ rg/​
2. McClave SA, Martindale RG, Vanek VW, McCarthy M, Roberts 10.​1002/​ncp.​10187
P, Taylor B et al (2009) Guidelines for the provision and assess- 17. Arabi YM, Aldawood AS, Haddad SH, Al-Dorzi HM, Tamim HM,
ment of nutrition support therapy in the adult critically ill patient: Jones G et al (2015) Permissive underfeeding or standard enteral
Society of Critical Care Medicine (SCCM) and American Society feeding in critically ill adults. N Engl J Med 372(25):2398–2408.
for Parenteral and Enteral Nutrition (A.S.P.E.N.). JPEN J Parenter https://​doi.​org/​10.​1056/​NEJMo​a1502​826
Enteral Nutr 33(3):277–316. https://d​ oi.​org/​10.1​ 177/​014860​ 7109​ 18. Doig GS, Simpson F, Heighes PT, Bellomo R, Chesher D, Cater-
335234 son ID et al (2015) Restricted versus continued standard caloric
3. Starke J, Schneider H, Alteheld B, Stehle P, Meier R (2011) intake during the management of refeeding syndrome in critically
Short-term individual nutritional care as part of routine clinical ill adults: a randomised, parallel-group, multicentre, single-blind

13
European Geriatric Medicine (2021) 12:1201–1212 1211

controlled trial. Lancet Respir Med 3(12):943–952. https://​doi.​ systematic review. BMC Geriatr 17(1):238. https://​doi.​org/​10.​
org/​10.​1016/​s2213-​2600(15)​00418-x 1186/​s12877-​017-​0625-y
19. Olthof LE, Koekkoek W, van Setten C, Kars JCN, van Blokland 35. Kraaijenbrink BVC, Lambers WM, Mathus-Vliegen EMH, Siegert
D, van Zanten ARH (2018) Impact of caloric intake in critically CEH (2016) Incidence of refeeding syndrome in internal medicine
ill patients with, and without, refeeding syndrome: a retrospective patients. Neth J Med 74(3):116–121
study. Clin Nutr 37(5):1609–1617. https://​doi.​org/​10.​1016/j.​clnu.​ 36. Zeki S, Culkin A, Gabe SM, Nightingale JM (2011) Refeeding
2017.​08.​001 hypophosphataemia is more common in enteral than parenteral
20. Rice TW, Wheeler AP, Thompson BT, Steingrub J, Hite RD, Moss feeding in adult in patients. Clin Nutr 30(3):365–368. https://​doi.​
M et al (2012) Initial trophic vs full enteral feeding in patients org/​10.​1016/j.​clnu.​2010.​12.​001
with acute lung injury: the EDEN randomized trial. JAMA 37. Xi W, Pennell ML, Andridge RR, Paskett ED (2018) Comparison
307(8):795–803. https://​doi.​org/​10.​1001/​jama.​2012.​137 of intent-to-treat analysis strategies for pre-post studies with loss
21. de Vries J, Antoine JM, Burzykowski T, Chiodini A, Gibney M, to follow-up. Contemp Clin Trials Commun 11:20–29. https://d​ oi.​
Kuhnle G et al (2013) Markers for nutrition studies: review of cri- org/​10.​1016/j.​conctc.​2018.​05.​008
teria for the evaluation of markers. Eur J Nutr 52(7):1685–1699. 38. White NM, Barnett AG (2019) Analysis of multisite interven-
https://​doi.​org/​10.​1007/​s00394-​013-​0553-3 tion studies using generalized linear mixed models. Infect Control
22. Bohannon RW (2019) Grip strength: an indispensable biomarker Hosp Epidemiol 40(8):910–917. https://d​ oi.o​ rg/1​ 0.1​ 017/i​ ce.2​ 019.​
for older adults. Clin Interv Aging 14:1681–1691. https://​doi.​org/​ 114
10.​2147/​cia.​S1945​43 39. Russell DM, Prendergast PJ, Darby PL, Garfinkel PE, Whitwell
23. McNicholl T, Dubin JA, Curtis L, Mourtzakis M, Nasser R, J, Jeejeebhoy KN (1983) A comparison between muscle function
Laporte M et al (2019) Handgrip strength, but not 5-meter walk, and body composition in anorexia nervosa: the effect of refeeding.
adds value to a clinical nutrition assessment. Nutr Clin Pract Am J Clin Nutr 38(2):229–237. https://d​ oi.o​ rg/1​ 0.1​ 093/a​ jcn/3​ 8.2.​
34(3):428–435. https://​doi.​org/​10.​1002/​ncp.​10198 229
24. Norman K, Stobäus N, Gonzalez MC, Schulzke JD, Pirlich M 40. Muscaritoli M, Anker SD, Argilés J, Aversa Z, Bauer JM, Biolo G
(2011) Hand grip strength: outcome predictor and marker of nutri- et al (2010) Consensus definition of sarcopenia, cachexia and pre-
tional status. Clin Nutr 30(2):135–142. https://​doi.​org/​10.​1016/j.​ cachexia: joint document elaborated by Special Interest Groups
clnu.​2010.​09.​010 (SIG) “cachexia-anorexia in chronic wasting diseases” and “nutri-
25. Kizilarslanoglu MC, Kilic MK, Gokce D, Sakalar T, Ulger Z tion in geriatrics.” Clin Nutr 29(2):154–159. https://​doi.​org/​10.​
(2017) Is it possible using handgrip strength instead of body mass 1016/j.​clnu.​2009.​12.​004
index in MNA-SF test to assess the nutritional status of geriatric 41. Cruz-Jentoft AJ, Bahat G, Bauer J, Boirie Y, Bruyère O, Ceder-
patients? J Nutr Health Aging 21(5):579–584. https://​doi.​org/​10.​ holm T et al (2019) Sarcopenia: revised European consensus on
1007/​s12603-​016-​0744-8 definition and diagnosis. Age Ageing 48(1):16–31. https://d​ oi.o​ rg/​
26. Flood A, Chung A, Parker H, Kearns V, O’Sullivan TA (2014) 10.​1093/​ageing/​afy169
The use of hand grip strength as a predictor of nutrition status 42. Camp MA, Allon M (1990) Severe hypophosphatemia in hospital-
in hospital patients. Clin Nutr 33(1):106–114. https://​doi.​org/​10.​ ized patients. Miner Electrolyte Metab 16(6):365–368
1016/j.​clnu.​2013.​03.​003 43. Rio A, Whelan K, Goff L, Reidlinger DP, Smeeton N (2013)
27. Zhang XS, Liu YH, Zhang Y, Xu Q, Yu XM, Yang XY et al Occurrence of refeeding syndrome in adults started on artificial
(2017) Handgrip strength as a predictor of nutritional status in nutrition support: prospective cohort study. BMJ Open. https://​
Chinese elderly inpatients at hospital admission. Biomed Environ doi.​org/​10.​1136/​bmjop​en-​2012-​002173
Sci 30(11):802–810. https://​doi.​org/​10.​3967/​bes20​17.​108 44. Kagansky N, Levy S, Koren-Morag N, Berger D, Knobler H
28. Jensen GL, Cederholm T, Correia M, Gonzalez MC, Fukushima (2005) Hypophosphataemia in old patients is associated with
R, Higashiguchi T et al (2019) GLIM criteria for the diagnosis of the refeeding syndrome and reduced survival. J Intern Med
malnutrition: a consensus report from the global clinical nutrition 257(5):461–468. https://​d oi.​o rg/​1 0.​1 111/j.​1 365-​2 796.​2 005.​
community. JPEN J Parenter Enteral Nutr 43(1):32–40. https://d​ oi.​ 01457.x
org/​10.​1002/​jpen.​1440 45. Fuentes E, Yeh DD, Quraishi SA, Johnson EA, Kaafarani H, Lee
29. Wagstaff G (2011) Dietetic practice in refeeding syndrome. J Hum J et al (2017) Hypophosphatemia in enterally fed patients in the
Nutr Diet 24(5):505–515. https://​doi.​org/​10.​1111/j.​1365-​277X.​ surgical intensive care unit: common but unrelated to timing of
2011.​01196.x initiation or aggressiveness of nutrition delivery. Nutr Clin Pract
30. Schulz KF, Altman DG, Moher D (2010) CONSORT 2010 State- 32(2):252–257. https://​doi.​org/​10.​1177/​08845​33616​662988
ment: updated guidelines for reporting parallel group randomised 46. Chen LJ, Chen HL, Bair MJ, Wu CH, Lin IT, Lee YK et al (2014)
trials. BMJ 340:332. https://​doi.​org/​10.​1136/​bmj.​c332 Refeeding syndrome in Southeastern Taiwan: our experience with
31. Bauer J, Biolo G, Cederholm T, Cesari M, Cruz-Jentoft AJ, Mor- 11 cases. World J Gastroenterol 20(30):10525–10530. https://​doi.​
ley JE et al (2013) Evidence-based recommendations for optimal org/​10.​3748/​wjg.​v20.​i30.​10525
dietary protein intake in older people: a position paper from the 47. Martínez MJ, Martínez MA, Montero M, Campelo E, Castro I,
PROT-AGE Study Group. J Am Med Dir Assoc 14(8):542–559. Inaraja MT (2006) Hypophosphatemia in postoperative patients
https://​doi.​org/​10.​1016/j.​jamda.​2013.​05.​021 with total parenteral nutrition: influence of nutritional support
32. Tazmini K (2014) Elektrolyttveileder 2 utgave Diakonhjemmet teams. Nutr Hosp 21(6):657–660
Sykehus 2014; 2 utgave. http://​www.​helse​bibli​oteket.​no/​nyhet​ 48. Goyale A, Ashley SL, Taylor DR, Elnenaei MO, Alaghband-
sarkiv/​ny-​utvid​et-​utgave-​av-​elekt​rolyt​tveil​ederen Zadeh J, Sherwood RA et al (2015) Predicting refeeding
33. Roberts HC, Denison HJ, Martin HJ, Patel HP, Syddall H, Cooper hypophosphataemia: insulin growth factor 1 (IGF-1) as a diag-
C et al (2011) A review of the measurement of grip strength in nostic biochemical marker for clinical practice. Ann Clin Biochem
clinical and epidemiological studies: towards a standardised 52(Pt 1):82–87. https://​doi.​org/​10.​1177/​00045​63214​523739
approach. Age Ageing 40(4):423–429. https://​doi.​org/​10.​1093/​ 49. Marik PE, Bedigian MK (1996) Refeeding hypophosphatemia in
ageing/​afr051 critically ill patients in an intensive care unit. A prospective study.
34. Sousa-Santos AR, Amaral TF (2017) Differences in hand- Arch Surg 131(10):1043–1047. https://​doi.​org/​10.​1001/​archs​urg.​
grip strength protocols to identify sarcopenia and frailty—a 1996.​01430​22003​7007

13
1212 European Geriatric Medicine (2021) 12:1201–1212

50. Reintam Blaser A, Hubert van Zanten AR (2020). Electrolyte dis- 54. Friedli N, Stanga Z, Culkin A, Crook M, Laviano A, Sobotka L
orders during the initiation of nutrition therapy in the ICU. Curr et al (2018) Management and prevention of refeeding syndrome
Opin Clin Nutr Metab Care. https://​doi.​org/​10.​1097/​mco.​00000​ in medical inpatients: an evidence-based and consensus-supported
00000​000730 algorithm. Nutrition 47:13–20. https://d​ oi.o​ rg/1​ 0.1​ 016/j.n​ ut.2​ 017.​
51. Ali S, Garcia JM (2014) Sarcopenia, cachexia and aging: diagno- 09.​007
sis, mechanisms and therapeutic options—a mini-review. Geron- 55. Järvinen TLN, Sievänen H, Kannus P, Jokihaara J, Khan KM
tology 60(4):294–305. https://​doi.​org/​10.​1159/​00035​6760 (2011) The true cost of pharmacological disease prevention. BMJ.
52. Roeland EJ, Bohlke K, Baracos VE, Bruera E, Del Fabbro E, https://​doi.​org/​10.​1136/​bmj.​d2175
Dixon S et al (2020) Management of cancer cachexia: ASCO
guideline. J Clin Oncol 38(21):2438–2453. https://​doi.​org/​10.​ Publisher's Note Springer Nature remains neutral with regard to
1200/​jco.​20.​00611 jurisdictional claims in published maps and institutional affiliations.
53. Arends J, Bachmann P, Baracos V, Barthelemy N, Bertz H, Boz-
zetti F et al (2017) ESPEN guidelines on nutrition in cancer
patients. Clin Nutr 36(1):11–48. https://​doi.​org/​10.​1016/j.​clnu.​
2016.​07.​015

13