Public Defence: Sissel Urke Olsen

Background: Malnutrition is prevalent in acute hospital settings where older adults are particularly vulnerable. It is associated with adverse outcomes, such as increased reconvalescence time, muscle wasting, and a longer hospital stay. When malnourished patients receive nutrition intervention, they may experience a detrimental outcome named the refeeding syndrome (RFS), which lacks a universal definition, diagnostic criteria, and treatment protocols. The syndrome can be characterized as severe electrolyte and fluid shifts occurring when the metabolism shifts from a catabolic state, using mainly fat and protein as an energy source, to an anabolic state, in which the metabolism regains access to carbohydrates. This shift may result in fluid and electrolyte imbalance—especially hypophosphatemia—and can cause pulmonary, cardiac, and neuromuscular complications.

Aim: The overall aim of this research project was to develop evidence-based knowledge regarding RFS among older malnourished hospitalized patients. A systematic review was conducted to obtain the current evidence available regarding RFS in older adults. Further, we explored if providing a more assertive initial refeeding protocol would improve nutrition status, as measured by hand grip strength (HGS), but that could potentially aggravate RFS. We also explored the mortality in relation to different diagnostic criteria of RFS.

Methods: This thesis consists of three papers, with different designs. The first paper is a systematic review. The outcomes of interest were incidence rate, when RFS occurs, the effect of different refeeding rates, and potential adverse events, including mortality connected to RFS. The second and third papers were a semi-double-blinded randomized trial (RCT) and a longitudinal study based on the data set of the RCT, respectively. The patients were admitted to the department of medicine or surgery because of acute illness. Inclusion criteria were patients 65 years or older and screened to be at risk for RFS. The exclusion criteria were end stage dementia, those already on artificial nutrition, not wanting to stay in the hospital for at least seven days, admitted directly to the intensive care unit, previous participation in the study, or those who were terminally ill or with a short life expectancy. Further, patients with severe kidney disease or congestive heart failure needed to be cleared by a medical doctor to participate. Patients both in the intervention and control groups received EN through a nasogastric feeding tube. The intervention group received a more assertive refeeding protocol, with an initial calorie level of 20 kcal/kg/day and titrated up to meet needs within 3 d, while the control group received a more cautious refeeding protocol, with an initial calorie level of 10 kcal/kg/day and titrated up to meet calorie needs within 7 d. The primary outcome was to evaluate if a more assertive refeeding protocol would improve HGS but potentially aggravate RFS and affect mortality. The longitudinal study evaluated incidence rates using three different diagnostic RFS criteria to examine if there was a difference in mortality in patients with and without RFS in any of the criteria.

Results: The systematic review resulted in 15 eligible papers, two cohort studies, one case control, and 12 case reports. Adverse events were noted in most of the case reports, and RFS occurred in up to 25% of the patients, even when refeeding cautiously. The quality of evidence was graded as low and moderate. In the RCT, 85 severely multimorbid patients were included with an overall age of 80 years. Patients receiving a higher initial refeeding rate and faster titration did not improve HGS at 3 months, with a mean difference between the groups of 0.78 kg. Neither did a higher initial refeeding rate aggregate a higher incidence of RFS, with 17.1% compared with 9.3% (p=0.29). Paper III demonstrated a wide range of incidence rates when using different diagnostic criteria of RFS, here between 12.9% and 65.9%, but no difference in mortality in patients with RFS compared with patients without RFS in any of the three criteria. The mortality rate was high at 3 months and 1 year at 36% and 56.5%, respectively. RFS occurred both in the systematic review and in the longitudinal study between the second and fourth days.

Conclusion: The patients in our study and who were at risk of RFS were severely malnourished, multimorbid, and with a high mortality rate. Providing more energy early during a hospital stay did not improve HGS, nor did it aggravate RFS, and mortality was not more prevalent in patients with RFS than those without. When applying different diagnostic RFS criteria, a wide range of RFS incidence rates was found. The systematic review and Paper III found RFS occurring between the second and fourth days, but electrolytes should be monitored daily during the refeeding period because RFS occur beyond the fourth day. This PhD project confirms the need of a unified global definition and diagnostic criteria to enable high-quality research and unified quality of care.