Article Text

Review: protein and energy supplements reduce mortality and complications in undernourished elderly people in hospital
  1. Nancy Bergstrom, RN, PhD, FAAN
  1. Center on Aging, University of Texas Health Science Center Houston
 Houston, Texas, USA

    Statistics from

 Q Do protein and energy supplements improve clinical outcomes in elderly people?


    Embedded ImageData sources:

    Cochrane Central Register of Controlled Trials (2005, issue 2); Medline (June 2005); EMBASE/Excerpta Medica, CINAHL, BIOSIS, and CAB abstracts (March 2004); HealthStar (March 2001); nutrition journals; reference lists of relevant articles; and manufacturers of nutritional supplements.

    Embedded ImageStudy selection and assessment:

    randomised and quasi-randomised controlled trials that evaluated use of an oral protein or energy supplement for ⩾1 week compared with placebo or control treatment in elderly people (average age ⩾65 y) in hospital, long term care, or the community. Randomised controlled trials (RCTs) involving patients in critical care or recovering from cancer treatment were excluded. 2 reviewers independently assessed study quality. 55 RCTs (n = 9187) met the selection criteria. The study interventions aimed to provide 175–1000 kcal or 8.5–37 g of protein daily. Intervention periods ranged from 10 days to 18 months, and duration of follow up was usually the same.

    Embedded ImageOutcomes:

    all cause mortality, complications (eg, infection, incomplete wound healing, pressure sores, and hospital admission), length of hospital stay, weight change, and arm muscle circumference change.


    Protein or energy supplements did not reduce mortality overall or in the 3 setting subgroups compared with the control condition but did reduce mortality in undernourished patients in hospital (table). Complications were reduced in patients in hospital, but not in elderly people in the other 2 settings (table). Supplements increased weight by about 2% and arm muscle circumference by about 1% in all settings. Supplements did not reduce length of hospital stay in either undernourished or nourished patients.

    Protein or energy supplements v routine care in elderly people*


    Protein or energy supplements reduce mortality and complications in undernourished elderly patients in hospital but do not provide benefits for well nourished patients or elderly people living in long term care or the community.


    Protein and energy supplementation of dietary intake is widely believed to reduce morbidity, mortality, and specific complications among older adults. Milne et al did a careful analysis of available studies, identifying methodological difficulties affecting the quality of individual studies. Only modest benefits were identified: mortality was reduced only among malnourished patients in hospital; complications were reduced only among patients in hospital; and small improvements were found for weight and arm circumference. The clinical, as opposed to statistical, significance of the results should be questioned.

    Before rejecting the use of nutritional support based on the small observed gains for a population subset, it is important to consider whether sufficient nutritional support was provided. Some participants received as little as 8.5 grams (range 8.5–37) of protein or 175 kcal (range 175–1000) for as little as 1 or 2 weeks. Even when Milne et al dichotomised dietary intake (< or >400 kcal), the amount of supplementation in the higher dose group may have been too small to make a difference in health outcomes. In fact, it would take 9.6 days to increase body weight by 0.5 kg (3850 kcal/0.5 kg weight gain / 400 kcal  =  9.6 days) if nutritional needs were otherwise being met. A recent observational study showed that nutritional support promoted healing of pressure ulcers but only when 30 kcal/kg of protein was ingested for 21 days.1 Nutritional support <30 kcal/kg did not provide significant benefits.

    A Mini Nutritional Assessment [] should be considered to identify older patients in hospital (supported by data in Milne et al) and nursing home residents (not supported by data) who are malnourished and can benefit from nutritional support. Although the meta-analysis found that overall mortality may not be improved, no data show that malnourished patients fare better than well nourished patients.


    View Abstract


    • For correspondence: MsA C Milne, Health Services Research Unit, University of Aberdeen, Aberdeen, UK. a.c.milne{at}

    • Sources of funding: Medical Research Council; Scottish Executive Health Department; Student Awards Agency for Scotland.

    Request permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.