- Top of page
Diarrhoea commonly complicates hospital admission, resulting in increased health care costs, in addition to impacting patient’s quality of life.1–3 While it is known that certain medications cause or contribute to non-infectious diarrhoea,1, 3–5 defining mechanisms that underlie the common occurrence of diarrhoea in patients receiving enteral nutrition (EN) remains poorly explored.6–10
Frequently, the enteral formula is blamed for inducing diarrhoea. Fibre has received the most attention with 13 controlled studies reported. Meta-analysis concluded that fibre significantly reduces the incidence of diarrhoea (OR 0.68), but the types of fibre used and the dosage of fibre varied considerably. Furthermore, the positive effect was not seen in populations from an intensive care setting and was mainly observed where the incidence of diarrhoea was very high in both the fibre and no-fibre control groups.11 The strategy of offering low osmolality enteral formulas to decrease the risk of diarrhoea remains unexplored. Other factors implicated in the cause of diarrhoea in EN include the mode of delivery,12, 13 contamination of EN equipment with micro-organisms,7 colonization of bacteria and fungi along enteral feeding tubes, acquisition of Clostridium difficile (C. difficile) from postpyloric feeding14 and the artificial nature of EN itself which alters digestion and possibly absorption.12
A recently presented but untested hypothesis is that diarrhoea may be induced by the presence in enteral formula of poorly absorbed short-chain carbohydrates, which have been collectively termed FODMAPs (fermentable oligo-, di- and mono-saccharides and polyols15).16 FODMAPs are found in a wide variety of foods including lactose (in milk), fructose in excess of glucose (in mango and honey), fructans (in onion, garlic, wheat and rye), galacto-oligosaccharides (in legumes), and polyols (in stone fruit and some artificial sweeteners).17 Restricting the intake of dietary FODMAPs has been shown to improve gastrointestinal symptoms, including diarrhoea, in a majority of patients with irritable bowel syndrome (IBS)18, 19 and inflammatory bowel disease.20 A randomized, placebo-controlled rechallenge trial confirmed that this response was not a placebo effect and was due to the reduction of FODMAPs.19 The mechanism by which restriction of FODMAPs provides this benefit is through their small molecular size and osmotic activity.21, 22 FODMAPs are also rapidly fermented by bacteria and the subsequent luminal distension might lead to secondary motility disturbance and diarrhoea.23
The current study aimed to examine the clinical predictors of diarrhoea in patients receiving EN, with specific attention to the association of the composition of the enteral formulas.
- Top of page
Diarrhoea is a frequent complication of EN as shown by almost two in three patients being affected in the present series. Diarrhoea is a major nursing problem in ill patients, including those in intensive care or with mobility problems following a stroke, and may compromise other management such as maintaining fluid and electrolyte balance. There is therefore a definite need to prevent its development in patients receiving EN.
There are many risk factors for hospitalized patients developing diarrhoea. Antibiotic use is one example, and the likelihood of these risk factors influencing patients increases with increasing LOS. Thus, to determine associations of enteral formula with the development of diarrhoea, all possible influences on diarrhoea had to be taken into account. Indeed, there was a four-fold increased risk of diarrhoea when LOS was greater than 21 days and the duration of EN more than 11 days, and two-fold increased risk when antibiotics were used. No clear associations with enteral formula and mode of delivery were identified.
Given that univariate analysis identified several factors apparently associated with the development of diarrhoea, it is important to adjust for potentially confounding variables. This was particularly important in the current study, as many patients were taking medications that are well documented to cause diarrhoea2, 3 (including antibiotics and laxatives). Multivariate analysis that included all variables, as shown in Table 4, was performed. One enteral formula was identified as protective, with a greater than five-fold reduction in risk of complicating diarrhoea, independent of other variables. The only characteristic that was unique to that protective formula was its lower FODMAP content.
Most previous works on the characteristics of EN that are associated with diarrhoea have concentrated on the risk of infectious diarrhoea due either to the enteral feeding equipment and sterility of the formula,7 or to the fibre content of the formula itself.28–31 Other issues often discussed, but with minimal evidence base, are the osmolality of the formulas and mode of EN delivery (bolus vs. continuous feeding). In the current study, strict adherence to EN hygiene practice protocols was maintained and this has been shown to decrease risk of diarrhoea associated with micro-organisms acquired from external sources.7 Mode of delivery, osmolality and fibre content of the formulas showed no association with the risk of developing diarrhoea. It might have been anticipated that fibre-supplemented formulas were protective of diarrhoea, especially since a meta-analysis of interventional studies of fibre in the feed indicated a reduction of 30% in the incidence of diarrhoea in hospitalized patients.11 Reasons for the lack of association in the present study remain uncertain. Furthermore, contrary to the meta-analysis findings, the only formula that was associated with a reduced risk of diarrhoea (Isosource 1.5) had a relatively low fibre content of 7.4 g per specified daily volume to meet recommended dietary and adequate intake for micronutrients. While the required amount of fibre needed to decrease the incidence of diarrhoea has not been determined, most studies included in the meta-analysis used a daily fibre intake of at least 14 g.11
Fermentable oligo-, di- and mono-saccharides and polyols are a recently described group of short-chain carbohydrates that are poorly absorbed. In high enough doses, they are laxatives in most people; sorbitol is an example of a FODMAP that is utilized as a laxative. A role of FODMAPs in enteral formulas in the induction of diarrhoea has recently been hypothesized.16 While lactose, the FODMAP most recognized in the induction of gastrointestinal symptoms (including diarrhoea), is routinely omitted from all enteral formulas, other FODMAPs are not similarly identified. As no other data existed on FODMAP content in enteral formulas, this was examined using well-established techniques17, 26 on the common formulas used in the present study. The formula with the lowest FODMAP content, of 10.6 g per daily volume, was Isosource 1.5, which contained nearly one half or less of the other formulas. Interestingly, this formula was the only one associated with a considerable reduction in risk of developing diarrhoea. Whether this association was due to lower load of FODMAPs could not be addressed in the study design, but the lower FODMAP content was the only feature unique to this formula. Clearly, a randomized controlled trial will need to address if this association is indeed due to FODMAP content.
To date, the impact of FODMAPs has only been investigated in food. Translating this concept into an EN model is based on theory and therefore must be put into practice to be substantiated. Furthermore, the theory of reducing FODMAPs to prevent or control diarrhoea has only been investigated in the presence of IBS.19 The FODMAP load known to control IBS symptoms, ≤0.5 g per sitting,16, 32 would equate to ≤4.0 g per daily volume of enteral formula, if a sitting were a 3-h interval (to represent the average time between a meal and snack). Ideally, a low FODMAP enteral formula would be comparable to this value to be suitable for an IBS population. As no formula in the present series is ≤4.0 g FODMAPs per daily volume, Isosource 1.5 is the most reflective of a low FODMAP formula. As most patients receiving EN would probably not require as stringent a restriction as in IBS populations, the lower FODMAP content of Isosource 1.5 would still highlight its protective feature of reduced FODMAP content in preventing diarrhoea. Only excessive doses of FODMAPs, as seen in most of the formulas in this study, are expected to trigger diarrhoea in those ordinarily asymptomatic. FODMAP loading is important in predicting diarrhoea and as such, arbitrary classifications are only guides in preventing diarrhoea rather than known therapeutic measures.
There are inherent weaknesses in the design of the current study. Its retrospective nature implies dependence on the completeness and quality of documentation. A second major issue is the definition and assessment of diarrhoea, which have been a troublesome aspect for all studies examining the complications of EN. While faecal frequency could be accurately monitored by the nursing staff, faecal consistency and weight did not undergo quantitative testing. The descriptions of bowel consistency and quantity are subject to the attending nurse’s opinion and relied on regular reporting. Varying opinions will result in inconsistent descriptions of faecal consistency and weight and therefore inaccuracies in the results. Auditing diarrhoea in these patients would be facilitated by the use of a validated system of reporting such as the King’s Stool Chart.33
Another limitation of the study design is the inability to exclude infectious diarrhoea in patients that did not have microbiological investigations. Testing of faecal specimens was conducted in only one-third of patients included in the study and C. difficile was confirmed in only two patients (5% of patients tested). This incidence is lower than anticipated on comparison with previous studies investigating sources of diarrhoea34, 35 and may be as a result of the preferred treatment method of the medical team. Ideally, all patients should be investigated for infectious diarrhoea and treated accordingly. A potential opportunity for treatment may have been lost in this current study. The suggested algorithm published by Barrett et al.16 would best guide management strategies for EN-associated diarrhoea.
The present study’s definition of diarrhoea (four or more bowel actions daily and/or descriptions of ‘loose’ or ‘ooze’ bowel actions) was chosen to favour bowel frequency and consistency. This definition reflects the two most important characteristics in determining diarrhoea33 and the most commonly documented information. The potential to under- or over-estimate its presence must be considered in the interpretation of the findings.
In conclusion, retrospective evaluation of patients receiving EN showed that long LOS and duration of EN were independent risk factors for the development of diarrhoea. The only independent protective factor was initiation of EN with the formula comprising lowest FODMAP content. While theoretical considerations regarding the effect of FODMAPs on bowel function support a cause–effect relationship, this can only be addressed by a placebo-controlled interventional study investigating FODMAP content of the enteral formula and controlling for all other EN regimen characteristics.