By their nature, the symptoms that arise as a result of gastroparesis can lead to reduced oral intake, including reduced macronutrient intake, weight loss and dehydration, as well as mineral and vitamin deficiency (Ogorek et al., 1991). Thus, the principle aim of dietary management is to restore and maintain nutritional status, and, in tandem, certain approaches can reduce symptoms. In diabetic patients, dietary interventions have the added aim of improving glycaemic control. In moderate to severe disease, dietary management becomes more challenging such that artificial nutrition support may be needed (Fontana & Barnett, 1996). However, there is a limited evidence base for dietetic management strategies because controlled studies are scarce; indeed, the majority of recommendations are based on clinical experience.
Because dietetic approaches to patients with gastroparesis can be complex, it is essential to have a specialist or advanced dietitian involved from the initial assessment. A detailed history of symptom frequency and timing is essential. A dietary history should be recorded with a focus on the type and consistency of foods tolerated, and the timing, content and size of meals should be recorded. Evaluation of nutritional and fluid status (including weight and anthropometric measurements), glycaemic control and the presence of any vitamin and mineral deficiency is also needed.
The main dietary components that need to be evaluated include particle size, meal size, and fat and fibre content of the meal. Alcohol and carbonated drinks are discouraged (Bujanda, 2000; Parkman et al., 2004). The overall dietary advice can be summarised as a diet of frequent, small, low-fibre and low fat meals with increased liquid nutrient intake (Parkman et al., 2004).
Because gastric emptying times for large volume meals are slower than smaller meals (Camilleri, 2006), a reduction in meal size is key in the dietary management of patients with gastroparesis. The exact meal size will need to be individualised in accordance with the patient response, although, clearly, meal numbers will need to be increased to compensate for reduced meal volume and ensure adequate nutritional intake.
The effect of dietary fat on gastric emptying and symptoms has not been evaluated directly in patients with gastroparesis, although it is established that fat delays gastric emptying in healthy volunteers (Hunt & Knox, 1968). Manipulation of the dietary fat content and meal size in patients with gastroparesis needs to be carried out with caution to ensure appropriate calorie delivery to the patient. The half emptying time of non-nutrient liquids is approximately 20 min; when plotted on a graph, gastric empting time is exponential, with a rapid initial transit into the small bowel. By contrast, liquids, with increased nutrients and energy loads, have a steadier, more linear gastric emptying curve (Camilleri, 2006). Thus, many patients may tolerate fats in liquid form, and therefore oral nutrition supplements often provide a useful way of encouraging intake of adequate calories in low volume.
An important function of the distal stomach is to regulate gastric emptying of solids by ‘grinding and sieving’ gastric food content until particles are small enough to pass the pylorus. The stomach normally grinds food to particles <2 mm in size (Meyer et al., 1981); this action is associated with the the lag phase of gastric emptying. The lag phase is followed by a linear phase, with controlled delivery of food in the small intestine (Camilleri, 2006). The importance of particle size was outlined in a recent study of seven patients with type 1 diabetes and gastroparesis and seven healthy volunteers, who were given two meals of similar composition but different particle size. Gastric emptying was measured by scintigraphy. The results obtained demonstrated that food of small particle size (minced beef, blended pasta and carrots) increased gastric emptying rate and reduced post-prandial blood glucose dip (Olausson et al., 2008). It is likely that gastric emptying was increased in the group receiving low particle size meals as a consequence of bypassing the lag phase of gastric emptying. Simple measures to reduce particle size, such as adequately chewing foods, may improve symptoms in patients with mild symptoms; in severe cases, a pureed or liquid diet may be required.
Fibre can be classified into soluble fibre (gums, gels and pectin) and insoluble fibre (bran, cellulose, hemicellulose and lignin). The impact of a modified fibre diet on gastric emptying in patients with gastroparesis has not been evaluated in a randomised study. In healthy subjects, the impact of modifying dietary fibre intake on gastric emptying can vary. The studies investigating the impact of dietary fibre are small, typically with eight to 14 patients. Methods of assessment of gastric emptying vary and most studies involve the addition of artificial fibre supplements (Holt et al., 1979; Schwartz et al., 1982, 1988; Bianchi & Capurso, 2002). Manipulation of dietary fibre intake (i.e. to reflect normal daily living) has been investigated in healthy individuals in a single study (Benini et al., 1995), which found that gastric emptying times were quick with both high- and low-fibre diets. Although gastric emptying was prolonged in the high-fibre cohort, the emptying time was only increased by an average of 3 min; furthermore, symptoms of bloating and satiety were not altered. Another study failed to demonstrate any difference in gastric emptying times when 10 healthy subjects modified their intake of soluble versus insoluble fibre (Bianchi & Capurso, 2002); however, that study was only conducted over a 24-h period. An earlier study had identified a reduction in gastric emptying with soluble fibre ingestion supplementation (Holt et al., 1979). Subsequently, in another study of 13 patients involving a 2-week ‘cross-over’ design, pectin was shown to reduce gastric emptying times, whereas cellulose (insoluble fibre) suplementation had no effect (Schwartz et al., 1982). Pectin supplementation had a similar effect in a 2-week ‘cross-over’ study in 12 diabetic pateints without established gastroparesis (Schwartz et al., 1988).
Clearly, the data available on the effect of fibre ingestion on gastric emptying are mixed and are derived from small studies; nevertheless, the ingestion of insoluble fibre has been associated with gastric bezoar formation in patients with gastroparesis. Bezoars are retained concretions of indigestible foreign material that accumulate in the stomach. Phytobezoars are composed of nondigestible food material, including cellulose, hemicellulose, lignin and fruit tannins, which are often found in raw vegetables, citrus fruits, celery, pumpkins, grapes, prunes and raisin. Avoiding foods high in insoluble fibre is advised in gastroparesis for the prevention of phytobezoars (Rider et al., 1984). Further study into the role of dietary fibre modification on symptoms and gastric emptying is required in patients with gastroparesis, although, at present, it would be reasonable for dietitians to advocate the avoidance of excessive soluble and insoluble fibre ingestion in this patient group.
Artificial nutritional support
When dietary measures and drug therapies have failed to improve symptoms and maintain nutritional status, more invasive forms of artificial nutrition support need to be considered. Slow-pump nasogastric (NG) feeding has been suggested as a treatment option (Patrick & Epstein, 2008), although, in practice, patients with severe gastroparesis are unlikely to tolerate the volume required to meet their nutritional requirement when fed directly into the stomach and, of course, gastric feeding may enhance the risk of aspiration in patients with delayed gastric emptying. Nasojejunal (NJ) feeding is more likely to be tolerated because feed bypasses the malfunctioning stomach; this modality of feeding may be considered as the route of choice in patients with gastroparesis. Indeed, the NJ rather than NG route is recommended by the National Institute for Health and Clinical Excellence (NICE) for patients with impaired gastric emptying (National Collaborating Centre for Acute Care, 2006). However, NJ tube placement can be time consuming and costly; bedside placement of NJ tubes can be attempted, although this requires peristaltic activity to guide the tube into the small bowel. There are evolving methods of bedside NJ placement; for example the Cortrak™ (Corpack MedSystems, Wheeling, IL, USA) system is a standard NJ tube with a sensor along the length, allowing visualisation on a bedside monitor (Windle et al., 2010). The ‘c’ shaped anatomy and midline position of the duodenum allows for typical images to indicate jejunal placement. When the tube crosses the mid line and then traverses back across the mid line, jejunal intubation is suggested. Tube position can be further confirmed by fluid aspiration and measuring the pH. However, in most hospitals, NJ tubes currently tend to be placed under endoscopic or radiological guidance, although these tubes are then, of course, susceptible to displacement, particularly in patients with recurrent vomiting; as a result, in our experience, oral feeding should be discouraged when patients receive NJ feeding, aiming to reduce vomiting risk and to allow an assessment of tolerance.
If patients are shown to benefit from a trial of NJ feeding and it is considered that the patient requires long-term enteral tube feeding, then percutanous placement of a jejunal tube can be considered. In general, two forms of longterm jejunal tubes can be considered: gastrostomy tubes with a jejunal extension (percutaneous endoscopic gastrostomy with jejunal extension or ‘PEG-J’ tube) or direct jejunostomy feeding tubes. PEG-J tubes may have the additional benefit of allowing the venting of stomach content and symptom relief (Kim & Nelson, 1998), although the jejunal extension can, not uncommonly, slip back into the stomach (Godbole et al., 2002). Direct jejunostomy feeding tubes can be placed endoscopically (percutaneous endoscopic jejunostomy or ‘PEJ’ tube) but a laparoscopic placement may be required (surgically-placed jejunostomy) (Fontana & Barnett, 1996). A retrospective observational study of 26 patients with severe diabetic gastroparesis, refractory to medical therapy, found that jejunostomy feeding improved nutritional and health status in 83% (Fontana & Barnett, 1996). In the same study, 14 patients required surgery as a result of complications, and minor complication occurred in 20 patients, with tube migration being a particular problem in PEG-J fed patients (Godbole et al., 2002).
Clearly enteral tube feeding should be attempted before considering parenteral nutrition (PN) because the latter is associated with significant catheter-related risks such as infection and thrombosis (Staun et al., 2009). This is in keeping with NICE guidance, which states that PN should be only be considered in a non-accessible or nonfunctioning gut and when there is failure of gut function to a degree that definitely prevents the adequate gastrointestinal absorption of nutrients (National Collaborating Centre for Acute Care, 2006). PN may take the form of temporary supplemental parenteral nutrition support or, in refractory cases, long-term PN.