Food refusal and failure to thrive are serious problems in young children. When accompanied by ejection of previously swallowed food, gastro-oesophageal reflux disease (GORD) is often suspected. The rationale is that reflux oesophagitis causes pain related to feedings and avoidance of feeds occurs in order to minimize the unpleasant experience (1). Such children, although hungry and eager to take the bottle at the beginning of the meal, subsequently become unsettled and irritable and push the bottle away. They often improve with anti-reflux measures such as antacids or histamine2 receptor antagonists (2).
Another potential cause of food refusal and ejection of feeds is the nausea and vomiting caused by activation of the emetic reflex. Activation of the emetic reflex is associated with autonomic prodromal symptoms such as nausea, followed by retching and vomiting caused by forceful contraction of the diaphragm and anterior abdominal wall muscles (2). In this situation, food refusal might be an indicator of antecedent nausea. This possibility was investigated in the current study by examining the relationship between food refusal and symptoms of the emetic reflex.
We used a structured interview in a prospective study of children scheduled for fundoplication to record symptoms associated with feeding and vomiting. The study included 20 neurologically impaired (age range, 3 months–8 years) and 11 neurologically normal (age range, 3 months—4 years) patients. The presence or absence of food refusal (distinct from dysphagia) was recorded in each patient. The children were divided into two groups according to whether their vomiting symptoms were typical of GORD (Group A: effortless regurgitation and/or possetting) or typical of activation of the emetic reflex (Group B: retching and forceful emesis). Groups were compared using Fisher exact test.
Pre-operatively, food refusal was more common in Group B than Group A and there was a significant association between food refusal and retching (Table 1, P < 0.01). The presence of preoperative food refusal was a strong predictor of postoperative retching in that 9 of 11 food refusers retched after fundoplication compared with 5 of 20 nonrefusers (P < 0.01).
Retching is caused by activation of the emetic reflex and usually is preceded by nausea and other autonomic prodromal features such as sweating and pallor due to peripheral vasoconstriction (2). Nausea is a very unpleasant sensation (2), perhaps a more aversive experience even than pain (3). Nausea is part of the defensive role of the emetic reflex. It facilitates the conditioning that results in subsequent avoidance of foods that previously caused illness. The role of learned food aversion is particularly apparent in animals unable to vomit, such as the rat. This conditioned food avoidance makes the rat a difficult animal to poison (2). Many chemotherapeutic agents used to treat malignancies are strongly emetogenic. Their repeated administration to children or adults in association with food may render those foods unpalatable to the patient. Such conditioning may result in specific food refusal behaviors (4,5).
Young children and the neurologically impaired are unable to verbalize their feelings and experiences. Instead, in a fashion similar to experimental animals (6), they change their behavior in response to negative experiences. They avoid what they perceive to be the stimulus for nausea by refusing feedings. They may become distressed when they see specific foods prepared and may have an aversion to the bottle. The behavior is easily misinterpreted as oppositional or lazy. Many children with chronic vomiting have features such as sweating, pallor, and retching suggesting that their symptoms are in fact caused by activation of the emetic reflex and not by the regurgitation of GORD (7). It follows that such children would not improve with medical antireflux therapy and many indeed receive antireflux surgery. Fundoplication may physically prevent vomiting, but the emetic reflex is still activated and these children may continue to retch postoperatively (7).
Food refusal may be difficult to identify in severely neurologically impaired children, and its incidence underestimated because of other factors such as dysphagia. However, it may still be apparent to careful observers. We recall the mother of an infant with spastic quadriplegia who described her child as “only happy when she is being fasted for a test.” The child's retching and vomiting improved following treatment with the anti-emetic metoclopramide.
A previous retrospective study of 600 children younger than 2 years with GORD reported that 25 had food refusal Thirteen of these had fundoplications to control vomiting (8). The prevalence of oesophagitis was no greater in the refusers than in nonrefusers, but all the food refusers experienced postoperative retching, implying that the emetic reflex was activated and suggesting that food refusal was a consequence of nausea caused by activation of the emetic reflex rather than pain from oesophagitis. This study is consistent with our findings of an association between food refusal and retching after fundoplication.
The association between food refusal and retching suggests that, in these children, food refusal in association with the ejection of gastric contents may be a consequence of activation of the emetic reflex and nausea, rather than a consequence of GORD and oesophagitis. Unpleasant experiences after food ingestion may lead to food avoidance. The response to nausea is qualitatively different from other noxious experiences in both humans and animals. It results in food/taste aversion (rather than simple avoidance), whereas heartburn (a classic symptom of GORD), bloating, and electric shock do not. Food aversion is a behavior which takes much longer to extinguish (3). A single pairing of a food with a nauseating experience has been shown sufficient to produce aversion (5,9) that may persist for years.
These observations have implications for management. These children may not be helped by antireflux medication or by antireflux surgery. Surgery may in fact make make symptoms worse because the children can no longer vomit. They may not be cured by behavioral therapy, and tube feeds may accentuate nausea and exacerbate symptoms. Investigation may show that some of them have a treatable cause for their emesis (e.g., cow's milk protein intolerance). In others, a specific cause will not be found. Some may have a form of functional dyspepsia or visceral afferent hypersensitivity. Others (especially the overtly neurologically impaired), may have a central nervous system disturbance (e.g., loss or dysfunction of central inhibitory emetic pathways). The treatment for these children should be anti-emetic strategies (7).
The authors thank Professor Spitz and Professor Milla for permission to report their patients.
1. Hyman PE. Gastroesophageal reflux: one reason why baby won't eat. J Pediatr 1994; 125:S103–9.
2. Andrews PLR. Vomiting: a gastro-intestinal tract defensive reflex. In: Andrews PLR, Widdicome JG, eds. Pathophysiology of the Gut and Airways. London: Physiological Society (Portland Press); 1993:97–113.
3. Pelchat ML, Rozin P. The special role of nausea in the acquisition of food dislikes by humans. Appetite 1982; 3:341–51.
4. Bernstein IL. Learned taste aversions in children receiving chemotherapy. Science 1978; 200:1302–3.
5. Bernstein IL, Webster MM. Learned taste aversions in humans. Physiology and Behaviour 1980; 25:363–6.
6. Hart BL. Biological basis for the behaviour in sick animals. Biobehav Rev 1988; 12:123–37.
7. Richards CA, Milla PJ, Andrews PLR, et al. Retching and vomiting in neurologically impaired children following Nissen fundoplication: predictive pre-operative factors. J Pediatr Surg 2001; 36:1401–4.
8. Dellert SF, Hyams JS, Treem WR, et al. Feeding resistance and gastroesophageal reflux in infancy. J Pediatr Gast Nutr 1993; 17:66–71.
9. Garb JL, Stunkard AJ. Taste aversion in man. Am J Psychiatry 1974; 131:1204–7.