This study prospectively evaluated 35 consecutive infants who presented with fresh blood mixed with stools. The mean age at onset of bleeding was approximately 4 weeks. All infants were otherwise asymptomatic and exhibited normal growth and physical examinations. None had evidence of a bleeding diathesis, viral or bacterial enteritis, or necrotizing enterocolitis. Con sent for limited colonoscopy and biopsy was obtained for 34 infants. Twenty-five had macroscopic colitis and 10 infants had marked nodular lymphoid hyperplasia. Thirty-one infants had histopathological evidence of colitis characterized by a marked eosinophilic infiltrate. Compared with 19 controls (age, 4.1 ± 3.6 months) the mean number of eosinophils per high-power field was greater in the patients with colitis. Patients with colitis also had an elevated mean absolute peripheral eosinophil count and a low mean serum albumin compared to control values.Summary
The 31 infants with colitis were receiving solely breast milk (10), cow's milk formula (9), soya formula (9), breast milk with cow's milk formula (2), or nutramigen (1) at the time of presentation. Nineteen infants had rapid resolution of frank bleeding and gradual correction of serum albumin with dietary change. While limited colonoscopy and biopsy were useful in establishing a definitive diagnosis, a low serum albumin and high peripheral eosinophil count suggested the diagnosis. Colitis characterized his-tologically by >20 eosinophils per high-power field is a common cause of rectal bleeding in otherwise healthy young infants. Resolution of bleeding and increase in serum albumin after dietary change suggest that this is an allergic colitis: however, the pathogenesis of this disorder needs further clarification.