Winter vomiting disease, 1929-2006

January 2006

John Zahorsky, who had previously described roseola infantum, addressed a meeting of the St. Louis Pediatric Society on April 12, 1929, at which time he described “hyperemesis hiemis or the winter vomiting disease.” In his presentation, he distinguished this new entity from “dyspepsia from a parenteral infection,” which he stated was “much overrated.”

He declared that epidemics of winter vomiting disease were quite circumscribed, and that he had observed “this disease for many winters in the last 30 years.” The illness was described as occurring mainly in infants and characterized by rather sudden onset with “peevishness and restlessness during the night before the vomiting begins.” Zahorsky added, “Within a few hours after the vomiting begins, a looseness of the bowels is observed” (Arch Pediatr. 1929;46:391-395).

There were many unsuccessful attempts to propagate the agent responsible for winter vomiting disease until material from infected infants, some of which had been filtered to eliminate the possibility that bacteria had caused the illness, were inoculated into young calves and produced diarrhea. Jack S. Light, a pediatrician, whom I met during my tenure at Cedars-Sinai in Los Angeles, gave me a reprint of the paper, which he had co-authored with Horace L. Hodes. The original work had been done while they were in Baltimore (J Exp Med. 1949;90:113-135). Hodes, who was later to become the first chair of pediatrics at Mount Sinai Hospital in New York, had preserved some of the material in a freezer for almost half a century. He sent this to the NIH, where rotavirus, now recognized as the causative agent of winter vomiting disease, was identified by electron microscopy.

In 1973, Ruth Bishop and colleagues in Australia (Lancet. 1973;1:1281-1283) studied intestinal biopsies of nine babies with diarrhea. They found viral particles in the villous cells and demonstrated a loss of disaccharidase activity in six. The particles were not present in convalescent biopsies in three of the infants. Although she believed the virus to be an orbivirus, it was later designated rotavirus, which is called this because of its wheel-like appearance by electron microscopy. Orbiviruses and rotaviruses are related viruses and classified in the family of reoviruses, which are double-stranded RNA viruses.

This seminal discovery opened the door to investigations of the pathophysiology and epidemiology of the disease and to the development of diagnostic tests and vaccines.

The studies of the clinical course and epidemiology of winter vomiting disease that followed substantially confirmed what Zahorsky had described almost a half century earlier. Rotavirus infection occurred mainly in the winter in the Northeast, affected infants mainly between the ages of 6 months and 12 months (median 10 months) and caused vomiting followed shortly by the onset of diarrhea both for an average of about two and a half days. , (In the Southwest, the illness occurs earlier than in the Northeast.) Fever generally was low grade. Dehydration was evident clinically and by slightly elevated blood urea nitrogen and urine specific gravity. The white blood cell count was generally unremarkable. Deaths in both reports were rare.

In the later studies, rotavirus was found to spread readily to contacts, which usually resulted in subclinical infection in adults. Moreover, it was apparent that antibodies did not offer complete protection against infection. Although reinfection did occur, it usually was milder than the first episode (New Engl J Med. 1976;294:965-972)(J Pediatr. 1977;91:188-193). At this time, the treatment described is withholding fluids for 24 hours and administration of IV fluids.

Emmett Holt, who had recently retired as the chair at Bellevue at that time, shared with us that shortly after World War II, babies in Europe with diarrhea were reported to recover as rapidly if fed from the onset of diarrhea as if the food was withheld. This experience was ignored during the era when “resting an injured part,” such as a sprained ankle or an infected gastrointestinal tract, was in vogue. “Resting the bowel” was the standard of care in that era. The use of for the treatment of cholera changed our approach to the management of diarrhea. Diarrhea had claimed millions of lives in developing countries, where it was impossible to hospitalize all of these patients for IV therapy. It was found that a properly constituted fluid given orally to these patients could successfully treat severe diarrhea. The original fluids contained a relatively high solute load as they were used to treat cholera, which produced a secretory diarrhea. Although the mechanism by which rotavirus causes diarrhea is complex (Microbes Infect. 2001;3:1145-1156), the resulting diarrhea is more hypotonic than that produced by cholera. Thus, the electrolyte composition of mixtures for enteral treatment of this diarrhea in the United States was adjusted accordingly.

Oral dehydration therapy of winter vomiting disease is encouraged, but it has not yet caught on in the United States. It avoids hospital admissions, is less expensive and is at least as effective as IV therapy for all but severely dehydrated patients. Physicians can still give oral dehydration therapy to infants who are vomiting, provided that initially it is given in teaspoon amounts. These infants are quite thirsty and will accept these preparations, although their flavor may leave something to be desired.

Glucose electrolyte fluids also can be frozen and offered as an ice pop. Babies who are being breast-fed should continue to be offered the breast. Indeed, the older practice of gradually introducing a normal diet has largely been abandoned. The AAP report states that children should receive age-appropriate diets as soon as they are rehydrated, and those who are not dehydrated should continue to receive their regular diet.

“If children are monitored to identify the few in whom signs of malabsorption develop, a regular age-appropriate diet, including full-strength mild can be used safely,” wrote researchers from an article published in Pediatrics (1996;97:424-435). Foods, which are well tolerated, include complex carbohydrates, lean meats, yogurt, fruits and vegetables. The traditional Bananas, Rice, Applesauce and Toast (BRAT) diet is felt to be “limited in energy density and fat.” There is a sense that “intractable diarrhea,” the continued downhill course of babies with diarrhea, may be decreasing with the liberalization of normal feeding in infants with diarrhea. Early feeding may reduce the volume of stools and duration of diarrhea (MMWR. 2003;52:RR-16).

We now are entering an era where prevention is possible. The first of the two rotavirus vaccines to be introduced has been recommended for licensure to the FDA. They already are being used in some foreign countries. In the interim, we must recognize that this virus is spread readily, and if children are admitted to the hospital, appropriate isolation procedures must be carried out. Hand washing, in hospitals, out of home care and other places where infants are together, should be emphasized. Alcohol-based washes have been shown to reduce transmission (MMWR. 2002;51:RR-16).

The major challenge that confronts us is making the new rotavirus vaccines available to children in the developing world, where rotavirus diarrhea is said to account for over 400,000 deaths annually (Emerg Inf Dis. 2003;9:565-572).