Epidemiologic considerations

The disease burden of acute gastroenteritis

Acute gastroenteritis is often considered to be exclusively a problem of children in developing countries. These children have 5—10 episodes of gastroenteritis each

Calicivirus Under Microscope

Rotavirus Adenovirus Astrovirus Calicivirus - NLV

Calicivirus Torovirus Picobirnavirus Enterovirus 22

FIG. 1. Viral agents of gastroenteritis as seen by direct electron microscopy. Bar= 100 nm.

year during their first 5 years of life and, though many of these are mild, some episodes can lead to malnutrition, dehydration and death (Bern et al 1992). In the developing world, about one child in 40 will die of diarrhoea, making this the first or second most common cause of death in this age group.

The problem of diarrhoea is not confined to children in developing countries. In the USA, nearly every person experiences an episode of gastroenenteritis each year, but severe and fatal illness is confined to those at the extremes of age, children and the elderly (Fig. 2). Children in the United States will experience only 1.5—2.5 episodes per year, most of which are mild, but each year, an estimated 2 million will visit a doctor or clinic and 160 000 will be hospitalized (Tucker et al 1998). Among adults, about 400 000 are discharged from hospital with diarrhoea reported on their discharge certificates, and about 3000 have diarrhoea reported on their death certificates (Glass et al 2000). For hospitalizations, diarrhoea is reported on discharge diagnoses of 10—12% of all children <5 years hospitalized in the USA and in about 1.5% of hospitalizations for adults, particularly those > 65 years. For many of these hospitalizations and deaths, diarrhoea is the principal cause, and for others it remains a problem of nosocomial infection or a complication of another severe or chronic disease.

Agents of diarrhoeal illness

In 1970, the causes of diarrhoeal illness were for the most part unknown. A few agents such as salmonella, shigella, Entamoeba histolytica and Giardia lamblia were commonly sought, but rarely detected, thus, a 'diagnostic void' existed for which the aetiology was attributed to malnutrition, weaning foods, physiologic conditions, drugs or 'idiopathic' (i.e. unknown) causes. The past three decades have seen an explosion in the number of agents identified as causes of diarrhoeal diseases, including many viruses, bacteria, parasites and some toxins.

The discovery of these new agents has filled in this 'diagnostic void' and allowed us to understand the striking differences in the epidemiology of diarrhoea between







FIG. 2. The estimated burden of gastroenteritis in the USA.








OPD Visits



FIG. 2. The estimated burden of gastroenteritis in the USA.

TABLE 1 Diarrhoeal illnesses among children in the USA and Bangladesh: the role of viruses






Total episodes by age 5












'Norwalk-like viruses'



'Sapporo-like viruses'


Adenoviruses (enteric)



Bacteria and parasites

± 1


Risk of death

1 in 12 000

1 in 40

children in developed and developing countries. To begin with, children in developing countries experience many more episodes of diarrhoea each year for the first 5 years of life than do US or British children — 4—7 episodes and 1—2.5 episodes per year, respectively (Table 1) (Bern & Glass 1994). The pathogens associated with these illnesses differ markedly, reflecting different levels of sanitation and hygiene. In an industrialized country like Finland, a recent study using the best diagnostics available could identify a virus in 60% of all diarrhoeal episodes in children < 2 years, and in 85% that were moderate or severe. Of these pathogens 24% were rotaviruses, 19% caliciviruses, 4% astroviruses and 4% adenoviruses (Pang et al 2000). In developing countries, all of these viruses are also present, but bacteria and parasites transmitted by faecally contaminated food and water play a more prominent role, and no agent can be identified for the remainder of cases. The observation that infection with these gastroenteritis viruses is universal and that all children, rich or poor, are exposed and acquire antibodies in the first few years of life suggests that they may be transmitted by means unrelated to contaminated food or water, perhaps by airborne droplets or person-to-person contact.

Much less is known about the aetiology of diarrhoea that sends adults to the hospital (Fig. 3). Studies in the past have failed to identify an agent in most patients. For example, an investigation of bacterial pathogens that applied the best diagnostic procedures to screen more than 30 000 patients admitted to 10 hospitals in the USA could identify a cause for fewer than 6% of cases, leaving 94% without an aetiology in the 'diagnostic void' (Slutsker et al 1997). Similarly,

FIG. 3. The diagnostic void for gastroenteritis: aetiology of gastroenteritis among hospitalized patients (A) (Slutsker et al 1997) and outbreaks investigated by CDC, 1973-1987 (B) (Bean & Griffin 1990).

investigation of nearly 7500 outbreaks in the USA by CDC between 1973-1987 identified a viral aetiology in less than 2%. Studies underway may help to clarify the extent to which many of these are due to enteric viruses. Some groups of adults, including those hospitalized or institutionalized, immunocompromised and the elderly, are at particular risk of viral gastroenteritis with caliciviruses, but unlike the situation for children, the proportion of illness caused by viruses has not yet been well defined.

'Endemic and epidemic patterns of disease

Viral gastroenteritis occurs in two distinct epidemiologic patterns, endemic childhood diarrhoea and epidemic disease. These patterns reflect differences in the pathogens, the host's response to these infections and their modes of transmission, and all have a direct bearing on strategies for prevention and control (Table 2). Childhood diarrhoea is best exemplified by the group A rotaviruses but the pattern is similar for enteric adenoviruses, astroviruses and the Sapporo-like virus genus of the caliciviruses. Infection with these agents is universal among children during the first few years of life. First infections are

TABLE 2 Epidemiologic patterns of viruses causing acute gastroenteritis

'Endemic childhood disease

Epidemic disease


Rotavirus (group A)

Caliciviruses (NLVs & SLVs)


Rotavirus (group B)

Enteric adenoviruses


Protavirus (group C)

Pcaliciviruses (NLVs & SLVs)

Mode of transmission

Unknown; Pcontact; fomites,

Food, water, Pcontact, droplets,

droplets & aerosols, or

and aerosols






High prevalence by 5 yrs of age

Seroconversion in epidemic



Short-term (calicivirus)

Virus variation

Limited discrete serotypes

Many antigenic variants of

(except caliciviruses)


Public health control

Vaccine— RV (group A)

Outbreak control; improved


food safety and handling

symptomatic and usually protect against subsequent severe disease, and therefore disease incidence decreases with increasing age, immunity is relatively good, and the strains have a limited number of specific and distinct serotypes. Since these viruses infect all children — rich or poor—everywhere in the world, improvements in the quality of water, food and sanitation are unlikely to decrease the incidence of disease, and vaccines hold promise for the prevention of severe illness. Epidemic disease is best exemplified by the human caliciviruses, particularly the 'Norwalk-like viruses' (NLVs). These viruses are the most important cause of outbreaks of gastroenteritis in the USA and affect people of all ages from children to adults (Glass et al 2000, Fankhauser et al 1998). This broad age distribution suggests that immunity is not complete or long-lasting and may also reflect the huge genetic and antigenic diversity of strains that is not yet fully understood. Often, these outbreaks are spread by faecally contaminated food or water or person-to-person contact, and therefore the success of control and prevention measures is directly related to the ability to detect their mode of transmission.

A number of agents do not completely fit these paradigms. For the NLV genus of caliciviruses, most children in developed countries and developing countries acquire antibodies in their first years of life, suggesting that childhood infections are universal even though we do not know whether the frequency of infections or the diversity of the infecting strains are comparable (Black et al 1982). Astroviruses have been associated with outbreaks in schools and affect older children and adults, suggesting that immunity to disease may be overcome, perhaps by larger-than-normal inocula of infections spread by atypical vehicles of infection. Finally, two antigenically distinct groups of rotaviruses — group B (Chen et al 1991) and group C (Saif & Jiang 1994) — are unusual. While they are structurally very similar to the common group A rotaviruses, they have been most notably identified in epidemics, and key features of their epidemiology and interactions with their human hosts are unknown.

0 0

Post a comment