The epidemiology of ETEC diarrhea and association of diarrhea and malnutrition in a cohort of young Egyptian children
[Thesis]
T. F. Wierzba
K. Nelson
The Johns Hopkins University
1998
173
Ph.D.
The Johns Hopkins University
1998
We studied the distribution, pathogenicity and virulence of enterotoxigenic E. coli diarrhea and the association between diarrhea and malnutrition in a cohort of children aged less than three years from a periurban area of Egypt. Home visits were made to each household twice weekly from Nov. 1993 to Sept. 1995. Clinical data and rectal swabs were obtained from each child with a loose stool. Fecal specimens were collected once a month from non-diarrheal participants. Anthropometric measurements were made at three month intervals. E. coli colonies were tested for heat labile (LT) and heat stable (ST) toxin and colonization factor antigens (CFAs). A diarrhea episode was defined as three or more loose or watery stools or one bloody stool in a 24 hour period. Among 242 children, diarrhea incidence was 2.9 episodes per year (epy), while ETEC was 0.6 epy. Children <12, 12-23 and 24-35 months had an ETEC incidence of 1.0, 0.6 and 0.1 epy, respectively. Twenty-three percent of ETEC expressed a known CFA. ST-ETEC incidence was 2.5 times more common in the warmer than cooler months, while LT-ETEC showed no seasonality. ETEC incidence increased when a garbage container was present in the house (RR = 1.5) and in crowded households. The presence of a sanitary latrine was protective (RR = 0.5). ST-ETEC, but not LT-ETEC, were more frequently isolated from cases than controls for children less than two years old. Twenty-four percent of cases reported vomiting and physicians reported dehydration in 16% of cases. Among 143 children included in the nutrition study, 358 diarrheal episodes were reported, 1% of which lasted usd\geusd14 days. Stunting, wasting and low weight-for-age were documented in 19%, 3% and 7%, respectively. An association was detected between greater than or equal to two diarrhea episodes and subsequent changes in weight-for-age (0.24 Z-score) and height for age (0.28 Z-scores) occurring over approximately three month intervals. This association did not hold, however, when analyzed over six month intervals if no diarrhea was reported in either the first or second half of this interval. When testing whether malnutrition predisposes to diarrhea, weight-for-age usd{<}\ {-}usd2 Z-scores among the poorest children was associated with diarrhea (RR = 1.8). Diarrhea itself was also associated with a subsequent attack (RR = 1.9). In this nominally well off population, diarrhea is moderately high with ETEC representing 20% of all episodes. The results suggest that ETEC epidemiology differs by toxin; ST-ETEC is more pathogenic and more frequent during warm months than LT-ETEC. Improved sanitation could reduce ETEC incidence. Diarrhea does not appear to substantially contribute to malnutrition when these children had diarrhea free time for catch up growth. Low weight-for-age among the poor and diarrhea itself was associated with subsequent risk of diarrhea.