Acta Paediatr Suppl 381: 39-44. 1992

Persistent diarrhea in Northeast Brazil: etiologies and interactions with malnutrition Aldo AM Lima, Guodong Fang, John B Schorling, Licio de Albuquerque, Jay F McAuliffe, Sulivan Mota, Roberio Leite and Richard L Guerrant Division of Geographic Medicine. Department of Medicine, University of Virginia School of Medicine, Charlottesville. Virginia, U S A ; Clinical Research Unit. Federal University of Ceara. Fortaleza. Cearu, Brazil

Lima AAM, Fang G , Schorling JB, de Albuquerque L, McAuliffe JF, Mota S, Leite R, Guerrant RL. Persistent diarrhea in Northeast Brazil: etiologies and interactions with malnutrition. Acta Paediatr 1992;(suppl 381):39-44. Stockholm. ISSN 0803-5326 With the improved control of acute diarrheal illness mortality with oral rehydration therapy, persistent diarrhea is now emerging as a major cause of childhood mortality in tropical developing areas like the impoverished populations in Brazil’s Northeast. “Graveyard surveillance” in the rural community of Guaiuba in northeastern Brazil revealed fully half of the 70% diarrhea mortality was due to persistent diarrheal illnesses. Furthermore, 1 1 ‘Yo of 14 or more diarrheal illnesses per child per year in an urban slum in Fortaleza persisted beyond 14 days, a definition that clearly identified the high risk children for heavy diarrhea burdens. Not only did heavy diarrhea burdens ablate the key “catch-up” growth seen in severely malnourished children and in children following previous diarrheal illnesses, but malnutrition significantly predisposed children to a greater incidence and duration of diarrhea as well as a greater incidence of persistent diarrhea. Etiologic studies of 37 children presenting with persistent diarrhea to Hospital das Clinicas in Fortalezd revealed that Cryptosporidium (in 13%) and enteroadherent E. coli (36% with aggregative, 29% with diffuse and 13% with localized adherence to HEp-2 cells) were the predominant potential pathogens found in the stool or upper small bowel. These findings suggest that persistent diarrhea is emerging as an important health problem in Brazil’s Northeast, that it identifies a high risk child for heavy diarrhea burdens, that important interactions occur with malnutrition and that Cryptosporidiurn and enteroadherent E. coli warrant further study as potential etiologies of this major cause of morbidity and mortality.

RL Guerrant. Division of Geographic Medicine, Department o j Medicine, University of Virginia School of Medicine, Charlottesville, Virginia, USA

With the advent of effective oral rehydration therapy for acute diarrheal illnesses, persistent diarrhea is now emerging as a major cause of morbidity and mortality in many developing areas. Nowhere is this more apparent than among the rural and urban impoverished populations around Fortaleza, a city of over two million people that is the capital of the state of Ceara in the heart of Brazil’s Northeast region. Regarding the impact of persistent diarrhea on mortality, “graveyard surveillance” with verbal autopsies of all childhood deaths was conducted by JFM each week over the 12-month period of August, 1983 through July, 1984 in the rural community of Guaiuba (population c. 8000) about 35 km outside Fortaleza. As noted in Table I , 70% of all deaths had been due to diarrhea, and half of the diarrhea deaths were due to illnesses that had persisted longer than two weeks. Our recent focus has been on the morbidity of persistent and other diarrheal illnesses in the urban slum area (favela), Gonqalves Dias, where illness rates exceed 14-1 5 illnesses per child-year in the second year of life, of which 1 1 % persisted beyond 14 days (Fig. 1) (1). This

contrasts with only 3% of 8-9 illnesses per child-year that had persisted according to similar surveillance methods in the rural community of Pacatuba, about 32 km from Fortaleza (2, 3). Furthermore, in the setting of heavy diarrhea burdens in the urban favela, Gonqalves Dias, diarrheal illnesses significantly blunt catch-up growth in two ways: ( I ) First, recurrent diarrhea in the two-month period following previous diarrhea (defined as > 30% of days spent with diarrhea) blunts the catchup growth normally seen by 5516% for weight and height gains respectively (Table 2), an effect that persisted when controlled for the confounding variables of age and prior nutritional status (4). Second, as shown in Fig. 2, while it has no effect on normally nourished children, the increasing diarrhea burdens progressively ablate the catch-up growth otherwise seen in severely malnourished children (defined as weight for age Z scores < 3) (4). In addition, malnutrition, defined as above, predisposed to a 100% increase (doubling) in the days a child spends with diarrhea. This is due to both an increase in rates as well as durations of illness (5) (Table 3).

40

A A M Lima et a ~ .

ACTA PAZDIATR SUPPL 381 (1992)

Table 1. Child mortality in Guaiuba, Ceara. Ages (years)

Acute diarrhea Prolonged diarrhea Other causes Total

la)

2 L + ? t d

o ! . , . 0.0

Furthermore, malnutrition, defined as either weightfor-age 16% of their days spent with diarrhea were identified by at least one persistent diarrheal illness over the 28month period of study. Conversely, any child who presents with a persistent illness beyond 14 days in this setting is clearly at a high risk for having a heavy diarrhea burden (defined as > 16% of days spent with diarrhea). Despite its importance, however, the pathogenesis and etiologies of persistent diarrhea remain poorly understood. Therefore we have been conducting studies of potential causes of persistent diarrhea in both clinic and hospitalized children in Fortaleza. We report here the results of studies of stool or small bowel aspirates or both in 37 children presenting with persistent diarrhea to the clinic at Hospital das Clinicas in Fortaleza.

1.0

1

.

1

.

1

.

2.0 3.0 4.0 AGE (YEARS)

1

5.0

Fig. I . Age-specific attack rates for diarrhea in Gonqalves Dias over 2 years ( I ) .

Table 2 . Impaired weight and height gains by recurrent diarrhea (defined as >30'!h of days spent with diarrhea over a second, sequential 2-month period).

N o recurrent diarrhea ( n = 393. 2-month periods) Recurrent diarrhea ( n = 134. 2-month periods)

Height gain (cm)

Weight gain (kg)

I .56

0.38

1.48*

0.32*

* p 45

96 of Days wlth Diarrhea

Fig. 2. Diarrhea impairs catch-up growth in severely malnourished children. -iZmedian.

Table 3. Malnutrition predisposes to increased incidence as well as duration of diarrhea among Gonqalves Dias children < 5 yo.

Wt. for age Z-score > -3 < - 3 o/. increase p value* Episodes/2 mos. Mean duration Total days of diarrhea

I .9 6.7 d 12.1

2.6 11.6d 24.2

37 13 100

0.001**

0.004 3,

-+-

weight for age < 3 SDs below the

NY; with 1YOD-mannose and 0.47 ml 7.5% NaHC03) and incubated at 37°C in 5% COZ for 3 h; cells were then washed with PBS, fixed with methanol and stained with Giemsa stain. HEp-2 cell adherence were characterized as local (LA), autoaggregative (AA) and diffuse (DA) traits. E. coli were also tested for enterotoxins and adhesins using DNA gene probe hybridization, done at the University of Virginia for LT, STa (NEN Research Products, Boston, MA). Bacterial colonies were transferred to Whatman 541 paper filters, treated with NaOH and heat to rupture the bacteria, render the DNA singlestranded and fix it to the solid substrate. The filters were then hybridized under stringent conditions (50% formamide; washed at 65°C) with the probes labeled by nick translation with biotinylated nucleotides using the BRL kit (Gaithersburg, MD).

PD/child-year

1%

Results Clinical data

* Numbers in parentheses are percentages cell adherence as follows: HEp-2 cells were grown in 4-well LAB-TEK slides (Nunc Inc., Naperville, IL) overnight, washed x 4 with HBSS (Hanks’ Balanced Salt Solution, Gibco Lab Inc.). Test and control strains form overnight growth on sheep blood agar plates were inoculated to vials with 1 YOtryptone/D-mannose broth, incubated shaking at 37°C overnight; 20 p1 of each broth culture was added per well with 1 ml of MEM medium (without glutamine, Gibco Lab. Tec. Inc. Grand Island,

Of 26 of these children with nutritional status determinations, 7(27%) had moderate or severe malnutrition as determined by weight-for-age more than two Z scores below the NCHS mean (3, 12% had weight-for-age more than three Z scores below the NCHS mean). Six (23%) had weight-for-age below the 75th percentile. Finally, despite their young age, none of these children were exclusively breast-fed; only 7 continued breast feeding with family food, and 2 1 had been fully weaned after a median of only one month of age (mean=2.86 months). Only two of these children had been breast-fed for longer than six months. Etiologic agents

In contrast to the usual predominance of enterotoxige-

42 AAM Lima el al.

ACTA P E D I A T R SUPPL 381 (1992)

NUMBER

of CHILDREN

2 4 6 8 1

0

1

2

1

4

1

6

1

8

~

P

2

4

~

~

~

~

%of DAYS SPENT WITH DIARRHEA Fig.3. Persistent diarrhea identifies the high-risk child. Distribution of 44 children i5 years old followed over 28 months in Gonqalves Dias by percentage of their days spent with diarrhea. W No illness lasting > 14 d. 0 One or more illnesses lasting > 14 d .

Table 5. E. coli adherence traits in fecal and small bowel studies of children with persistent diarrhea. Stool ( n = 19)

Small bowel (n= 28)( 16[57'X,]had > 10' GNR/ml, 12 had strains tested)*

Either (n=31)

(%)

(%)

(X,)

HEp-2 adherence Local (LA) Aggregative (AA) Diffuse (DA)

",

L , 25 (3/12) 25 (3/12) ,L{

M R H A E. coli (Human A RBC) Hydrophobic E. coli

35.5 (11/31) 29 (9f/31)

8.3 (1**/12) 8.3 (1/12)

9.7 (3/31) 6.5 (2131)

* Three of these ( I 1% of total) had > lo4 coliform/ml. t One patient had strains from small bowel exhibiting LA and strains from stool exhibiting DA to HEp-2 cells. T One oatient had strains from stool exhibitine. both AA and D A to HEo-2 cells. _. . . . . . . . . . . - . . .- . . . 9; 1 hese three patients. strains also exhibited UA to Hkp-2 cells. YThese two patients' strains also exhibited DA ( I ) or AA (1) to HEp-2 cells. ** This patient's strain was also hydrophobic and exhibited D A to HEp-2 cells.

-

~

nic E. coli and rotaviral etiologies seen with acute diarrhea in this setting, the leading potential pathogens seen with persistent diarrhea were Cryptosporidium and enteroadherent, LT-, ST-coliform organisms. The leading pathogen (other than enteroadherent E. coli), Cryptosporidium, was seen in 13% of children with stools . . . .. ... ,-,., . . examinea ror persistent aiarrnea, witn 6./ y o naving Giardia lamblia, 3.7% with rotavirus and none with enterotoxigenic E. coli, Shigella, Salmonella, or C .jejuni. Shown in Table 5 are the results of studies of HEp-2 cell adherence traits, mannose-resistant hemagglutination of human type A red blood cells (MRHA, for colonization factor antigens I, I1 or IV) (7) and hydrophobicity testing of E. coli isolated from the stool or small bowel of these children with persistent diarrhea. .

C

I

An impressive 53% of 19 children with fecal E. coli tested had organisms that exhibited aggregative adherence to HEp-2 cells (Fig. 4); 37% had diffusely adhereent E. coli (one patient had both DA and AA strains); and 16% had locally adherent organisms. In addition, 57% (16) of 28 children with small bowel aspirates had > lo2 .. .... aerobic crram negative coiirorm organisms per mi in me aspirate. Of these, 12 had strains tested for adherence traits, including 3 (25%) with aggregative adherence (AA), 3 (25%) with diffuse adherence (DA), and 2 (1 7%) with localized adherence (LA) to the HEp-2 cells. Altogether, studies of stool or small bowel aspirates showed 35.5% with AA, 29% with DA and 12.9% with LA (one patient had both DA and AA strains isolated from stool, another patient had an LA strain from the

-

..C

~

3

~

ACTA PiEDlATR SUPPL 381 (1992)

Persistent diarrhea and malnutrition in Northeast Brazil

43

Fig. 4. Patterns of E. coliadherence, showing localized (a), diffuse (b) and aggregative (c) adherence distribution on HEp-2 cells in tissue culture.

study [ l]), and that a significant 5 1 % of the children with persistent diarrhea had aggregative adherent E. coli in their stool (comparable to 53% of children with fecal isolates tested in the present study, and 35.5% overall) ( 1 1, 12). Of interest, Cravioto also noted an association of aggregative adherent E. coli with bloody diarrhea in children with acute diarrhea. Whether the acute bloody diarrhea or the persistent diarrhea relates to bloody diarrhea in ligated rabbit intestinal loops or to the invasive ability reported with aggregative adherent E. coli reported by Vial et al. ( I 3) and by Baudry et al. (14) or to a new heat stable toxin (15) remains to be determined. Analogous to the experience of Wanke et al. ( 1 6) and Schlager et al. (1 7), and to five of nine AA strains tested from seven hospitalized patients (1 8) the capacity of these AA small bowel organisms to cause diarrhea was documented in the RITARD model. However, the mechanism by which EAgg E. coli cause diarrhea remains unclear, and may involve a newly described 2-5 kDa heat stable, plasmid encoded toxin (EAST) that is animal loop negative but increases short Discussion circuit current in bowel mucosa mounted in Ussing The predominant potential pathogen found in these chambers (19). Of interest, like Fang’s AA small bowel children (36%) was E. coli with the ability to show isolates from hospitalized children, the AA E. coli aggregative adherence (EAggEC) to HEp-2 cells in isolated from the small bowel of the patient in this study tissue culture, comparable to 20% (8/40) of cases of (and RITARD positive) were negative by gene probe persistent diarrhea and a significantly, greater number (14). than in controls (2/38,5%) or acute diarrhea cases (4/50, Also like Cravioto, we found that 29% of our patients 18%) in a previous study by Wanke et al. in this same had diffusely adherent E. coli, which, taken with population (8). While Wanke found no LA E. coli in her Wanke’s data from a previous study in this population is controls, she did find 18% (7/38) of controls with DA E. not significantly different from controls (8). This is in coli. A similar association of aggregative adherent E. contrast to Giron et al., who found diffusely adherent E. coli with persistent diarrhea has been described by Bhan coli in 58% of 24 cases of diarrhea without other et al. in children in India (9, lo). In a two-year, pathogens studied over a three-week period during the prospective cohort study in a rural village in Mexico, peak diarrhea season in a remote southern Mexican Cravioto found that 9% of all diarrheal illnesses Mayan village (20). Finally, Fang et al. have noted a persisted beyond 14 days (comparable to 1 1 % in our similar association of aggregative adherent E. coli and

small bowel and a DA strain from stool). Only 5 (16%) of patients’ strains from both small bowel and stool showed no HEp-2 cell adherence. Controls from an earlier study by Wanke et al. in this same population showed only 5% (2 of 38) with AA E. coli(and 8%, 4/50 with acute diarrhea having AA E. coli) (8). The rates of AA E. coliin this study (10 of 19 tested from stool, 53%, or 11 of 31 overall, 36%) were significantly greater than either the control (2/38, 5 % ) or acute diarrhea (4/50, 8%) group in Wanke’s study in this same population ( p < 0.01). In contrast to AA E. coli, Wanke et al. found 18% (7/38) ofcontrols in this population had DA E. coli (8). Finally, we tested two AA E. coli (that had been isolated at 8 x 104/mlfrom the upper small bowel of one patient) in the reversible ileal tie model in rabbits and found that both colonized at 108-109 and caused diarrhea (that was persistent until sacrifice at 72 h in one animal and fatal at 27 h in the other).

44

AAM Lima et al.

Cryptosporidiurn with persistent diarrhea in children hospitalized in Fortaleza as well (1 8). In conclusion, we note that persistent diarrhea is emerging as a major cause of mortality and morbidity in Brazil’s Northeast. It appears to predispose to as well as result from malnutrition in this setting, and it clearly identifies a high risk child for a heavy burden of diarrhea. Finally, enteroadherent E. coli, especially those with an aggregative pattern of adherence to HEp2 cells, and Cryptosporidiurn may be associated with approximately half of these cases and clearly warrant further investigation. Acknowledgements.-This work was supported in part by grants from the World Health Organization and by the Rockefeller Foundation.

References I . Schorling JB, Wanke CA, Schorling SK, McAuliffe J F , de Souza MA, Guerrant RL. A prospective study of persistent diarrhea among children in an urban Brazilian slum: patterns of occurrence and etiologic agents. Am J Epidemiol 1990;132:144-56 2. McAuliffe JF, Shields DS, de Souza MA, Sakell J, Schorling J, Guerrant RL. Prolonged and recurring diarrhea in the Northeast of Brazil: examination of cases from a community-based study. J Pediatr Gastroenterol Nutr 1986;5:902-6 3. Guerrant RL, Kirchhoff LV, Shields DS, et al. Prospective study of diarrheal illnesses in northeastern Brazil: patterns of disease, nutritional impact, etiologies, and risk factors. J Infect Dis 1983;148:986-97 4. Schorling JB, Guerrant RL. Diarrhea and catch-up growth. Lancet 1990;335:599-600 5. Schorling JB, McAuliffe J F , de Souza MA, Guerrant RL. Malnutrition is associated with increased diarrhoea incidence and duration among children in an urban Brazilian slum. Int J Epidemiol 1990;19:728-35 6. Schorling JB, de Souza MA, Guerrant RL. Identification of the high risk child. In: Guerrant RL, de Souza MA, Nations MK, eds. At the edge of development: health crises in a transitional society. New York: Oxford University Press, 1991 7. Guerrant RL, Lingwood CA. Interactions of microbial adhesins and toxins with the gastrointestinal mucosa. Glycoconjugate receptors for adhesins and toxins. In: Marshall BJ, McCallum RW, Guerrant RL, eds. Helicobacter pylori, peptic ulceration and gastritis. Boston: Blackwell, 1991: 66-80

ACTA PADIATR SUPPL 381 (1992)

8. Wanke CA, Schorling JB, Barrett LJ, de Souza MA, Guerrant RL. Adherence traits of Escherichiu coli,alone and in association with other stool pathogens: potential role in pathogenesis of persistent diarrhea in an Urban Brazilian slum. Pediatric J Infect Dis 1991;10:746-51 9. Bhan M K , Khoshoo V, Sommerfelt H, Pushker R, Sazawal S, Srivastava R. Enteroaggregative Escherichiu coli and Salmonellu associated with nondysenteric persistent diarrhea. Pediatr Infect Dis J 1989;8:499-502 10. Bhan M K , Raj P, Levine MM, et al. Enteroaggregative Escherichiu coli associated with persistent diarrhea in a cohort of rural children in India. J Infect Dis 1989;159:1061-4 1 1, Cravioto A, Reyes RE, Trujillo F, et al. Risk of diarrhea during the first year of life associated with initial and subsequent colonization by specific enteropathogens. Am J Epidemiol 1990; I3 1386-904 12. Cravioto A, Tello A, Navarro A, et al. Association of Escherichiu coli HEp-2 adherence patterns with type and duration of diarrhoea. Lancet 1991;337:262-4 13. Vial P, Robins-Browne R, Lior H, et al. Characterization of enteroadherent-aggregative Escherichiu coli, a putative agent of diarrheal disease. J Infect Dis 1988;158:70-9 14. Baudry B, Savarino SJ, Vial P, Kaper JB, Levine M M . A sensitive and specific DNA probe to identify enteroaggregative Escherichiu coli, a recently discovered diarrheal pathogen. J Infect Dis 1990;161:1249-5 I 15. Savarino SJ, Fasano A, Robertson D, Levine M M . Characterization of a putative enterotoxin elaborated by enteroaggregative Escherichiu coli. In: Abstracts 90th Annual Meeting, Washington, DC. Am SOCMicrobiol 1990;45(abstr BI 12) 16. Wanke CA, Cronan S, Goss C, Chadee K, Guerrant RL. Characterization of binding of Escherichiu coli strains which are enteropathogens to small-bowel mucus. Infect Immun 1990; 58:794-800 17. Schlager TA. Wanke CA. Guerrant RL. Net fluid secretion and impaired villous function induced by colonization of the small intestine by nontoxigenic colonizing Escherichiu coli. Infect lmmun 1990;58:1337-43 18. Fang G D , Lima AM, Wanke CA. Kaper JB, Levine MM, Guerrant RL. HEp2 cell-adherent E. coli: potential causes of persistent diarrhea of multiple genotypes and different adherence phenotypes. Clin Res 1991;39:223A 19. Savarino SJ, Fasano A, Robertson DC, Levine MM. Enteroaggregative Escherichiu coli elaborate a heat-stable enterotoxin demonstrable in an in vitro rabbit intestinal model. J Clin Invest 1991;87:1450-55 20. Giron JA, Fry J, Frankel G, et al. Diffuse-adhering Escherichiu coli (DAEC) as a putative cause of diarrhea in Mayan children in Mexico. J Infect Dis 1991;163:507-13