NIH Public Access

Author Manuscript
Gastroenterology. Author manuscript; available in PMC 2011 October 1.
Published in final edited form as: Gastroenterology. 2010 October ; 139(4): 11561164. doi:10.1053/j.gastro.2010.05.076.

NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript

Prolonged episodes of acute diarrhea reduce growth and increase risk of persistent diarrhea in children
Sean R. Moore1,2,*, Nolia L. Lima3, Alberto M. Soares3, Reinaldo B. Ori3, Relana C. Pinkerton4, Leah J. Barrett4, Richard L. Guerrant4, and Aldo A.M. Lima3 1Division of Gastroenterology, Hepatology, & Nutrition, Cincinnati Children's Hospital Medical Center, 3333 Burnet Ave., MLC 2010, Cincinnati, OH, USA
2Center

for Global Child Health, Department of Pediatrics, Cincinnati Children's Hospital Medical Center, 3333 Burnet Ave., MLC 2010, Cincinnati, OH, USA
3Instituto

de Biomedicina, Faculdade de Medicina, Universidade Federal do Cear, Fortaleza, Cear, Brazil

4Division

of Infectious Disease and International Health, Department of Medicine, University of Virginia, Charlottesville, VA, USA

Abstract
Background & AimsProlonged episodes of acute diarrhea (ProD, duration 713 days) or persistent diarrhea (PD, duration 14 days) are important causes of undernutrition, yet the epidemiology and nutritional impact of ProD are poorly understood. MethodsWe conducted a 10-year cohort study of 414 children from a Brazilian shantytown who were followed from birth; data were collected on diarrhea, enteric pathogens, and anthropometry. ResultsDuring 1,276 child-years of observation, we recorded 3,257 diarrheal episodes. ProD was twice as common as PD (12% and 5% of episodes, respectively); ProD and PD together accounted for 50% of all days with diarrhea. ProD was more common in infants whose mothers had not completed primary school (relative risk [RR]=2.1; 95% confidence interval=1.022.78). Early weaning was associated with earlier onset of ProD (Spearman's , 0.309; P=0.005). Infants with ProD were twice as likely to develop PD in later childhood (log rank P=0.002) compared to infants with only acute diarrhea (AD, duration <7 days), even after controlling for confounders. Children's growth was more severely stunted before their first episode of ProD, compared with AD (mean height-forage Z score (HAZ) -0.81 vs. -0.51, P<0.05, unpaired t test). Following ProD, HAZ (HAZ=-0.232) and weight-for-age (WAZ=-0.26) significantly decreased (P<0.005 in paired t tests). ProD was associated with Cryptosporidium and Shigella infections.

Correspondence to: Sean R. Moore, MD, MS, Division of Gastroenterology, Hepatology, & Nutrition, Cincinnati Children's Hospital Medical Center, 3333 Burnet Avenue, MLC 2010, Cincinnati, OH 45229-3039, Phone (513) 636-4415, Fax (513) 636-7805, sean.moore@cchmc.org. Study design: SRM, NLL, RLG, ALL; Figures: SRM, RCP; Data collection: AMS, RLG, AAL, LJB; Data analysis/interpretation: All authors; Writing: SRM, RCP, RLG, AAL. Disclosures: No conflicts of interest. Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Moore et al.

Page 2

ConclusionsProD accounts for significant morbidity and identifies children at risk of a vicious cycle of diarrhea and malnutrition. Further studies are needed to address the recognition and control of ProD and its consequences in resource-limited settings and assess its role in PD pathogenesis. Keywords Longitudinal study; chronic diarrhea; cryptosporidiosis; shigellosis

NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript

Introduction
Despite substantial reductions in diarrhea mortality over recent decades, diarrhea and its reciprocal relationship to undernutrition remain leading causes of global disability-adjusted life years (DALYs).1 Diarrhea causes 1 to 2.5 billion illnesses and 1.5 to 2.5 million deaths per year in children under the age of 5 in developing countries.2, 3 Furthermore, heavy burdens of diarrhea and undernutrition in early childhood profoundly impair growth and development. 4 The majority of diarrheal illnesses are acute in duration (AD, <14 days); however, a subset of illnessesknown as persistent diarrhea (PD)begin acutely then persist 14 days or longer, thereby resulting in a disproportionate share of diarrhea-related morbidity and mortality.5 Moreover, PD illnesses are associated with growth faltering,6 micronutrient deficiencies,7, 8 impaired neurodevelopment,9 and increased morbidity and mortality from other childhood diseases.10 Advances in the prevention and treatment of PD were the recent focus of an international PD Working Group who identified several research priorities for PD.11 Chief among these was the need to characterize prolonged episodes of acute diarrhea (ProD, duration 713 days) in order to further understand the pathogenesis and classification of PD, as well as the potential impact of interventions to halt the progression to persistent diarrhea. Although a number of studies have addressed the epidemiology of AD and PD, no studies have yet examined ProD as a distinct category of diarrhea. Thus, ProD's specific risk factors, etiologies, nutritional impact, and relationship to PD are unknown. To the extent that ProD represents an intermediate stage in the continuum from acute to persistent diarrhea, such studies would greatly enhance our understanding of PD's evolution and potential strategies for its control. To elucidate the epidemiology and impact of ProD in a highly endemic setting, we analyzed diarrheal episodes from a 10-year prospective birth cohort study of persistent diarrhea in Gonalves Dias, an urban shantytown in Brazil's developing Northeast region. During the first 4 years of surveillance, we reported on PD and its associated pathogens,6 followed by an analysis of community-wide improvements in the incidence of diarrhea and prevalence of malnutrition, due, in part, to intense surveillance and an increase in exclusive breastfeeding. 12 Now, with over a decade of cohort data collected, we report herein that ProD accounts for significant morbidity and is a key predictor of PD, with distinct risk factors, etiologies, and associations with undernutrition.

Methods and Materials

Study site and design The study was conducted from August 1989 to March 2000 in a 5-block area of a shantytown, Gonalves Dias; located in Fortaleza, Brazil. Fortaleza (population 3.5 million) is the capital city of the northeastern state of Cear (population 8.2 million). In 1994, Fortaleza's infant mortality was 59/1,000 live births. At the study's midpoint, Gonalves Dias was a community of 1,826 people, 13.5% (247) of whom were children <5 years old. The study team identified
Gastroenterology. Author manuscript; available in PMC 2011 October 1.

Moore et al.

Page 3

all pregnant women in the community and invited them to enroll in the project. Women choosing to participate provided informed consent and completed a detailed demographic and socioeconomic questionnaire with the assistance of a study nurse. In the first 45 months of the study, nurses visited the home of each newborn child three times weekly to record diarrheal illnesses. Thereafter, nurses visited homes twice weekly. At each visit, the study team also asked detailed information on the child's diet, including breastfeeding. Mothers (or guardians) of children with diarrhea were asked to give detailed clinical information about each illness, including accompanying symptoms and stool consistency and character. Study nurses visited children with diarrhea daily until 48 h after resolution of the illness. We used World Health Organization (WHO) guidelines to define diarrhea as 3 looser than normal stools in the preceding 24-hr period.13 An episode of diarrhea was defined as lasting 1 day and separated from another episode by 2 or more days without diarrhea. Discrete episodes of diarrhea were classified by their total duration. Acute diarrhea (AD) was defined as an episode that lasted <7 days, a prolonged episode of acute diarrhea (ProD) as an episode that lasted 7 and <14 days, and persistent diarrhea (PD) as an episode that lasted 14 days. During the time period of the study, rotavirus vaccine was not yet available in Northeast Brazil, 14 nor had WHO recommendations for zinc supplementation during diarrhea been published; thus, participants did not receive these interventions. When children developed severe or bloody diarrhea (or other worrisome conditions), they were promptly referred by a study nurse to a pediatrician, who prescribed antimicrobials when appropriate. We did not; however, collect detailed data on children's use of medications. Human investigation committees at the Federal University of Cear, the Johns Hopkins University School of Medicine, and the University of Virginia approved the study protocol. Nutritional assessment We measured weights and lengths of all children at 3-month intervals. A calibrated-sling Salter scale was used to obtain lightly-clothed weights to the nearest 0.1 kg. Lengths were measured supine to the nearest centimeter. We used NutStat (Epi Info version 3.5.1, Center for Disease Control and Prevention, Atlanta) to calculate weight-for-age, height-for-age, and weight-forheight Z scores (WAZ, HAZ, and WHZ, respectively); sex- and age-adjusted measures of the number of standard deviations above or below the median for the international reference population.15 Microbiologic studies Detailed methods and materials for stool collection and examination have been described elsewhere.6 Briefly, specimen cups were distributed to homes of children with diarrhea and caretakers were instructed to collect the first stool the following day. To provide controls, specimens were also obtained every 3 months from children who had no diarrhea for at least 14 days. Control specimens were not matched by age or season to diarrheal specimens. Stools were not collected from children who had ingested antibiotics in the previous 3 days. Specimens were transported on ice and processed within 4 h of collection. All specimens were examined by microscopy for parasites and leukocytes. One aliquot from each specimen was frozen and stored for viral testing by use of ELISA and PCR assays. Stool samples were screened for Norwalk-like viruses by RT-PCR; rotavirus antigen by ELISA (Rotaclone; Cambridge Biotech, Worcester, MA); torovirus antigen by ELISA after sucrose partial purification of virus from stool by sucrose gradient purification; and adenovirus by ELISA (Adenoclone, Cambridge Biotech). Initial bacterial cultures were done by use of MacConkey's, xylose-lysinedeoxycholate, and Gram-negative broth enrichment and subculture, TCBS (thiosulfate, citrate, bile salt, and sucrose) agar, and ampicillin blood agar

NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript

Gastroenterology. Author manuscript; available in PMC 2011 October 1.

Moore et al.

Page 4

plates (with 10 mg/mL ampicillin). Lactose-fermenting colonies were selected from MacConkey's agar plates and saved for further toxin, adherence, and serologic testing.

NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript

Statistical analyses Statistical analyses were performed using SPSS software (version 17.0; SPSS, Chicago) and Epi-Info. Statistical tests included Spearman's rank correlation for age of first ProD episodes with age at weaning; paired t-tests for comparison of nutritional status before and after a ProD episode; X2 or Fisher's exact test for associations of pathogens with prolonged, persistent, acute, and non-diarrheal stools. All tests were 2-tailed, and P values <0.05 were considered statistically significant. To determine whether early childhood ProD is a risk factor for subsequent PD, we used KaplanMeier analysis. Children were classified by whether or not they experienced ProD in their first year of life. Age in days at the time of a child's first episode of PD was used as the time variable. Children without PD were censored as they left the study or at the end of the surveillance period. The 8 children who developed PD prior to their first episode of ProD were excluded, giving a total of 406 children in the overall analysis. To account for possible confounders, we performed Cox regression to control for covariates associated (P<0.20) with persistent diarrhea. Covariates were identified by univariate analyses. Variables considered for inclusion in the multivariate model included: birthweight, child gender, type of house, number of people in household, type of sanitation, mother's level of education, age at weaning from exclusive breastfeeding, and date of birth (day, month, and year). Of these, number of people in household, type of sanitation, mother's years of education, age at weaning from exclusive breastfeeding and birth date were associated at P<0.20 using appropriate univariate analyses. These variables were subsequently included as covariates in a series of Cox regression equations.

Results
We enrolled 414 children (193 males, 47%) during the study period. Children were followed a mean of 1,125 days, a median of 803.5 days (range, 103,769), and the cumulative number of days of observation was 465,953 (1,276 child-years). Overall, these children had 2.55 episodes of diarrhea per child-year (3,257 episodes per 1,276 child-years), spending an average of 10.4 days per year with diarrhea (13,724 days per 1,276 child-years). Of the 414 children in the cohort, 68 (16.4%) had no diarrhea recorded during the study; however, this subset was followed a median of only 83 days (range, 10 to 2194) compared to the median of 1,094 days (range, 24 to 3,769) for children with at least one recorded episode of diarrhea. The mean duration of a diarrheal episode was 4.2 days (SD, 5.2); the median duration was 3 days. Of 3,257 episodes, 2,722 (83.6%) were AD, 380 (11.7%) were ProD, and 155 (4.7%) were PD illnesses. Figure 1 shows age-specific attack rates and days with diarrhea for total, AD, ProD, and PD episodes. Total, AD, and PD all peaked at 6-12 months of age (5.15, 4.22, and 0.68 episodes per child-year, respectively). PD peaked at 12-24 months of age (0.26 episodes per child-year). As shown in Figure 2, ProD accounted for only 11.7% (380/3,257) of diarrheal episodes but 25.2% (3463/13,724) of all days of diarrhea. Similarly, PD accounted for only 4.7% (155/3257) of episodes but 24.5% (3368/13,724) of days of diarrhea. Importantly, PD episodes accounted for 155 (29%) of the 535 diarrheal episodes 7 days in duration in our study. Thus, when a child's diarrheal episode progressed from acute to prolonged, the overall risk that the episode would evolve into PD increased from 4.8% to 29% (RR=6.09; 95% CI, 4.96-7.45).

Gastroenterology. Author manuscript; available in PMC 2011 October 1.

Moore et al.

Page 5

Maternal education and early weaning were significant risk factors for ProD. Of the 210 children with >11 months of surveillance data in the first year of life, the 157 whose mothers did not complete primary school had a 2-fold higher incidence of prolonged diarrhea (relative risk [RR], 2.1; 95% confidence interval [CI] 1.02-2.78) than children whose mothers finished primary school. In the 112 infants with complete follow-up and ProD the first year of life, age at weaning from exclusive breastfeeding was positively correlated with age at first ProD episode (Spearman's =0.309; P=0.005), i.e., early weaning was associated with earlier onset of ProD. Birthweight, family income, crowding, and lack of a toilet in the home were not significantly associated with ProD. Table 1 shows etiologic agents found in stools collected from children with ProD and controls in the first four years of the study. We obtained 132 samples from a total of 130 ProD episodes (34% of all ProD episodes), as well as 442 control samples from children without diarrhea. Cryptosporidium and Shigella species were isolated more frequently from ProD than control specimens (P=0.0178 and P=0.0223, respectively). Ascaris ova were found more frequently in controls than in ProD specimens (P=0.008). To further identify etiologies specific to ProD, we compared ProD data with previous results for AD and PD from this same cohort of children. 5 ProD episodes were delineated from AD episodes to create four categories of specimens: AD (<7 days), ProD (7 and <14 days), PD ( 14 days), and control. Cryptosporidium was seen more frequently in ProD than AD specimens (P=0.023). Giardia was more common in PD than ProD episodes (P=0.012). No other significant differences between ProD and other categories were detected. To evaluate the effects of ProD on subsequent risk for PD we used Kaplan-Meier analysis and Cox regression. Time to first PD episode was compared for 112 children with ProD before age 1 year and 294 children with no ProD before age 1 year. Prior to the analysis, we noted that 57 (80.3%) of the 71 children who developed PD during the study also experienced ProD. Of these 57 children, 30 (52.6%) had ProD in their first year of life. Kaplan-Meier analysis showed that children with ProD before age 1 year were significantly more likely to develop PD later in childhood compared to those without ProD (Log Rank P=0.002). As shown in Figure 3, children who experienced at least one ProD episode of diarrhea before age 1 were nearly twice (15.5% (s.e.=2.5%)) versus 29.9% (s.e.=5.0) as likely to experience PD by age 2 years. Cox Regression showed (OR 2.2, 95% CI; [1.323.54], P=0.002) a similar overall increased risk of PD by age 6 years. ProD before age 1 year remained a significant (P=0.025) independent predictor of later PD even after number of people in household, type of sanitation, mother's years of education, age at weaning from exclusive breastfeeding, and birth date were included in the equation. Only date of birth was a significant (P=0.004) independent predictor of PD due to previously documented improvements over time in the incidence of PD in this cohort.
11

NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript

The incidence of AD was significantly higher among the 195 (47%) children who experienced at least one episode of ProD or PD (mean incidence of 3.12 (s.d=2.3) AD episodes/child-year) versus children who experienced only AD or no diarrhea (mean incidence 1.91 (s.d.=2.7) AD episodes/child-year; P<0001, unpaired t test). Only 11 children experienced ProD and/or PD with no episodes of AD. Figure 4 shows the impact of AD (duration, <7 days), ProD, and PD illnesses on a child's anthropometry. To determine changes in nutritional status after a diarrheal illness, we identified children's first AD, ProD, and PD episodes and compared anthropometric Z scores 3 months prior to and following the episode. We found significant decreases in WAZ (-0.19 to -0.45, P=0.0001) and HAZ (-0.81 to -1.40, P=0.0002) in the period after children's first ProD illness. Similarly, first AD episodes were followed by declines in WAZ (0.02 to -0.28, P<00001) and

Gastroenterology. Author manuscript; available in PMC 2011 October 1.

Moore et al.

Page 6

HAZ (-0.51 to -0.82, P<0.0001). Furthermore, children's first PD illnesses were accompanied by declines in both WAZ (-0.46 to -0.74, P=00012) and WHZ (0.44 to 0.17, P=0.0015).

NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript

We also noted striking downward trends in WAZ and HAZ as children progressed from first AD to ProD and PD episodes. Children's HAZ scores (Figure 4C) prior to ProD or PD episodes were significantly worse than children's HAZ before their first AD episode (P<005). Likewise, WAZ prior to PD was significantly worse than WAZ prior to AD episodes (Figure 4B).

Discussion
In this decade-long cohort study of children followed from birth in an urban Brazilian shantytown, we found that ProD accounts for significant morbidity and is associated with early weaning, low maternal schooling, and undernutrition. Furthermore, we showed that ProD in infancy is a strong indicator of children's future risk of PD. To our knowledge, this is the first study to examine ProD as a distinct class of diarrhea, thus our findings have several important public health implications that merit further consideration. First, although ProD and PD comprise only a small portion of diarrheal episodes (12% and 5%, respectively), together they account for half of all days of diarrhea. In addition, once an acute episode of diarrhea progresses to ProD, the relative risk that the episode will evolve into PD is 6-fold higher. These two findings suggest that targeted interventions to halt the progression from AD to ProD would not only mitigate PD and its consequences, but have a substantial impact on overall diarrhea burdens as well. Our second key finding is that infants with ProD have a 2.2-fold higher risk of developing PD in later childhood. In our analysis, ProD remains a significant risk factor for PD even after controlling for a number of confounders. ProD may therefore identify infants whose pathogens, nutritional status, diet, environment, or genes predispose them to PD. Alternatively, ProD may play a direct role in the evolution of PD via effects on growth, innate immunity, disruption of intestinal barrier function, or alteration of gut flora. Regardless of its nature, ProD clearly highlights infants at risk of both PD and growth faltering who therefore warrant increased vigilance and support in order to prevent a vicious cycle of diarrhea and malnutrition. Third, we find that lack of a primary school education and early weaning from exclusive breastfeeding are significant maternal risk factors for infantile ProD. Several studies from developing countries indicate that children of uneducated mothers are at increased risk of diarrhea.16-18 Despite limited schooling, a number of educational campaigns have been shown to reduce the incidence of childhood diarrhea in these settings, e.g., handwashing interventions in Pakistan,19, 20 and exclusive breastfeeding campaigns in India.21 Our present study indicates that early weaning from exclusive breastfeeding was associated with earlier onset of ProD. This supports previous findings by our group and others that breastfeeding is highly protective against diarrhea and limits the duration of diarrheal illnesses.6 Thus, promotion of breastfeeding must remain an essential component of diarrheal control programs. Fourth, we demonstrate that a first episode of ProD is associated with undernutrition both before and after the illness. Prior to ProD, children's mean HAZ was already significantly below the WHO median. Following ProD, we see further declines in both HAZ (indicating stunting, a marker of chronic undernutrition) and WAZ (indicating underweight, a marker of more acute malnutrition). In addition, we find baseline decrements in HAZ and WAZ prior to both ProD and PD versus AD episodes. This trend is not seen for low WHZ (an indicator of wasting), which was less common in our study due to the number of children who were both underweight and stunted. We interpret these results as further evidence of a vicious cycle of diarrhea and malnutrition in which diarrhea causes undernutrition and, in turn, poor nutritional status predisposes to further, lengthier episodes of diarrhea. Importantly, these acute effects of ProD
Gastroenterology. Author manuscript; available in PMC 2011 October 1.

Moore et al.

Page 7

on nutritional status likely have chronic implications for children's growth. Despite catch-up growth following diarrheal illnesses, our studies and a more recent multi-country analysis have shown that early childhood diarrhea remains highly predictive of stunting at age 2 years and beyond.22, 23 Finally, we found that Shigella and Cryptosporidium species were etiologies significantly associated with ProD episodes. Shigella is well-recognized as a major cause of dysentery in developing countries and has consistently been shown to associate with growth faltering, as well as ProD and PD.24 Because of this impact and the worldwide emergence of antibioticresistant strains, WHO has targeted the development of a Shigella vaccine as a high priority. 25 Similarly, Cryptosporidium is increasingly recognized as an important cause of severe diarrhea and undernutrition in tropical, developing regions; however, existing therapies are only partially effective.26 The antiparasitic drug nitazoxanide has demonstrated efficacy for childhood cryptosporidiosis,27 however it is has not been shown effective for cryptosporidiosis in undernourished children or patients with HIV/AIDS.28 In addition to the pressing need for effective vaccines and therapies for Shigella and Cryptosporidium, recent WHO reports highlight more immediate approaches to control childhood diarrhea in developing countries, which are of course relevant to ProD. These include: 1) improved outcomes through rotavirus and measles vaccination; promotion of early breastfeeding and Vitamin A supplementation; promotion of handwashing with soap; improved water supply and quality, including household water treatment and safe storage of household water; and community-wide sanitation, and 2) improved management of diarrhea by scaling up implementation of low-osmolarity ORS and zinc supplementation worldwide.2,
29

NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript

Of note, Ascaris was found more frequently in control stools than in ProD specimens, but not AD specimens. We speculate that Ascaris might have a mitigating effect on the duration of intestinal infections by altering intestinal immune responses to diarrheal pathogens. Alternatively, ProD may purge Ascaris ova from the digestive tract, making them more difficult to identify in stools. The major strengths of our study include a large sample of children followed intensively from birth for diarrheal, nutritional, and microbiologic surveillance over multiple seasons in a highly endemic setting. Our study was limited by several factors. First, we recorded substantial improvements over time in the incidence of diarrhea and undernutrition in this community changes not seen in nearby shantytowns.12 Second, loss to follow-up was common and previous analyses suggest that dropouts were more likely to be underweight.22 Therefore, our current results may represent a best-case scenario, which underestimates the true impact of ProD in communities where undernutrition and diarrhea morbidity remain endemic. Third, stools were collected only during the first four years of the study and only 34% of ProD episodes were sampled. Fourth, all stools obtained from children with ProD were examined for bacterial and parasitic pathogens; however only 26% of samples were tested for viral pathogens. Previous findings from our study site (in a different cohort of children) revealed that the use of antibiotics is very common, often inappropriate, and may be a risk factor for diarrhea.30, 31 Lack of adequate data on antibiotic use in the current study prevented us from defining the extent to which antibiotics or other drugs may influence the risk of ProD or PD. Lastly, our study represents results from only one community and may not be generalizable across multiple settings. Several of these limitations will be addressed by the recently launched Global Enterics Multi-Center Study and MAL-ED Network, multi-country investigations of the complex interrelationships of malnutrition and a broader range of enteric infections.32, 33

Gastroenterology. Author manuscript; available in PMC 2011 October 1.

Moore et al.

Page 8

There is limited consensus between pediatric and adult subspecialties on definitions of acute, prolonged, persistent, and chronic episodes of diarrhea. In general, infectious disease and pediatric GI texts use the WHO cutoff of 2 weeks to delineate persistent from acute episodes. 34,35 Some experts use chronic for illnesses lasting >30 days.36 Conversely, adult GI texts use chronic to describe episodes lasting >2 or >4 weeks.37, 38 Because our data indicate that ProD identifies children at increased risk of PD and the vicious cycle of diarrhea and malnutrition, we propose that the current WHO classification of acute diarrhea as <14 days be revised as follows: 1) acute diarrhea, <7 days and 2) prolonged episodes of acute diarrhea, 7 and <14 days. We conclude that ProD: 1) accounts for a substantial portion of diarrhea burdens in this tropical, developing setting; 2) robustly predicts persistent diarrhea; and 3) associates with undernutrition, growth faltering, early weaning, and low maternal educational status. Further studies are needed to assess the role of ProD in the pathogenesis of persistent diarrhea, as well as the recognition, prevention and treatment of ProD in resource-limited settings, with the overall goal of preventing persistent diarrhea, as well as long-term growth and neurodevelopmental deficits in at-risk children.

NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript

Acknowledgments
The authors thank the participating children and their families; field workers Sayonara S.B. Alencar, Maria de Lourdes P. Rodrigues, Maria Luzia F. Melo and Rosania Silva; and laboratory technicians Maria do Carmo, N. Pinho, Maria Isabel L. McAuliffe, Jania Teixeira, and Coneicao Nogueria Raulino. This work was supported by the National Institute of Allergy and Infectious Diseases grant U01AI226512 to RLG and AAML. SRM was a Pediatric Scientist Development Program Fellow sponsored by the Eunice Shriver Kennedy National Institute of Child Health and Human Development (NICHD Grant Award K12-HD000850) and is now supported by NICHD Grant Award K12-HD028827. This project was supported in part by PHS Grant P30 DK 078392. The authors also thank Lee Denson and Mitchell Cohen for their thoughtful comments on the manuscript. Funding: This work was supported by NIAID grant U01AI226512 to RLG and AAML. SRM was a Pediatric Scientist Development Program Fellow sponsored by the Eunice Shriver Kennedy National Institute of Child Health and Human Development (NICHD Grant Award K12-HD000850) and is now supported by NICHD Grant Award K12-HD028827. Supported in part by PHS Grant P30 DK 078392.

References
1. WHO. The global burden of disease: 2004 update. 2008. 2. UNICEF/WHO. Diarrhoea: Why children are still dying and what can be done. 2009. 3. Kosek M, Bern C, Guerrant RL. The global burden of diarrheal disease, as estimated from studies published between 1992 and 2000. Bull World Health Organ 2003;81:197204. [PubMed: 12764516] 4. Petri WA Jr, Miller M, Binder HJ, Levine MM, Dillingham R, Guerrant RL. Enteric infections, diarrhea, and their impact on function and development. J Clin Invest 2008;118:127790. [PubMed: 18382740] 5. McAuliffe JF, Shields DS, Auxiliadora de Sousa M, 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;5:9026. [PubMed: 3794909] 6. Lima AAM, Moore SR, Barboza MS, et al. Persistent diarrhea signals a critical period of increased diarrhea burdens and nutritional shortfalls: A prospective cohort study among children in northeastern Brazil. Journal of Infectious Diseases 2000;181:16431651. [PubMed: 10823764] 7. Lukacik M, Thomas RL, Aranda JV. A meta-analysis of the effects of oral zinc in the treatment of acute and persistent diarrhea. Pediatrics 2008;121:32636. [PubMed: 18245424] 8. Villamor E, Mbise R, Spiegelman D, et al. Vitamin A supplements ameliorate the adverse effect of HIV-1, malaria, and diarrheal infections on child growth. Pediatrics 2002;109:E6. [PubMed: 11773574] 9. Guerrant DI, Moore SR, Lima AAM, Patrick PD, Schorling JB, Guerrant RL. Association of early childhood diarrhea and cryptosporidiosis with impaired physical fitness and cognitive function four-

Gastroenterology. Author manuscript; available in PMC 2011 October 1.

Moore et al.

Page 9

seven years later in a poor urban community in northeast brazil. American Journal of Tropical Medicine and Hygiene 1999;61:707713. [PubMed: 10586898] 10. Black RE, Allen LH, Bhutta ZA, et al. Maternal and child undernutrition: global and regional exposures and health consequences. Lancet 2008;371:24360. [PubMed: 18207566] 11. Bhutta ZA, Nelson EA, Lee WS, et al. Recent advances and evidence gaps in persistent diarrhea. J Pediatr Gastroenterol Nutr 2008;47:2605. [PubMed: 18664885] 12. Moore SR, Lima AA, Schorling JB, Barboza MS Jr, Soares AM, Guerrant RL. Changes over time in the epidemiology of diarrhea and malnutrition among children in an urban Brazilian shantytown, 1989 to 1996. Int J Infect Dis 2000;4:17986. [PubMed: 11231179] 13. Persistent diarrhoea in children in developing countries: memorandum from a WHO meeting. Bull World Health Organ 1988;66:70917. [PubMed: 3266111] 14. Gurgel RG, Bohland AK, Vieira SC, Oliveira DM, Fontes PB, Barros VF, Ramos MF, Dove W, Nakagomi T, Nakagomi O, Correia JB, Cunliffe N, Cuevas LE. Incidence of rotavirus and all-cause diarrhea in northeast Brazil following the introduction of a national vaccination program. Gastroenterology 2009;137:19705. [PubMed: 19632228] 15. WHO. WHO child growth standards: length/height-for-age, weight-for-age, weight-for-length, weight-for-height and body mass index-for-age: methods and development. 2006. 16. Hatt LE, Waters HR. Determinants of child morbidity in Latin America: a pooled analysis of interactions between parental education and economic status. Soc Sci Med 2006;62:37586. [PubMed: 16040175] 17. Sur D, Deen JL, Manna B, et al. The burden of cholera in the slums of Kolkata, India: data from a prospective, community based study. Arch Dis Child 2005;90:117581. [PubMed: 15964861] 18. Dargent-Molina P, James SA, Strogatz DS, Savitz DA. Association between maternal education and infant diarrhea in different household and community environments of Cebu, Philippines. Soc Sci Med 1994;38:34350. [PubMed: 8140461] 19. Luby SP, Agboatwalla M, Painter J, Altaf A, Billhimer WL, Hoekstra RM. Effect of intensive handwashing promotion on childhood diarrhea in high-risk communities in Pakistan: a randomized controlled trial. JAMA 2004;291:254754. [PubMed: 15173145] 20. Luby SP, Agboatwalla M, Feikin DR, et al. Effect of handwashing on child health: a randomised controlled trial. Lancet 2005;366:22533. [PubMed: 16023513] 21. Bhandari N, Bahl R, Mazumdar S, Martines J, Black RE, Bhan MK. Effect of community-based promotion of exclusive breastfeeding on diarrheal illness and growth: a cluster randomised controlled trial. Lancet 2003;361:141823. [PubMed: 12727395] 22. Moore SR, Lima AAM, Conaway MR, Schorling JB, Soares AM, Guerrant RL. Early childhood diarrhoea and helminthiases associate with long-term linear growth faltering. Int J Epidemiol 2001;30:14571464. [PubMed: 11821364] 23. Checkley W, Buckley G, Gilman RH, et al. Multi-country analysis of the effects of diarrhoea on childhood stunting. Int J Epidemiol 2008;37:81630. [PubMed: 18567626] 24. Black RE, Brown KH, Becker S. Effects of diarrhea associated with specific enteropathogens on the growth of children in rural Bangladesh. Pediatrics 1984;73:799805. [PubMed: 6374599] 25. Kosek M, Yori PP, Pan WK, et al. Epidemiology of highly endemic multiply antibiotic-resistant shigellosis in children in the Peruvian Amazon. Pediatrics 2008;122:e5419. [PubMed: 18710884] 26. Kosek M, Alcantara C, Lima AA, Guerrant RL. Cryptosporidiosis: an update. Lancet Infect Dis 2001;1:2629. [PubMed: 11871513] 27. Rossignol JF, Ayoub A, Ayers MS. Treatment of diarrhea caused by Cryptosporidium parvum: a prospective randomized, double-blind, placebo-controlled study of Nitazoxanide. J Infect Dis 2001;184:1036. [PubMed: 11398117] 28. Sears CL, Kirkpatrick BD. Is nitazoxanide an effective treatment for patients with acquired immune deficiency syndrome-related cryptosporidiosis? Nat Clin Pract Gastroenterol Hepatol 2007;4:136 7. [PubMed: 17245327] 29. WHO. WHO/FCH/CAH/04.7. WHO; 2004. Clinical management of acute diarrhoea. 30. Schorling JB, De Souza MA, Guerrant RL. Patterns of antibiotic use among children in an urban Brazilian slum. Int J Epidemiol 1991;20:2939. [PubMed: 2066238]

NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript

Gastroenterology. Author manuscript; available in PMC 2011 October 1.

Moore et al.

Page 10

31. Schorling JB, De Souza MA, Guerrant RL. Antibiotic use among children in an urban Brazilian slum: a risk factor for diarrhea? Am J Public Health 1991;81:99100. [PubMed: 1983925] 32. The Global Enterics Multi-Center Study (GEMS). [December 22, 2009]. Available from: http://medschool.umaryland.edu/GEMS/ 33. MAL-ED: A Network for the Study of Malnutrition and Enteric Diseases. [December 22, 2009]. Available from: http://mal-ed.fnih.org/ 34. Steiner, TS.; Guerrant, RL. Principles and syndromes of enteric infections. In: Mandell, GL., et al., editors. Principles and practice of infectious diseases. 7th. Vol. 1. Philadelphia: Church Livingstone; 2009. 35. Guarino, A.; De Marco, G. Persistent diarrhea. In: Kleinman, RE., et al., editors. Pediatric Gastrointestinal Disease. 5th. Vol. 1. Hamilton: BC Decker Inc.; 2008. 36. Guerrant RL, Van Gilder T, Steiner TS, et al. Practice guidelines for the management of infectious diarrhea. Clin Infect Dis 2001;32:331351. [PubMed: 11170940] 37. Tarr, PI.; Bass, DM.; Hecht, GA. Bacterial, viral, and toxic causes of diarrhea, gastroenteritis, and anorectal infections. In: Yamada, Tadataka, editor. Textbook of gastroenterology. 5th. Vol. 1. Oxford: Wiley-Blackwell; 2008. 38. Schiler, LR.; Sellin, JH. Diarrhea. In: Feldman, M., editor. Slesinger and Fordtran's gastrointestinal and liver disease. 8th. Vol. 1. Philadelphia: Saunders; 2010.

NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript

ST

Abbreviations
AD ProD PD HAZ WAZ WHZ LT EPEC EAggEC EIEC acute diarrhea prolonged episodes of acute diarrhea persistent diarrhea height-for-age Z score weight-for-age Z score weight-for-height Z score heat labile heat stable enteropathogenic E. coli enteroaggregative E. coli enteroinvasive E. coli

Gastroenterology. Author manuscript; available in PMC 2011 October 1.

Moore et al.

Page 11

NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript

Gastroenterology. Author manuscript; available in PMC 2011 October 1.

Moore et al.

Page 12

NIH-PA Author Manuscript NIH-PA Author Manuscript

Figure 1.

A. Diarrhea attack rates of acute (<7 days), prolonged (7 and <14 days), and persistent (14 days) episodes per child-year, by age, among 414 children in Gonalves Dias in Fortaleza, Brazil, from August 1989 through March 2000. B. Days with diarrhea per child-year, by age. Episodes and days of diarrhea for prolonged and persistent diarrhea peaked at 6-12 and 1224 months of age, respectively.

NIH-PA Author Manuscript

Gastroenterology. Author manuscript; available in PMC 2011 October 1.

Moore et al.

Page 13

NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript

Gastroenterology. Author manuscript; available in PMC 2011 October 1.

Figure 2.

Proportions of total diarrhea morbidity accounted for by acute (<7 days), prolonged (7 and <14 days), and persistent (14 days) episodes and by days of illness.

Moore et al.

Page 14

NIH-PA Author Manuscript

Figure 3.

NIH-PA Author Manuscript NIH-PA Author Manuscript

Proportional hazard curves of time to first persistent diarrheal (PD, duration 14 days) episode in children with (n=112) and without (n=294) a prolonged episode of diarrhea (ProD, duration 7 and <14 days) in the first year of life. Eight children who developed PD prior to their first episode of ProD were excluded. Children who experienced at least one ProD episode of diarrhea before age 1 were nearly twice (15.5% (s.e.=2.5%) versus 29.9% (s.e.=5.0) as likely to experience a PD episode by age 2 years. Cox Regression showed (OR 2.2, 95% CI; [1.32-3.54], P=0.002) a similar overall increased risk of PD by age 6 years. ProD before age 1 year remained a significant (P =0.025) independent predictor of later PD even after controlling for household crowding, type of sanitation, maternal education, age at weaning from exclusive breastfeeding, and birth date.

Gastroenterology. Author manuscript; available in PMC 2011 October 1.

Moore et al.

Page 15

NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript

Gastroenterology. Author manuscript; available in PMC 2011 October 1.

Figure 4.

Impact of acute (<7 days), prolonged (7 and <14 days), and persistent (14 days) diarrhea on anthropometry. Nutritional Z-scores were compared 3 months before and after children's first acute (n=308), prolonged (n=145), and persistent (n=62) episodes. A. Weight-for-height Z (WHZ) scores declined significantly following persistent episodes, but not acute or prolonged episodes. B. Weight-for-age Z (WAZ) scores decreased with all episodes types and mean WAZ prior to acute diarrhea was significantly greater than mean WAZ prior to persistent diarrhea. C. Height-for-age Z (HAZ) scores declined following acute and prolonged episodes, but not persistent episodes. HAZ prior to acute diarrhea was greater than HAZ prior to prolonged or persistent diarrhea. (Error bars indicate SEM; *P<0.005, paired t-test of Z scores before and after diarrhea; P<0.05, unpaired t-test of Z scores prior to acute vs. prolonged, acute vs. persistent, or prolonged vs. persistent episodes.)

Moore et al.

Page 16

Table 1

Etiologic studies of 130 prolonged episodes of acute diarrhea (duration 7 and <14 days) among children in Fortaleza, Brazil.
Prolonged Diarrheaa Etiologic Agent Virus Norwalk-like virus Rotavirus Torovirus Adenovirus Parasite (all pathogens) Cryptosporidium species Giardia lamblia Ascaris species Trichuris species Entamoeba histolytica 38.5% (5/13) 2.9% (1/34) 0.0% (0/5) 5.9% (2/34) 21.2% (28/132) 12.2% (12/98) 6.8% (9/132) 5.3% (7/132) 4.5% (6/132) 0.8% (1/132) 0.0% (0/132) 0.0% (0/8) 12.1% (16/132) 4.5% (6/132) 0.0% (0/132) 0.0% (0/132) 0.8% (1/132) 0.8% (1/132) 6.1% (8/132) 3.0% (4/132) 3.0% (4/132) 4.4% (4/90) 31.1% (23/74) 5.6% (5/90) 6.7% (6/90) 0.0% (0/90) 1.2% (1/84) 50.8% (67/132) <0.0001 <0.0001 0.0223 0.008 0.0178 24.0% (12/50) 2.4% (2/82) 0.0% (0/17) 0.0% (0/82) 25.1% (111/442) 5.2% (15/289) 7.9% (35/442) 13.8% (61/442) 7.7% (34/442) 1.1% (5/442) 0.5% (2//442) 4.0% (1/25) 1.8% (8/440) 1.1% (5/440) 0.2% (1/440) 0.0% (0/440) 0.0% (0/440) 0.0% (0/440) 7.2% (14/195) 5.1% (10/195) 2.0% (4/195) 7.0% (14/195) 35.0% (64/184) 10.0% (19/195) 8.0% (9/112) 0.0% (0/112) 0.0% (0/112) 6.3% (28/442) % (no. with agent/no. tested) P Controlb % (no. with agent/no. tested)

NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript

Strongyloides species Microsporidium species Bacteria (all pathogens) Shigella species Salmonella species Vibrio cholera Vibrio parahemolyticus Yersinia species LT or ST E. Coli LT E. Coli ST E. Coli EPEC EAE probe EAggEC (Hep-2) AA probe DAEC EHEC probe EIEC Fecal leukocytes Lactoferrin All Mild High Any pathogen Multiple pathogens

77.1% (27/35) 45.7% (16/35) 31.4% (11/35) 74.5% (73/98) 26.5% (26/98)

0.0047 0.0102

49.5% (47/95) 25.3% (24/95) 24.2% (23/95) 73.0% (143/194)

0.0058

40.7% (82/194)

Data represent the presence or absence of specified etiologies in case or control fecal specimens. P values are for differences between controls and cases. LT, heat labile; ST, heat stable; EPEC, enteropathogenic E. coli; EAggEC, enteroaggregative E. coli, EIEC, enteroinvasive E. coli.

Gastroenterology. Author manuscript; available in PMC 2011 October 1.

Moore et al.
a b

Page 17

Includes 132 specimens representing 130 illnesses. Includes 442 control specimens.

NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript

Gastroenterology. Author manuscript; available in PMC 2011 October 1.