Clinical, environmental, and behavioral characteristics associated with Cryptosporidium infection among children with moderate-to-severe diarrhea in rural western Kenya, 2008–2012: The Global Enteric Multicenter Study (GEMS)

Background Cryptosporidium is a leading cause of moderate-to-severe diarrhea (MSD) in young children in Africa. We examined factors associated with Cryptosporidium infection in MSD cases enrolled at the rural western Kenya Global Enteric Multicenter Study (GEMS) site from 2008-2012. Methodology/Principal findings At health facility enrollment, stool samples were tested for enteric pathogens and data on clinical, environmental, and behavioral characteristics collected. Each child’s health status was recorded at 60-day follow-up. Data were analyzed using logistic regression. Of the 1,778 children with MSD enrolled as cases in the GEMS-Kenya case-control study, 11% had Cryptosporidium detected in stool by enzyme immunoassay; in a genotyped subset, 81% were C. hominis. Among MSD cases, being an infant, having mucus in stool, and having prolonged/persistent duration diarrhea were associated with being Cryptosporidium-positive. Both boiling drinking water and using rainwater as the main drinking water source were protective factors for being Cryptosporidium-positive. At follow-up, Cryptosporidium-positive cases had increased odds of being stunted (adjusted odds ratio [aOR] = 1.65, 95% CI: 1.06–2.57), underweight (aOR = 2.08, 95% CI: 1.34–3.22), or wasted (aOR = 2.04, 95% CI: 1.21–3.43), and had significantly larger negative changes in height- and weight-for-age z-scores from enrollment. Conclusions/Significance Cryptosporidium contributes significantly to diarrheal illness in young children in western Kenya. Advances in point of care detection, prevention/control approaches, effective water treatment technologies, and clinical management options for children with cryptosporidiosis are needed.


Introduction
The Global Enteric Multicenter Study (GEMS) was undertaken to assess the burden and etiology of moderate-to-severe diarrhea (MSD) in seven countries, three in South Asia and four in sub-Saharan Africa. In all African sites, Cryptosporidium was the second-highest enteric pathogen attributable to infant MSD; in GEMS Kenya, Cryptosporidium was a major pathogen across all age groups (0-11, 12-23, and 24-59 months) [1]. Cryptosporidium was also identified as one of five pathogens with the highest attributable burden of infant diarrhea in a study of malnutrition and enteric disease (MAL-ED), a cohort study that compared diarrheal and nondiarrheal stools in children under two years old collected at community surveillance visits at 8 sites in South America, Africa, and Asia [2]. Based on GEMS data, it has been estimated that there are nearly three million annual diarrhea episodes attributable to Cryptosporidium in young children in sub-Saharan Africa [3]. Globally, acute Cryptosporidium infections are estimated to cause 48,000 annual deaths in children under five years old [4]. Cryptosporidium infections in young children in low-and middle-income countries have been associated with excess mortality [5], an excess burden of diarrhea later in life [6], and growth faltering, the deficits of which may not be recovered for those children infected during infancy [7]. Cryptosporidium has been associated with decreases in height-for-age z-scores in children, even in the absence of diarrhea symptoms [4]. Cryptosporidium infections have been associated with persistent diarrhea in Kenya [8,9]. Cryptosporidium is highly tolerant to disinfection with chlorine [10].
Nitazoxanide can treat cryptosporidiosis in immunocompetent children 1-11 years old [11]; however, it is not often available in developing countries [12] and is presently not approved for infants [11]. There is currently no vaccine available for Cryptosporidium; however, the evidence of acquired immunity suggests that one could be effective [12].
Although outbreaks of Cryptosporidium in developed countries have been studied in detail, less is known about risk factors for cryptosporidiosis in countries where it is endemic [10]. Reviews of risk factors for Cryptosporidium infection identified malnutrition, contact with domestic animals, non-exclusive breastfeeding in infants, lack of sanitation facilities, and crowded living conditions as possible risk factors for infection in low-and middle-income countries [13,14]. Few studies have examined risk factors for Cryptosporidium infection in Kenyan children [15][16][17]. In Kenya, risk factors for Cryptosporidium in children include being HIV-positive [17], or having an HIV-positive mother [15].
We describe the prevalence of Cryptosporidium infections in Kenyan children under five years old with MSD, assess clinical, environmental, and behavioral characteristics associated with Cryptosporidium infection, and describe the outcomes and consequences of cryptosporidiosis.

Global Enteric Multicenter Study (GEMS)
We evaluated data collected in Kenya from cases enrolled in GEMS, a four-year, prospective, age-stratified, health facility-based matched case-control study of MSD among children aged 0-59 months residing within a defined and enumerated population. The rationale, study design, clinical and microbiologic methods, and assumptions of GEMS have been described elsewhere [18,19]. Briefly, GEMS enrolled MSD cases from selected sentinel health facilities in each of three age strata (0-11, 12-23, and 24-59 months old), along with 1-3 matched community controls who had not had diarrhea in the week before enrollment. MSD was defined as having three or more loose stools in the previous 24 hours, with onset within the 7 days prior to enrollment, and having one or more of the following illness severity characteristics: loss of skin turgor, sunken eyes, required intravenous fluid rehydration, dysentery (blood in stool), or required hospitalization [18].
At enrollment, demographic, clinical, epidemiological information, and stool samples were collected. Cryptosporidium oocyst antigens were detected in whole stool specimens by enzyme immunoassay (EIA; TechLab, Inc, VA, USA). Detailed laboratory methods are described elsewhere [19]. DNA was extracted from a subset of stools that were EIA-positive for Cryptosporidium. Restriction fragment length polymorphism analyses and DNA sequencing of polymerase chain reaction (PCR) products were used to identify Cryptosporidium genotypes for these specimens at the U.S. Centers for Disease Control and Prevention (CDC) [20].
To assess each child's health status, a home visit including focused physical exams and anthropometric measurements was conducted~60 days (acceptable range 50-90 days) following enrollment. Mortality that occurred at any time between enrollment and this follow-up was recorded.

Study site
In Kenya, children were enrolled between January 31, 2008 and January 29, 2011, and again from October 31, 2011 to September 30, 2012. The study was conducted in Siaya County, in the areas of Gem and Asembo, and during the second enrollment period, in the areas of Asembo and Karemo due to the Kenya Medical Research Institute (KEMRI)/CDC health and demographic surveillance system moving activities. This health and demographic surveillance system has been operating in these communities since 2001. The study setting has high rates of child mortality, malaria, HIV, and tuberculosis, and has been described elsewhere [21,22].

Study definitions
Days of diarrhea. At enrollment caretakers were (1) asked how many days the case child had been experiencing diarrhea in the week before presenting at the health facility and (2) instructed on how to record the presence/absence of diarrhea for 14 days following enrollment on a Memory Aid form [18]. The total number of days of diarrhea was calculated as the sum of these durations, with a maximum of 21 possible diarrhea days, which may include multiple episodes [8]. Acute diarrhea was defined as having 1-6 days of diarrhea in the 21-day period; having 7-13 and !14 days refer to prolonged and persistent diarrhea, respectively. Anthropometry measurements. Height and weight for each child were measured at enrollment and at 60-day follow-up; details on measurement methods are described elsewhere [18]. Height/length-for-age, weight-for-age, and weight-for-height/length z-scores (HAZ, WAZ, and WHZ) were calculated using a WHO SAS macro with the WHO Child Growth Standards as the reference population [23,24]. For case children requiring rehydration, weight measurements after rehydration were used. If the child was under observation for more than four hours, weight measurements were repeated upon discharge from the health facility. To account for rehydration, the last recorded weight measurement was used as the baseline value. Stunted, underweight, and wasted were defined as having HAZ<-2, WAZ<-2, and WHZ<-2 respectively. Children with HAZ<-3, WAZ<-3, or WHZ<-3 were considered severely malnourished. All anthropometric measures identified as extreme measures using the WHO definition (|HAZ|>6, WAZ<-6, WAZ>5, or |WHZ|>5) were excluded [23], as were outliers identified using median absolute deviation methods [25].
HIV testing. A subset of cases and their biological mothers had HIV status captured. Children !18 months old had a rapid HIV antibody test and children <18 months old had a confirmatory PCR test. Persons identified as HIV positive were referred to care and treatment.
Water definitions. Surface water was defined as drinking water coming from a pond, lake, river, stream, or dam, as is standard practice [26], or water coming from an earth pan/ water pan, which is western Kenya specific. Standard definitions were used to define improved and unimproved water sources [26].
Breastfeeding definitions. Partial breastfeeding refers to providing children with breast milk and supplementing with other food or liquid, whereas exclusive breastfeeding refers to providing children only breast milk without supplementation.

Statistical analysis
Analyses were performed in SAS 9.4 (SAS Institute, Inc., Cary, NC) and R 3.4.0 (R Foundation for Statistical Computing, Vienna, Austria). To assess variables associated with Cryptosporidium positivity, univariable logistic regression models were used to compute odds ratios (ORs) and 95% confidence intervals (CIs). Since Cryptosporidium risk factors may be modified by age and the sample size might limit detection of interactions, we assessed for effect modification by age category for all variables in separate models with a p<0.05 cutoff for significance. To assess whether each risk factor was confounded by socioeconomic status (SES) we ran models with and without SES and considered confounding if effect sizes changed >10%.
Two multivariable models were generated to identify clinical, demographic, environmental, and behavioral characteristics associated with being Cryptosporidium-positive. The first model examined the clinical presentation at enrollment of case children, including age strata and sex (see Table 1) and all variables in Table 2 (except duration of diarrhea, which includes information collected post-enrollment). The second model sought to identify demographic, environmental, and behavioral characteristics that may be risk factors for Cryptosporidium infection (all variables in Table 1, and the following caretaker-reported water, sanitation, and hygiene characteristics collected at enrollment: primary source of drinking water, whether water was always available from the main drinking water source, whether the child was given stored water in the two weeks prior to enrollment, whether the caretaker boiled or filtered drinking water, whether there was a facility for feces disposal, whether the caretaker uses soap when washing hands, and whether the caretaker washes their hands at the following times: before eating, after defecating, before nursing, before cooking, after cleaning a child, and after touching an animal). Breastfeeding was not considered in either model as collinearity with age was identified, and information on breastfeeding is only available for children under two years old in the first three years of GEMS (n = 1,083); questions on breastfeeding changed during the fourth study year. Model selection was performed using the Least Absolute Shrinkage and Selection Operator (LASSO) method using the minimum error lambda [27,28]. While LASSO methods were used to identify variables for inclusion, parameter estimates and CIs are derived by standard logistic regression maximum likelihood methods. Anthropometry analyses. The proportions of children who were stunted, underweight, and wasted at enrollment and at follow-up were compared for Cryptosporidium-positive and Cryptosporidium-negative cases using ORs and 95% CIs. Models with follow-up anthropometry measures as outcomes controlled for baseline measures. Changes in HAZ, WAZ, and WHZ were calculated by subtracting the baseline measure from the follow-up measure. A Wilcoxon rank sum test was used to evaluate differences in distributions of z-scores for these changes for Cryptosporidium-positive and Cryptosporidium-negative cases. Only those with follow-up measurements within an acceptable range of 50-90 days were included.
For anthropometry analyses, effect modification by age and sex was assessed by including interaction terms one-by-one into the models. If interaction could be assessed (i.e., the model converged) and the interaction was significant at alpha<0.05, results were stratified accordingly.
Sensitivity analyses. Children could be re-enrolled as a case 90 days post-enrollment. Children who were enrolled more than once as a case into GEMS Kenya were included in all analyses, as were children with multiple enteric pathogens detected in stool. To assess bias in this approach, two sensitivity analyses were performed. Models were re-run on the subset of case children (n = 1,483) who were only enrolled as a case once. We also compared Cryptosporidium-positive cases with a single pathogen to Cryptosporidium-positive cases with multiple enteric pathogens (S1 and S2 Tables).

Ethical review
Written informed consent was collected from all parents of children who participated in GEMS. The GEMS protocol was approved by the Scientific and Ethical Review Committees of KEMRI (Protocol #1155) and the Institutional Review Board (IRB) of the University of Maryland School of Medicine, Baltimore, MD, USA (UMD Protocol #H-28327). CDC (Atlanta, GA, USA) formally deferred its review to the UMD IRB (CDC Protocol #5038).

Demographic and household characteristics
Among the 1,778 MSD case children enrolled, Cryptosporidium was identified in 195 cases (11.0%). Cryptosporidium infections were more frequently identified in infants (<12 months old), with a peak in Cryptosporidium infection at 6-11 months old (Table 1 and Fig 1). Compared to case children aged 24-59 months, infants had over triple the odds of being Cryptosporidium-positive (OR = 3.32; 95% CI: 2.08-5.31). Other demographic and household characteristics were similar between Cryptosporidium-positive and Cryptosporidium-negative cases ( Table 1). A non-statistically significant relationship between Cryptosporidium status and having agricultural land was confounded by SES. As no other variable was confounded by SES, only unadjusted effect measures are shown in Table 1.

Clinical presentation and outcomes
The clinical presentation of Cryptosporidium-positive and Cryptosporidium-negative cases was similar ( Table 2). Mucus in the stool was significantly associated with being Cryptosporidiumpositive (OR = 1.72, 95% CI: 1.21-2.51). Only age category and mucus in stool remained in the final multivariable clinical model (not presented). The findings were the same when children who were enrolled multiple times as a case were excluded. Having mucus in the stool remained significantly associated with Cryptosporidium infection controlling for age (aOR = 1.50; 95% CI: 1.05-2.20).
Approximately two-thirds (66%) of Cryptosporidium-positive cases with daily information on diarrhea experienced prolonged or persistent diarrhea, compared to approximately half (51%) of Cryptosporidium-negative cases ( Table 2). Compared to cases experiencing acute diarrhea, cases experiencing prolonged diarrhea were significantly more likely to be Cryptosporidium-positive (OR = 1.68; 95% CI: 1.18-2.37); cases experiencing persistent diarrhea were also significantly more likely to be Cryptosporidium-positive compared to cases experiencing acute diarrhea (OR = 3.43; 95% CI: 1.97-5.98).

Indicators of malnutrition
At enrollment, sex was a significant effect modifier of the relationship between Cryptosporidium and stunting/severe stunting. Among girls, Cryptosporidium-positive cases had  (1) asked how many days the case child had been experiencing diarrhea in the week before presenting at the health facility and (2) instructed on how to record the presence/absence of diarrhea for 14 days following enrollment on a pre-piloted Memory Aid form. The total number of days of diarrhea was calculated as the sum of these durations, with a maximum of 21 possible diarrhea days, which may include multiple episodes. significantly greater odds of being stunted at baseline than Cryptosporidium-negative cases (OR = 1.82, 95% CI: 1.10-3.01). There were no other statistically significant differences in malnutrition indicators between Cryptosporidium-positive and Cryptosporidium-negative cases at enrollment (Table 3). At the 60-day follow-up, Cryptosporidium-positive cases had significantly greater odds of being stunted (aOR = 1.65, 95% CI: 1.06-2.57), underweight (aOR = 2.08, 95% CI: 1.34-3.22), or wasted (aOR = 2.04, 95% CI: 1.21-3.43) compared to Cryptosporidium-negative cases, controlling for baseline status for each measure (Table 3).
Cryptosporidium-positive cases had significantly larger negative changes in HAZ and WAZ measures from baseline to follow-up. When considering HAZ by sex, female Cryptosporidiumpositive cases had significantly larger negative changes in HAZ compared to female Cryptosporidium-negative cases (Table 3).

Hospitalizations and deaths
There was no significant difference in the proportion of Cryptosporidium-positive cases and Cryptosporidium-negative cases who were hospitalized at enrollment (13.3% vs. 10.5%, p = 0.24). Among those with 60-day follow-up information, 9 (4.8%) of 187 Cryptosporidiumpositive cases and 53 (3.5%) of 1,531 Cryptosporidium-negative cases died by the time of follow-up (p = 0.35). The cause of death, as per verbal autopsy, for the 9 children who died and had Cryptosporidium identified in their stool was as follows: HIV/AIDS related (n = 5), diarrhea/gastroenteritis (n = 1), pneumonia (n = 1), and anemia (n = 1); one child did not have a verbal autopsy completed.

Water, sanitation, and hygiene (WASH) characteristics
The most common caretaker-reported primary drinking water sources were rainwater (35%), surface water (31%), other improved water sources (23%), and other unimproved water sources (11%); (Table 5). Compared to cases in households that used rainwater as the primary source of drinking water, case children living in households using other improved sources or unimproved sources (other than surface water) had significantly higher odds of Cryptosporidium infection (OR = 1.72; 95% CI: 1.14-2.58 and 2.12; 95% CI: 1.31-3.41, respectively). Few caretakers of GEMS-Kenya case children reported boiling or using a ceramic filter to treat drinking water; however, those reporting one of these methods had significantly lower odds of Cryptosporidium infection compared to those who didn't (OR = 0.52, 95% CI: 0.25-0.96), predominantly driven by those who boiled water. Only two households reported filtering.
Reported handwashing behavior was similar among the caretakers of Cryptosporidium-positive and Cryptosporidium-negative cases (Table 5).
Only age category remained in the final multivariable demographic, environmental, and behavioral characteristics model; thus, this model is not presented. Cryptosporidium genotyping DNA was extracted from a random subset of 64 (40%) of the 160 Cryptosporidium-positive stool specimens from GEMS-Kenya case children enrolled in the first three years. Nested 18S PCR detected Cryptosporidium in 43 (67%) of these specimens. Of the 43 specimens, 35 (81%) were of the species C. hominis and 6 (14%) were C. parvum. C. meleagridis and C. canis were found in one specimen each. a. In the two weeks prior to enrollment; as reported by caretaker b. Other improved water sources: water piped into the house/yard, public taps, deep tube wells, covered wells, protected springs, or boreholes. Surface water sources: pond, lake, river, stream, dam, or earth pan/water pan. Other unimproved water sources: open wells, shallow tube wells, unprotected springs, and purchased water (such as bottled water). c.
Caretakers were asked what method they used most often when treating water at home d. Caretakers were asked to list times they usually wash their hands without probing from the questionnaire administrators https://doi.org/10.1371/journal.pntd.0006640.t005 Characteristics associated with Cryptosporidium infection in GEMS Kenya case children

Sensitivity analysis
Of the 195 Cryptosporidium-positive cases, 142 (72.8%) also had one or more additional enteric pathogens identified in their stool; enteric co-infections were common throughout the study population (Fig 2). The characteristics of case children with only Cryptosporidium detected in their stool were generally similar to Cryptosporidium-positive case children with multiple enteropathogens (S1 and S2 Tables); the only significant clinical difference was in the child's mental state at enrollment (S1 Table). Variables chosen for multivariable models were unchanged when excluding children who were enrolled more than once as an MSD case.

Discussion
This study evaluated the clinical, environmental, and behavioral characteristics associated with Cryptosporidium infection among children under five years old with MSD in rural western Kenya. Overall, 11% of children with MSD had Cryptosporidium identified in their stool; the majority (81%) of genotyped samples were C. hominis. Among MSD cases, being an infant, having mucus in stool, and having prolonged or persistent duration diarrhea were associated with being Cryptosporidium-positive. Boiling drinking water and using rainwater as the main drinking water source appeared to protect against Cryptosporidium infection in MSD cases. Among girls, Cryptosporidium-positive cases were more likely to be stunted at baseline compared to Cryptosporidium-negative cases. Cryptosporidium-positive cases had longer-term consequences in terms of malnutrition, as these children were more likely to stunted, underweight, or wasted at follow-up (controlling for baseline status), and have significantly larger negative changes in height-and weight-for-age z-scores. Except for having mucus in stool, which could be associated with Cryptosporidium adhering to the small intestine mucosa, possibly causing inflammation [29], the clinical presentation of children with MSD was similar for Cryptosporidium-positive and Cryptosporidium-negative cases, as was observed in another study of Kenyan children with diarrhea [16]. This finding highlights the difficulty in clinically diagnosing Cryptosporidium among children with MSD in this setting and underscores the need for point of care rapid diagnostics for Cryptosporidium.
Infants were over three times more likely to have Cryptosporidium identified in their stool compared with children aged 24-59 months. The peak of infection at 6-11 months in this study is similar to the age pattern of Cryptosporidium infections previously reported in sub-Saharan Africa, though an earlier peak than other studies in Kenya [9,16]. This timeframe may coincide with the introduction of complementary foods or drinking water. The high prevalence of Cryptosporidium infections in young children is concerning as Cryptosporidium infections in early childhood have been associated with numerous poor outcomes, sometimes lasting beyond the initial infection [6,7], as evident in our findings.
Prolonged and persistent duration diarrhea, and growth shortfalls subsequent to enrollment were significantly more pronounced among Cryptosporidium-positive cases compared to other children with MSD. Prolonged and persistent diarrheal episodes occurring in infants have been previously associated with growth shortfalls [30]. The proportion of Cryptosporidium-positive cases who were underweight and wasted increased from baseline to follow-up. This could result from many days of diarrhea experienced by these children. There was also an increase in the proportion of Cryptosporidium-positive cases who were stunted from baseline (29%) to follow-up (39%). Undernutrition and stunting among children in low-and middleincome countries have predicted decreased performance in school and on cognitive tests in previous research [31], thus even longer-term consequences could be appreciable although unexplored in the current study. It is estimated that growth faltering contributes substantially to the overall global burden of disease from Cryptosporidium infections in children [4]. By the time of follow-up, 4.8% of Cryptosporidium-positive and 3.5% of Cryptosporidium-negative cases had died. Although the difference was not statistically significant, this warrants close future attention since other research has shown an association between Cryptosporidium and excess mortality for children who became infected in infancy [5].
Like other studies in Kenya [9,16,32], our findings indicate that person-to-person transmission is likely the predominant route for Cryptosporidium infection in rural western Kenya, since the main host for C. hominis is humans [10]. Infections may thus more commonly result from exposure to human feces than animal feces. The presence of animals in the compound was examined in univariable analyses but was not reported in detail or included in the risk factor model selection, as the ownership of many types of animals was not associated with Cryptosporidium infection (S3 Table), though it may be associated with unmeasured confounders (e. g., higher income). Notably, in another study, C. hominis was associated with more severe clinical symptoms in Kenyan children compared to C. parvum [32], although we have too little data in GEMS Kenya to examine this.
Using rainwater as the main drinking water source was common and was significantly protective against Cryptosporidium infections. Rainwater may be less contaminated with Cryptosporidium, or this finding could be related to the seasonality of Cryptosporidium infections. The proportion of households using rainwater as the main drinking water source varied by month, ranging from 3%-74%. We did not explore Cryptosporidium infections by month/season, as the biweekly enrollment targets for GEMS make interpretation of pathogen-specific seasonal analyses challenging. However, the fact that using rainwater as the primary drinking water source and boiling drinking water were both protective against Cryptosporidium infections indicates that drinking water source choices and certain treatment options may be effective in reducing Cryptosporidium infections and signals that water may play a role in transmission.
A limitation of this work is that a comparison could not be made between Cryptosporidiumpositive GEMS cases and GEMS controls, as reliable population weights were not available at the time of analysis. The factors associated with Cryptosporidium compared to other individuals with MSD may be different from those risk factors that would be seen when compared to healthy controls. Using MSD as a condition for inclusion for our study may lead to spurious associations, as it is potentially a common effect of both the exposures and the outcome of interest [33]. Our ability to explore data on those with only Cryptosporidium infections was limited due to the small number of single-pathogen Cryptosporidium infections; however, those who presented with Cryptosporidium alone had similar characteristics to those who presented with multiple-pathogen Cryptosporidium infections. We were not able to assess the association between breastfeeding and Cryptosporidium because of (1) the collinearity between breastfeeding and age, and (2) the small number of children who were either exclusively or not breastfed in certain age groups. We also could not examine anthropometric outcomes by age or by the number of enteric pathogens isolated in stool, due to the small number of Cryptosporidium-positive children in some age groups and the small number of cases who presented with single-pathogen Cryptosporidium infections. We examined HIV status and malnutrition in our analyses, and performed a sensitivity analysis related to enteric co-infections; however, we did not have information on other co-morbidities that may be associated with Cryptosporidium infection.
The burden of diarrhea attributable to Cryptosporidium differed between GEMS and MAL-ED, especially for children 1-2 years old; however, GEMS generally considered more severe cases of diarrhea than MAL-ED. Other differences between the studies have been described elsewhere [34]. While GEMS and MAL-ED found Cryptosporidium to be significantly associated with diarrhea among infants, there were differences between study sites; in other studies Cryptosporidium has been isolated in non-diarrheal stools as often as in diarrheal stools [35,36]. Dissimilarities in study design (e.g., the time from diarrhea onset to stool collection) or laboratory methods may partially explain the differences observed, and host susceptibility and other risk factors are likely to vary across settings [35]. However, Cryptosporidium infections have been associated with growth shortfalls in asymptomatic children without diarrhea, thus identification and treatment of Cryptosporidium should remain a priority for young children in settings where it is endemic [35,36].
The high prevalence of cryptosporidiosis among young children in our study, coupled with other research that shows extended long-term effects of Cryptosporidium infections and diarrhea early in life, underscores the need for preventive measures aimed at households with young children, as well as improved diarrhea case management. Early diagnosis and management of cryptosporidiosis may mitigate subsequent growth deficits and other long-term consequences. Increased availability of nitazoxanide or new treatments, point of care rapid diagnostics for Cryptosporidium, additional insights into the role of appropriate WASH practices and technologies in childhood cryptosporidiosis, and vaccine development could reduce the burden of disease in such settings. Since Cryptosporidium-positive cases experienced more days of diarrhea and subsequent malnutrition than other MSD cases, increased promotion of the use of zinc in the management of diarrhea, and continued feeding of children with diarrhea should be undertaken, per WHO/UNICEF guidelines [37]. As rotavirus vaccine coverage increases, potentially leading to an altered enteric pathogen landscape, continuing to examine the impact and relative importance of Cryptosporidium infection among infants should remain a priority.

Disclaimer
The findings and conclusions in this report are those of the authors and do not necessarily represent the views of the U.S. Centers for Disease Control and Prevention.