Skip to main content
  • Loading metrics

The Interaction of Deworming, Improved Sanitation, and Household Flooring with Soil-Transmitted Helminth Infection in Rural Bangladesh

  • Jade Benjamin-Chung ,

    Affiliation School of Public Health, University of California, Berkeley, Berkeley, California, United States of America

  • Arifa Nazneen,

    Affiliation Centre for Communicable Diseases, International Centre for Diarrhoeal Disease Research, Bangladesh, Dhaka, Bangladesh

  • Amal K. Halder,

    Affiliation Centre for Communicable Diseases, International Centre for Diarrhoeal Disease Research, Bangladesh, Dhaka, Bangladesh

  • Rashidul Haque,

    Affiliation Centre for Communicable Diseases, International Centre for Diarrhoeal Disease Research, Bangladesh, Dhaka, Bangladesh

  • Abdullah Siddique,

    Affiliation Centre for Communicable Diseases, International Centre for Diarrhoeal Disease Research, Bangladesh, Dhaka, Bangladesh

  • Muhammed Salah Uddin,

    Affiliation Centre for Communicable Diseases, International Centre for Diarrhoeal Disease Research, Bangladesh, Dhaka, Bangladesh

  • Kim Koporc,

    Affiliation Children Without Worms, Task Force for Global Health, Atlanta, Georgia, United States of America

  • Benjamin F. Arnold,

    Affiliation School of Public Health, University of California, Berkeley, Berkeley, California, United States of America

  • Alan E. Hubbard,

    Affiliation School of Public Health, University of California, Berkeley, Berkeley, California, United States of America

  • Leanne Unicomb,

    Affiliation Centre for Communicable Diseases, International Centre for Diarrhoeal Disease Research, Bangladesh, Dhaka, Bangladesh

  • Stephen P. Luby,

    Affiliation Department of Medicine, Stanford University, Stanford, California, United States of America

  • David G. Addiss,

    Affiliation Children Without Worms, Task Force for Global Health, Atlanta, Georgia, United States of America

  • John M. Colford Jr.

    Affiliation School of Public Health, University of California, Berkeley, Berkeley, California, United States of America



The combination of deworming and improved sanitation or hygiene may result in greater reductions in soil-transmitted helminth (STH) infection than any single intervention on its own. We measured STH prevalence in rural Bangladesh and assessed potential interactions among deworming, hygienic latrines, and household finished floors.


We conducted a cross-sectional survey (n = 1,630) in 100 villages in rural Bangladesh to measure three exposures: self-reported deworming consumption in the past 6 months, access to a hygienic latrine, and household flooring material. We collected stool samples from children 1–4 years, 5–12 years, and women 15–49 years. We performed mini-FLOTAC on preserved stool samples to detect Ascaris lumbricoides, Enterobius vermicularis, hookworm, and Trichuris trichiura ova. Approximately one-third (32%) of all individuals and 40% of school-aged children had an STH infection. Less than 2% of the sample had moderate/heavy intensity infections. Deworming was associated with lower Ascaris prevalence (adjusted prevalence ratio (PR) = 0.53; 95% CI 0.40, 0.71), but there was no significant association with hookworm (PR = 0.93, 95% CI 0.60, 1.44) or Trichuris (PR = 0.90, 95% CI 0.74, 1.08). PRs for hygienic latrine access were 0.91 (95% CI 0.67,1.24), 0.73 (95% CI 0.43,1.24), and 1.03 (95% CI 0.84,1.27) for Ascaris, hookworm, and Trichuris, respectively. Finished floors were associated with lower Ascaris prevalence (PR = 0.56, 95% CI 0.32, 0.97) but not associated with hookworm (PR = 0.48 95% CI 0.16,1.45) or Trichuris (PR = 0.98, 95% CI 0.72,1.33). Across helminths and combinations of exposures, adjusted prevalence ratios for joint exposures were consistently more protective than those for individual exposures.


We found moderate STH prevalence in rural Bangladesh among children and women of childbearing age. This study is one of the first to examine independent and combined associations with deworming, sanitation, and hygiene. Our results suggest that coupling deworming with sanitation and flooring interventions may yield more sustained reductions in STH prevalence.

Author Summary

Soil-transmitted helminth infections remain prevalent in many low-resource areas of the world. The World Health Organization recommends that schoolchildren in countries where these infections remain common receive deworming medication two times a year. However, previous research has shown that people who live in countries where these infections are common are frequently reinfected within 6 months of taking deworming medication. Programs that improve sanitation and hygiene might help complement deworming programs to reduce reinfection and prevent transmission. We conducted a survey of women and children in rural Bangladesh to understand potential sanitation and hygiene interventions that could complement deworming. We found that people who took deworming medication and had access to a hygienic latrine had a lower worm infection prevalence than people who only took deworming medication. We also found that people who took deworming medication and had a house with a finished floor had a lower prevalence than people who only took deworming medication. Our results suggest that coupling deworming with sanitation and flooring interventions may be a more successful strategy for reducing STH transmission in the long run.


The World Health Organization recommends mass drug administration of albendazole or mebendazole to school-aged children to control soil-transmitted helminths in endemic countries; in addition, they recommend improved sanitation and hygiene to ensure long-term sustainability of deworming efforts [1]. While anthelminthics are highly efficacious in the short term, it is estimated that within six months, on average, 68% of those treated become reinfected with Ascaris, 67% with Trichuris, and 55% with hookworm [2]. Improvements to sanitation [38] and installation of finished flooring in homes may contribute to more sustainable reductions in STH transmission and prevalence either when delivered alone or in combination with deworming. However, to date, STH control has largely been a separate enterprise from the control of enteric pathogens through sanitation and hygiene interventions [9]. There has been a call to consider the joint effects of anthelminthic treatment and water, sanitation, and hygiene interventions [7,9], yet few studies have done so rigorously [1013]. Considering the increased concerns about the potential for resistance to anthelminthic drugs [14,15], it is increasingly important to identify interventions that may more sustainably reduce transmission and minimize the duration of mass drug administration campaigns. Evidence that the combination of deworming and improved household environmental conditions results in greater risk reductions than either alone (i.e., evidence of a synergistic interaction) would motivate the delivery of combined interventions to more sustainably reduce the incidence and transmission of STH. Furthermore, such evidence could inform the targeting of deworming interventions to households with environmental conditions in which deworming is most effective.

Several studies have suggested that improved sanitation can reduce the risk of STH infection by reducing the shedding of helminth ova into the environment, leading to reduced transmission [7,1618]. Another potential intervention that may reduce transmission is the provision of finished floors (e.g., cement or wood floors). Since STH eggs must be deposited in the soil to reach their infective stages, providing finished floors to households to replace earthen floors may reduce transmission by reducing the number of infective stages in the living space. The few studies that have systematically explored whether finished flooring is associated with STH infection found evidence of reduced risk but did not adjust for potentially strong confounders, such as household wealth [1922]. To date, no studies have formally explored interactions between water, sanitation, and flooring interventions and deworming [23].

In Bangladesh, the Ministry of Health and Family Welfare has implemented mass drug administration of mebendazole to control STH infection in schools in 27 out of 64 districts bi-annually since 2008. In addition, the Bangladesh Expanded Program on Immunization offers mebendazole to pre-school children in Bangladesh. To control lymphatic filariasis infection, the Ministry has offered albendazole and diethylcarbamazine in endemic areas each year to all individuals over 1 year of age who are not pregnant since 2001; as of 2008, the program has operated in 20 districts. Prior to mass drug administration, an estimated 80% of Bangladeshi school-age children in rural areas were infected with STH [24]. To our knowledge, there have not been any systematic surveys of STH prevalence since mass drug administration began in Bangladesh.

Our objectives were to estimate the prevalence of STH infection among children and women of childbearing age in rural Bangladesh and to estimate the separate and combined associations of deworming, hygienic latrines, and finished floors with STH prevalence.


Study population and sample

In this study, we collected stool samples from a subset of individuals who participated in an ongoing cross-sectional study led by the International Centre for Diarrhoeal Disease Research, Bangladesh (ICDDR,B) in rural Bangladesh and conducted a secondary analysis of survey data from that study. The original study evaluated the Sanitation Hygiene Education and Water Supply in Bangladesh (SHEWA-B) program, which was implemented by UNICEF and the Government of Bangladesh from 2007–2012. Local hygiene promoters visited mothers of children under five years old in underserved areas of rural Bangladesh and delivered key messages about safe water, sanitation, and hygiene practices. The program did not offer deworming or improvements to household flooring. In a small subset of villages with high poverty levels, the program offered subsidized latrines [25]. The data used in this cross-sectional study were collected in 2012 in 68 sub-districts, 19 districts, and 50 intervention and 50 control village clusters in rural Bangladesh to evaluate SHEWA-B in 2012. The intervention, selection of control areas, eligibility, and sampling of clusters have been described elsewhere [26]. In this study, we did not stratify by whether a respondent participated in SHEWA-B or lived in a control cluster because the impact evaluation of SHEWA-B found that there was no increase in access to improved latrines, safe disposal of feces, availability of a handwashing station, or safe drinking water storage among SHEWA-B participants [25]. Thus, we considered it unlikely that SHEWA-B participation would be a potential confounder of the association between our exposures of interest and STH infection. The field team collected stool samples and questionnaires about socio-demographic information and anthelminthic treatment in October 2012. In December 2012, they administered a questionnaire to the same households to measure access to hygienic latrines, finished floors, and other environmental exposures. Latrine and flooring status were ascertained following stool sample collection due to field logistics constraints and the need to complete stool collection prior to national mass drug administration in early November 2012.

Stool specimen collection and analysis

In each cluster, stool was collected from individuals in 17 randomly selected households; this number was determined by the sample size calculations from the original SHEWA-B evaluation. Field workers aimed to stratify sample collection by age such that six people within each of the following age and sex categories provided stool in each cluster: children 1–4 years, children 5–14 years, and women 15–49 years. We stratified by these groupings because in Bangladesh pre-school children (1–4 years) are offered deworming through the Expanded Program on Immunization, and school-aged children (5–14 years) are offered deworming through a separate program administered by the Ministry of Health. We collected stool from women of childbearing age because our exposure assessment focused on household-level exposures, and in rural Bangladesh, women in this age group spend most of their time at home, so these exposures are more relevant for women than men. In addition, women of childbearing age are also more likely to be exposed to STH ova shed by young children than older women.

Field workers provided households with plastic sheets and stool collection tubes and returned within 24 hours to collect samples. They stored 1g of stool in 20 ml of 4% sodium acetate-acetic acid-formalin. The maximum time between defecation and stool processing was 12 hours. Samples were transported to Dhaka, Bangladesh, for laboratory analysis at the International Centre for Diarrhoeal Disease Research, Bangladesh.

Helminth ova were detected using mini-FLOTAC, a copromicroscopic diagnostic technique appropriate for preserved stool [27,28]. Laboratory staff centrifuged samples at 1500 RPM for 3 minutes and then discarded the supernatant and suspended the sedimented stool in 20 ml of flotation solution 2 (saturated sodium chloride). They then mixed the contents thoroughly and filled each of the two chambers of the mini-FLOTAC device with 1 ml of the mixed sample. Staff recorded the number of eggs of Ascaris lumbricoides, hookworm, Trichuris trichiura, and Enterobius vermicularis in each chamber. For each helminth, we averaged the number of eggs in each chamber and multiplied the number by a factor of 10 to quantify the number of eggs per gram of stool. Laboratory analyses were conducted within 9 months of stool sample collection.

Outcome and exposure definitions

Outcomes included the presence of any helminth ova as well as the intensity of helminth infection. Moderate/high intensity infections were defined as ≥5,000 eggs/gram for Ascaris, ≥1,000 eggs/gram for Trichuris, and ≥2,000 eggs/gram for hookworm [29]. Exposures included access to a hygienic latrine, household flooring material (earth/bamboo or cement/wood), and self-reported deworming in the last six months. We defined hygienic latrines as flush latrines connected to a piped sewer system, septic tank, off-set pit latrine, pit latrine with slab and functional water seal, pit latrine with slab, lid and no water seal, or a composting latrine. We defined “unhygienic” latrines as those that would likely fail to separate feces from the environment effectively including flush latrines connected to canal or ditch, pit latrines with or without a slab, no or broken water seal or a hanging latrine. This definition was developed by the International Centre for Diarrhoeal Disease Research, Bangladesh (ICDDR,B) and is intended to more accurately categorize latrines that isolate feces from the environment in the Bangladeshi context than the commonly used World Health Organization Joint Monitoring Programme (JMP) definition [30] (see S1 Table). Respondents reported whether each person who provided a stool sample took deworming medication in the last six months. If so, they were asked approximately how many weeks or months ago they took deworming. Field workers recorded whether deworming was received as part of a campaign and the source of deworming (e.g. clinic, school).

We calculated the cluster-level deworming coverage as the percentage of respondents in the sample who reported taking deworming in the prior six months in a given cluster. To estimate cluster-level sanitation and finished floor coverage, we calculated the percentage of respondents with a hygienic latrine, finished household floor, or who reported deworming consumption in a cluster.

We controlled for the following potential confounders in statistical models: age, sex, household wealth, cluster-level wealth, mother's education level, and district of residence. We used information about household assets (e.g. refrigerator, mobile phone) to develop an index of household wealth using principal components analysis [31] (see S2 Table). Households in the lowest three quintiles of the first principal component were classified as lower household wealth and those in the highest two quintiles were classified as higher household wealth. Cluster-level wealth was calculated as the percentage of households in the two highest quintiles of household wealth.

Sample size

Since age- and sex-specific STH prevalence estimates were not available for the study areas, we assumed the prevalence of all helminths to be 50%. We assumed a design effect of 2.6 based on intraclass correlation coefficients estimated in a study measuring STH infection in a similar population in India [32] since such information for Bangladesh was not available. Because our study was nested within the ongoing SHEWA-B evaluation, our calculations assumed a fixed sample size of 1,700 (100 village clusters x 17 individuals per cluster). Under these assumptions, the precision associated with an estimate of prevalence of 50% is ±4%.

Statistical analysis

We calculated pooled and age- and sex-specific prevalence by species of helminth. To examine the association between prevalence and cluster-level variables, we produced scatter plots of the observed variables and used locally weighted scatter plot smoothing (LOWESS) with normal-based pointwise 95% confidence bands to explore patterns in each scatter plot [33]. We also estimated the intraclass correlation coefficient for each STH infection within each cluster using a one-way analysis of variance.

To estimate adjusted prevalence ratios we used modified Poisson regression [34] adjusting for the potential confounders defined above. For each of the three exposures, we adjusted for the other two exposures in each model (e.g., the models for deworming were adjusted for hygienic latrines and finished floors). We estimated robust standard errors clustered at the village level to account for potential within-village outcome correlation. We excluded individuals with missing outcomes from the analysis, which assumes that they were missing at random conditional on covariates in our model.

Standard statistical models for binary outcomes predict outcomes and assess statistical interaction on the multiplicative scale. In this study, we chose to estimate interaction on the additive scale, which is useful when one is interested in the extent to which a primary exposure may yield greater health improvements by introducing a secondary exposure. In our study population, we consider deworming to be a primary exposure due to ongoing school-based deworming activities in Bangladesh. Our approach allows us to assess whether secondary exposures in addition to deworming, such as access to hygienic latrines, were associated with lower STH prevalence than deworming alone [35]. Specifically, we estimated the relative excess risk due to interaction (RERI), a measure of additive interaction [36]. When the exposures of interest are associated with only a lower or higher prevalence, an RERI>0 indicates a synergistic interaction between exposures [37,38]. If exposures could be associated with either increased or decreased prevalence, then the RERI must be greater than 1 for synergistic interaction to be present [37,38]. Since we expected associations to be protective, we recoded variables prior to RERI calculation so that the stratum with the prevalence ratio furthest from the null was reassigned as the reference group [39]. Because data were clustered at the village level, we used the bootstrap and resampled clusters to estimate 95% confidence intervals [40,41]. We did not estimate confidence intervals for any point estimates for which there were strata with fewer than 5 units. Analyses were conducted in Stata version 12 and in R version 3.1.3.

Ethics statement

This study was approved by the Institutional Review Board (IRB) at the Centers for Disease Control and Prevention (CDC) (Protocol #6061). Study participants provided written consent. The CDC IRB approved this consent procedure. We offered a single dose of mebendazole to all participants who provided a stool sample.


The field team collected stool samples and demographic information from 1,795 individuals in October 2012 and collected exposure information from 1,655 individuals in December 2012 (Fig 1). There were 1,630 individuals in the complete dataset; 140 individuals were not home during follow-up, and 25 identification numbers were mismatched between survey rounds. The number of missing observations for each variable of interest is listed in S3 Table.

Fig 1. Data collected.

This figure shows the number of surveys returned, samples analyzed, and final sample size for this analysis.

Less than half (40%, n = 656) of mothers of the youngest child in each household had received at least a primary education. About a third of households (32%, n = 527) had access to a hygienic latrine, and 13% (n = 207) of households had finished floors. Respondents reported that 49% of children 1–4 years, 52% of children 5–14 years, and 21% of women of childbearing age took deworming medication in the prior six months. Child caregivers reported that less than half (47%) of school-age children took deworming drugs at school.

Approximately one-third (32%) of individuals sampled had an STH infection, and 9% had multiple infections (Table 1). Across all age groups, Trichuris was most prevalent, infecting 17% of children 1–4 years, 28% of children 5–14 years, and 18% of women of childbearing age. For all helminths and age groups, fewer than 2% had moderate/heavy intensity Ascaris or Trichuris infections; there were no moderate/heavy intensity hookworm infections. Prevalence of Ascaris and Trichuris were highest in areas in Dhaka and northern Chittagong divisions (Fig 2).

Fig 2. Map of soil-transmitted helminth prevalence in Bangladesh.

We mapped STH prevalence in each study cluster for which valid GPS coordinates were available (n = 99). Panel A shows the cluster-level prevalence of any STH infection, Panel B shows the prevalence of Ascaris lumbricoides, Panel C shows the prevalence of hookworm, and Panel D shows the prevalence of Trichuris trichiura.

Table 1. Helminth infection prevalence by age, helminth, and infection intensity.

We estimated associations between individual exposures and STH infection (Table 2). Deworming was associated with 47% lower Ascaris prevalence (95% CI 29%, 60%); adjusted associations with hookworm and Trichuris were null. There were no statistically significant associations with hygienic latrine access for any helminth. Finished floors were associated with 44% lower Ascaris prevalence (95% CI 3%, 68%). There was no statistically significant association between finished floors and hookworm or Trichuris. Participation in SHEWA-B was not associated with STH infection (S4 Table).

Table 2. Prevalence ratios for deworming, hygienic latrine access, and finished floors.

To explore potential interactions among deworming, hygienic latrines, and finished floors, we estimated adjusted prevalence ratios for each separately and in combination and the relative excess risk due to interaction RERI (Tables 3 and 4). The combination of deworming and hygienic latrines was associated with lower Ascaris and Trichuris prevalence than associations with separate exposures, although this association was only statistically significant for Ascaris (Table 3, Panel A; Fig 3). For example, deworming without hygienic latrine access (denoted D+L- in Fig 3) was associated with 45% lower Ascaris prevalence (95% CI 23%, 60%). Ascaris prevalence was equivalent for those with hygienic latrine access who did not take deworming (D-L+). The combination of deworming and hygienic latrine access (D+L+) was associated with 59% lower Ascaris prevalence (95% CI 27%, 76%). While individual and combined exposures were associated with lower hookworm prevalence, the associations did not follow the same pattern and were not statistically significant.

Fig 3. STH prevalence by exposure to deworming, hygienic latrines, and finished floors.

Panel A shows the prevalence of each helminth among those who did not receive deworming (D-), those who did receive deworming (D+), those who did not have access to a hygienic latrine (L-), and those who did have access to a hygienic latrine (L+). Panel B shows the prevalence of each helminth among those who did not receive deworming (D-), those who did receive deworming (D+), those whose household did not have a finished floor (F-), and those whose household did have a finished floor (F+).

Table 3. Adjusted prevalence ratios for single and combined deworming and hygienic latrine access and relative excess risk due to interaction.

Table 4. Adjusted prevalence ratios for single and combined deworming and household finished floor and relative excess risk due to interaction.

Our estimates of the RERI for Ascaris and Trichuris were less than zero and not statistically significant, indicating no clear evidence of synergistic interaction between deworming and hygienic latrines on the additive scale. However, there was some suggestion of synergistic interaction for hookworm (RERI = 0.49; 95%CI -0.73, 1.16), although the confidence interval included 0.

For the combination of deworming and finished floors, we found the same pattern for all three helminths: the combination was associated with a lower prevalence than either exposure on its own (Table 3, Panel B; Fig 3). Deworming without finished floors (denoted D+F- in Fig 3) was associated with 47% lower Ascaris prevalence (95%CI 29%, 61%). Finished floors without deworming (D-F+) was associated with 40% lower Ascaris prevalence (95%CI -15%, 68%). The combination of deworming and finished floors (D+F+) was associated with 72% lower Ascaris prevalence (95%CI 89%, 25%). For Ascaris, the RERI was 0.56 (95%CI -3.64, 2.40). This pattern was also evident for hookworm and Trichuris.

We plotted the cluster-level prevalence of each helminth across the observed range of cluster-level deworming, hygienic latrine, and finished floor coverage (Figs 2 and S1 and S2). As hygienic latrine coverage increased, Ascaris and hookworm prevalence remained approximately the same and Trichuris prevalence increased slightly (Fig 4). As deworming and finished floor coverage increased, there was no substantial change in the prevalence of any helminth (S1 and S2 Figs). The village cluster level intraclass correlation coefficients were 0.11 (95% CI 0.08, 0.16) for Ascaris, 0.02 (95% CI 0.00, 0.05) for hookworm, and 0.21 (95% CI 0.15, 0.27) for Trichuris.

Fig 4. Cluster-level STH prevalence by cluster-level hygienic latrine coverage.

Panel A shows the cluster-level prevalence of Ascaris, Panel B shows the prevalence of hookworm, and Panel C shows the prevalence of Trichuris by the proportion of respondents with hygienic latrines in each cluster.


In this cross-sectional study of rural, low-income households in Bangladesh, where biannual national school-based deworming has been implemented for the last five years, the prevalence among school-aged children was 14% for Ascaris, 8% for hookworm, and 21% for Trichuris. Forty percent of school-aged children had at least one STH infection, and 12% had multiple infections. Approximately half the children under 15 years were reported to have taken deworming drugs in the prior six months. When we compared associations between STH infection and single versus combined exposures, we found that for deworming and hygienic latrines, Ascaris and Trichuris prevalence was lower among those with both exposures compared to those with only one exposure. Among those exposed to both deworming and finished floors, we found a lower prevalence for the combination of the two exposures than for either exposure separately. While our measures of additive interaction (the RERI) did not consistently indicate synergy and were not statistically significant, we consider the pattern of lower prevalence for combined exposures, particularly for deworming and finished floors, worthy of further exploration.

Comparison of our prevalence estimates to prior estimates

The prevalence of any STH among school-aged children was 40% compared to 80% in rural areas reported by the Ministry of Health and Family Welfare prior to the initiation of school-based deworming [24]. The prevalence we observed is consistent with studies of school-based deworming with one-year follow-up and very high coverage [4244]. Transmission theory [4547] and empirical findings [48,49] suggest that prevalence decreases in pre-school age and adult populations when deworming coverage is high. We found that 26% of children 1–4 years and 30% of women of childbearing age had an STH infection. The similar prevalence in these two groups to the prevalence among school-aged children may suggest that, despite low levels of intensity, transmission is still ongoing in those population subgroups.

Associations between STH and individual exposures

We observed associations between STH infection and deworming, finished floors, and hygienic latrines that were generally consistent with existing studies.


We found that self-reported deworming was associated with a lower prevalence of Ascaris and hookworm, but the association was only significant for Ascaris. There was no association with Trichuris prevalence. These findings are consistent with those of randomized controlled trials estimating the efficacy of single-dose mebendazole and albendazole [5052].

Hygienic latrines.

Among those exposed to hygienic latrines, we found protective but not statistically significant associations with Ascaris and hookworm, but there was no association with Trichuris. We also examined associations between STH infection and access to any type of latrine and found similar results. These findings differ from those reported in two meta-analyses led by Ziegelbauer et al. and Strunz et al. examining sanitation and STH infection [7,53]. Ziegelbauer et al. identified 36 studies evaluating the association between latrine access and STH infection. They found protective associations for all three helminths for access to any type of latrine and for latrine usage [7]. The meta-analysis found evidence of publication bias, which is one possible explanation for our differing findings. Strunz et al. conducted a similar systematic review of water, sanitation, hygiene, and STH infection. They also found statistically significant protective associations between access to latrines of any kind and Ascaris and Trichuris but did not find an association with hookworm [53].

Finished floors.

Among those exposed to finished floors, we observed a strong protective association with Ascaris, a protective association with hookworm that was not significant, and no association with Trichuris. The existing literature has for the most part reported that prevalence is lower for these helminths among those with finished floors; however, many previous studies did not adjust for household wealth, an important potential confounder of this association [1922,5456]. In this analysis, prevalence ratios for finished floors, adjusted for household wealth, were closer to the null than unadjusted estimates for each helminth, and for hookworm the association was no longer statistically significant following adjustment.

The null associations for all three individual exposures with Trichuris, the most prevalent helminth in this population, are noteworthy. There is some evidence that the prevalence of Trichuris decreases more slowly than that of other STH and that reinfection with Trichuris occurs more rapidly following intervention [5759]. This may be because of longer survival of adult worms or because Trichuris has a higher reproductive rate than other STH [57,58]. The higher observed prevalence of Trichuris compared to Ascaris and hookworm in this study likely reflects these parasite-specific differences in biology and response to intervention.

Associations between cluster-level exposures and STH prevalence

Contrary to what we would expect based on modeling studies [4547], we did not find evidence of an association between STH prevalence and coverage of exposures at the cluster level. We expected that prevalence would decrease as the coverage of deworming, hygienic latrines, and finished floors increased. Such a pattern could be explained by herd effects of these exposures, which result from decreased shedding of infective stages in feces into the environment and reduced transmission [47]. Interestingly, our estimates of intraclass correlation coefficients suggest that cluster membership accounts for 25% of Trichuris prevalence and 11% of Ascaris prevalence; these findings are similar to the intraclass correlation coefficients reported in the literature [60]. Thus, village membership appears to be an important predictor of STH infection despite the lack of associations with cluster-level exposure coverage. It is possible that we did not find cluster-level associations because herd effects were small or absent for these exposures. It is also possible that such herd effects might have been higher if STH prevalence were higher. Alternatively, because we only collected data for 16 households per cluster on average, our sample might not have been sufficient to characterize cluster-level prevalence patterns. Additionally, variation in local population density in our sample could explain the lack of association because herd effects are likely to be stronger in densely populated areas. We did not account for population density in this analysis, and further work is needed to explore its role as a potential effect modifier.


This study is subject to several limitations. First, deworming consumption was self-reported over a six-month period. There is evidence that recall of medication consumption is under-reported with longer recall periods [61]. It is also possible that reporting was subject to courtesy bias so that deworming was over-reported. Given that it is unlikely that respondents knew if they had STH infection at the time of deworming reporting, we posit that it is unlikely that misclassification of deworming consumption differed by STH infection status. If non-differential misclassification occurred, it would bias point estimates towards the null [62]. It is also possible that individuals who were dewormed in the prior six months were reinfected prior to stool collection in this study. For this reason and because of the possible misclassification of deworming due to poor recall, the associations between deworming and STH infection do not necessarily measure the reduction in STH attributable directly to deworming.

Second, on average, stool samples were stored for 6 months prior to mini-FLOTAC analysis (min = 4, max = 8 months). This long storage period may have resulted in degradation of ova and underestimated prevalence, particularly for hookworm [63]. The sensitivity of mini-FLOTAC is estimated to be 75.5% for Ascaris, 79.2% for hookworm, and 76.2% for Trichuris [64]. Thus, given the sensitivity of the assay and the long storage period prior to analysis, our prevalence estimates are likely a lower bound of the true values. In addition, the sample size was powered to estimate prevalence but not to estimate interactions between exposures. Thus, our estimates of the RERI were in most cases underpowered, particularly for improved floors, which were relatively rare in this population.

Third, exposure measurement occurred following outcome measurement. This is suboptimal because it is possible that the outcome status of an individual would trigger a change in exposure status, leading to reverse causation. However, for the exposures measured—access to a hygienic latrine and finished floors—we consider it highly unlikely that the respondents' exposure status changed in the two months following outcome measurement. Furthermore, respondents were likely unaware of their outcome status throughout the study, so it is unlikely that they changed their sanitation infrastructure or flooring material because of their outcome status. Fourth, our analysis assumes that the sanitation and flooring infrastructure we observed were present within the past six months (the recall window for deworming consumption). We consider this to be a reasonable assumption; however, if these exposures were misclassified, prevalence ratios would be biased towards the null and the effects on measures of interaction would be unpredictable [65].


This study provides the first estimates of STH prevalence following the initiation of mass drug administration for STH control among children and women of childbearing age in representative rural areas of Bangladesh. Our results suggest that combining deworming with sanitation and flooring interventions may yield greater reductions in STH prevalence than deworming alone.

Supporting Information

S1 Fig. Cluster-level STH prevalence by cluster-level deworming floor coverage.

Panel A shows the cluster-level prevalence of Ascaris, Panel B shows the prevalence of hookworm, and Panel C shows the prevalence of Trichuris by the proportion of respondents who took deworming in the past six months in each cluster.


S2 Fig. Cluster-level STH prevalence by cluster-level finished floor coverage.

Panel A shows the cluster-level prevalence of Ascaris, Panel B shows the prevalence of hookworm, and Panel C shows the prevalence of Trichuris by the proportion of respondents with finished floors in each cluster.


S1 Table. Prevalence ratios for improved vs. hygienic latrine access.


S2 Table. Means of each variable used in the principal components analysis by quintile of the index.


S3 Table. Missing observations for outcome, exposure, and confounder variables.


S4 Table. Association between SHEWA-B participation and STH infection.



The authors would like to thank Davide Ianeillo and Andrew Majewski for providing training to the ICDDR,B laboratory team on mini-FLOTAC and the University of Naples for providing mini-FLOTAC devices.

Author Contributions

Conceived and designed the experiments: JBC BFA JMC SPL DGA KK LU. Performed the experiments: AKH AN AS RH MSU. Analyzed the data: JBC. Contributed reagents/materials/analysis tools: AEH RH. Wrote the paper: JBC.


  1. 1. World Health Organization. Deworming for health and development [Internet]. Geneva, Switzerland; 2004.
  2. 2. Jia T-W, Melville S, Utzinger J, King CH, Zhou X-N. Soil-Transmitted Helminth Reinfection after Drug Treatment: A Systematic Review and Meta-Analysis. PLoS Negl Trop Dis. 2012;6: e1621. pmid:22590656
  3. 3. Utzinger J, Raso G, Brooker S, De Savigny D, Tanner M, Ornbjerg N, et al. Schistosomiasis and neglected tropical diseases: towards integrated and sustainable control and a word of caution. Parasitology. 2009;136: 1859–1874. pmid:19906318
  4. 4. Spiegel JM, Dharamsi S, Wasan KM, Yassi A, Singer B, Hotez PJ, et al. Which New Approaches to Tackling Neglected Tropical Diseases Show Promise? PLoS Med. 2010;7: e1000255. pmid:20502599
  5. 5. Bartram J, Cairncross S. Hygiene, Sanitation, and Water: Forgotten Foundations of Health. PLoS Med. 2010;7: e1000367. pmid:21085694
  6. 6. Asaolu S., Ofoezie I. The role of health education and sanitation in the control of helminth infections. Acta Trop. 2003;86: 283–294. pmid:12745145
  7. 7. Ziegelbauer K, Speich B, Mäusezahl D, Bos R, Keiser J, Utzinger J. Effect of Sanitation on Soil-Transmitted Helminth Infection: Systematic Review and Meta-Analysis. PLoS Med. 2012;9: e1001162. pmid:22291577
  8. 8. Campbell SJ, Savage GB, Gray DJ, Atkinson J-AM, Soares Magalhães RJ, Nery SV, et al. Water, Sanitation, and Hygiene (WASH): A Critical Component for Sustainable Soil-Transmitted Helminth and Schistosomiasis Control. PLoS Negl Trop Dis. 2014;8: e2651. pmid:24722335
  9. 9. Freeman MC, Ogden S, Jacobson J, Abbott D, Addiss DG, Amnie AG, et al. Integration of Water, Sanitation, and Hygiene for the Prevention and Control of Neglected Tropical Diseases: A Rationale for Inter-Sectoral Collaboration. PLoS Negl Trop Dis. 2013;7: e2439. pmid:24086781
  10. 10. Freeman MC, Greene LE, Dreibelbis R, Saboori S, Muga R, Brumback B, et al. Assessing the impact of a school-based water treatment, hygiene and sanitation programme on pupil absence in Nyanza Province, Kenya: a cluster-randomized trial. Trop Med Int Health. 2012;17: 380–391. pmid:22175695
  11. 11. Hesham Al-Mekhlafi M, Surin J, Atiya AS, Ariffin WA, Mohammed Mahdy AK, Che Abdullah H. Pattern and predictors of soil-transmitted helminth reinfection among aboriginal schoolchildren in rural Peninsular Malaysia. Acta Trop. 2008;107: 200–204. pmid:18582430
  12. 12. Henry FJ. Reinfection with Ascaris lumbricoides after chemotherapy: a comparative study in three villages with varying sanitation. Trans R Soc Trop Med Hyg. 1988;82: 460–464. pmid:3232186
  13. 13. Cort WW, Schapiro L, Stoll NR. A Study of Reinfection After Treatment with Hookworm and Ascaris in Two Villages in Panama. Am J Epidemiol. 1929;10: 614–625.
  14. 14. Albonico M, Engels D, Savioli L. Monitoring drug efficacy and early detection of drug resistance in human soil-transmitted nematodes: a pressing public health agenda for helminth control. Int J Parasitol. 2004;34: 1205–1210. pmid:15491582
  15. 15. Geerts S, Gryseels B. Anthelmintic resistance in human helminths: a review. Trop Med Int Health. 2001;6: 915–921. pmid:11703846
  16. 16. Barreto M, Genser B, Strina A, Teixeira MG, Assis AMO, Rego RF, et al. Impact of diarrhea of a city-wide sanitation programme in Norhteast Brazil. Inst Saude Coletiva Fed Univ Bahia. 2005.
  17. 17. Mascarini-Serra LM, Telles CA, Prado MS, Mattos SA, Strina A, Alcantara-Neves NM, et al. Reductions in the Prevalence and Incidence of Geohelminth Infections following a City-wide Sanitation Program in a Brazilian Urban Centre. PLoS Negl Trop Dis. 2010;4: e588. pmid:20126396
  18. 18. Freeman MC, Clasen T, Brooker SJ, Akoko DO, Rheingans R. The Impact of a School-Based Hygiene, Water Quality and Sanitation Intervention on Soil-Transmitted Helminth Reinfection: A Cluster-Randomized Trial. Am J Trop Med Hyg. 2013;89: 875–883. pmid:24019429
  19. 19. Faust EC, Giraldo LE. Parasitological surveys in Cali, Departamento del Valle, Colombia VI. Strongyloidiasis in Barrio Siloé, Cali, Colombia. Trans R Soc Trop Med Hyg. 1960;54: 556–563. pmid:13698369
  20. 20. Hall A, Conway DJ, Anwar KS, Rahman ML. Strongyloides stercoralis in an urban slum community in Bangladesh: factors independently associated with infection. Trans R Soc Trop Med Hyg. 1994;88: 527–530. pmid:7992327
  21. 21. Walker M, Hall A, Basáñez M-G. Individual predisposition, household clustering and risk factors for human infection with Ascaris lumbricoides: new epidemiological insights. PLoS Negl Trop Dis. 2011;5: e1047. pmid:21541362
  22. 22. Narain K, Rajguru SK, Mahanta J. Prevalence of Trichuris trichiura in relation to socio-economic & behavioural determinants of exposure to infection in rural Assam. Indian J Med Res. 2000;112: 140–146. pmid:11200680
  23. 23. Blot WJ, Day NE. Synergism and Interaction: Are They Equivalent? Am J Epidemiol. 1979;110: 99–100. pmid:463868
  24. 24. Ministry of Health & Family Welfare. A Situation Analysis: Neglected Tropical Diseases in Bangladesh [Internet]. Government of Bangladesh; 2010.
  25. 25. Unicomb L. SHEWA-B Program Health Impact Study Report [Internet]. 2014.
  26. 26. Luby SP, Halder AK, Huda T, Unicomb L, Johnston RB. The Effect of Handwashing at Recommended Times with Water Alone and With Soap on Child Diarrhea in Rural Bangladesh: An Observational Study. Plos Med. 2011;8.
  27. 27. Barda BD, Rinaldi L, Ianniello D, Zepherine H, Salvo F, Sadutshang T, et al. Mini-FLOTAC, an Innovative Direct Diagnostic Technique for Intestinal Parasitic Infections: Experience from the Field. PLoS Negl Trop Dis. 2013;7: e2344. pmid:23936577
  28. 28. Barda B, Zepherine H, Rinaldi L, Cringoli G, Burioni R, Clementi M, et al. Mini-FLOTAC and Kato-Katz: helminth eggs watching on the shore of Lake Victoria. Parasit Vectors. 2013;6: 220. pmid:23902918
  29. 29. WHO. Prevention and control of schistosomiasis and soil-transmitted helminthiasis: Report of a WHO expert committee. World health Organization (Geneva); 2002. Report No.: WHO Technical Report Series, 912.
  30. 30. WHO. Core Questions on Drinking-water and Sanitation for Household Surveys. In: WHO [Internet]. 2006 [cited 23 Dec 2013].
  31. 31. Vyas S, Kumaranayake L. Constructing socio-economic status indices: how to use principal components analysis. Health Policy Plan. 2006;21: 459–468. pmid:17030551
  32. 32. Patil SR, Arnold BF, Salvatore AL, Briceno B, Ganguly S, Colford JM Jr, et al. The effect of India’s total sanitation campaign on defecation behaviors and child health in rural Madhya Pradesh: a cluster randomized controlled trial. PLoS Med. 2014;11: e1001709. pmid:25157929
  33. 33. Hastie T, Tibshirani R, Friedman JH. The Elements of Statistical Learning: Data Mining, Inference, and Prediction, Second Edition. Springer; 2009.
  34. 34. Zou G. A Modified Poisson Regression Approach to Prospective Studies with Binary Data. Am J Epidemiol. 2004;159: 702–706. pmid:15033648
  35. 35. VanderWeele TJ, Tchetgen Tchetgen EJ. Attributing Effects to Interactions: Epidemiology. 2014;25: 711–722. pmid:25051310
  36. 36. Rothman K. Modern Epidemiology. Boston: Little, Brown; 1986.
  37. 37. VanderWeele TJ, Robins JM. The Identification of Synergism in the Sufficient-Component-Cause Framework: Epidemiology. 2007;18: 329–339. pmid:17435441
  38. 38. VanderWeele TJ. Sufficient Cause Interactions and Statistical Interactions: Epidemiology. 2009;20: 6–13. pmid:19234396
  39. 39. Rothman KJ, Greenland S, Lash TL. Modern Epidemiology. Philadelphia, PA: Lippincott Williams & Wilkins; 2008.
  40. 40. VanderWeele TJ, Knol MJ. A Tutorial on Interaction. Epidemiol Methods. 2014;3: 33–72.
  41. 41. Assmann SF, Hosmer DW, Lemeshow S, Mundt KA. Confidence Intervals for Measures of Interaction. Epidemiology. 1996;7: 286–290. pmid:8728443
  42. 42. Albonico M, Shamlaye N, Shamlaye C, Savioli L. Control of intestinal parasitic infections in Seychelles: a comprehensive and sustainable approach. Bull World Health Organ. 1996;74: 577–586. pmid:9060217
  43. 43. Albonico M, Stoltzfus RJ, Savioli L, Chwaya HM, d’Harcourt E, Tielsch JM. A controlled evaluation of two school-based anthelminthic chemotherapy regimens on intensity of intestinal helminth infections. Int J Epidemiol. 1999;28: 591–596. pmid:10405869
  44. 44. Massa K, Magnussen P, Sheshe A, Ntakamulenga R, Ndawi B, Olsen A. The effect of the community-directed treatment approach versus the school-based treatment approach on the prevalence and intensity of schistosomiasis and soil-transmitted helminthiasis among schoolchildren in Tanzania. Trans R Soc Trop Med Hyg. 2009;103: 31–37. pmid:18771789
  45. 45. Anderson RM, May RM. Helminth Infections of Humans: Mathematical Models, Population Dynamics, and Control. Advances in parasitology; 1985.
  46. 46. Anderson RM, Medley GF. Community control of helminth infections of man by mass and selective chemotherapy. Parasitology. 1985;90: 629–660. pmid:3892436
  47. 47. Anderson RM, Truscott JE, Pullan RL, Brooker SJ, Hollingsworth TD. How Effective Is School-Based Deworming for the Community-Wide Control of Soil-Transmitted Helminths? PLoS Negl Trop Dis. 2013;7: e2027. pmid:23469293
  48. 48. Bundy DA, Wong MS, Lewis LL, Horton J. Control of geohelminths by delivery of targeted chemotherapy through schools. Trans R Soc Trop Med Hyg. 1990;84: 115–120. pmid:2345909
  49. 49. Asaolu SO, Holland CV, Crompton DWT. Community control of Ascaris lumbricoides in rural Oyo State, Nigeria: mass, targeted and selective treatment with levamisole. Parasitology. 1991;103: 291–298. pmid:1745554
  50. 50. Albonico M, Smith PG, Hall A, Chwaya HM, Alawi KS, Savioli L. A randomized controlled trial comparing mebendazole and albendazole against Ascaris, Trichuris and hookworm infections. Trans R Soc Trop Med Hyg. 1994;88: 585–589. pmid:7992348
  51. 51. Vercruysse J, Behnke JM, Albonico M, Ame SM, Angebault C, Bethony JM, et al. Assessment of the Anthelmintic Efficacy of Albendazole in School Children in Seven Countries Where Soil-Transmitted Helminths Are Endemic. PLoS Negl Trop Dis. 2011;5: e948. pmid:21468309
  52. 52. Steinmann P, Utzinger J, Du Z-W, Jiang J-Y, Chen J-X, Hattendorf J, et al. Efficacy of Single-Dose and Triple-Dose Albendazole and Mebendazole against Soil-Transmitted Helminths and Taenia spp.: A Randomized Controlled Trial. PLoS ONE. 2011;6: e25003. pmid:21980373
  53. 53. Strunz EC, Addiss DG, Stocks ME, Ogden S, Utzinger J, Freeman MC. Water, Sanitation, Hygiene, and Soil-Transmitted Helminth Infection: A Systematic Review and Meta-Analysis. PLoS Med. 2014;11: e1001620. pmid:24667810
  54. 54. Kightlinger LK, Seed JR, Kightlinger MB. Ascaris lumbricoides intensity in relation to environmental, socioeconomic, and behavioral determinants of exposure to infection in children from southeast Madagascar. J Parasitol. 1998;84: 480–484. pmid:9645843
  55. 55. Conway DJ, Hall A, Anwar KS, Rahman ML, Bundy DAP. Household aggregation of Strongyloides stercoralis infection in Bangladesh. Trans R Soc Trop Med Hyg. 1995;89: 258–261. pmid:7660426
  56. 56. Holland CV, Taren DL, Crompton DWT, Nesheim MC, Sanjur D, Barbeau I, et al. Intestinal helminthiases in relation to the socioeconomic environment of Panamanian children. Soc Sci Med. 1988;26: 209–213. pmid:3347848
  57. 57. Bundy DAP. Epidemiological aspects of Trichuris and trichuriasis in Caribbean communities. Trans R Soc Trop Med Hyg. 1986;80: 706–718. pmid:3299888
  58. 58. Bundy DAP, Thompson DE, Cooper ES, Golden MHN, Anderson RM. Population dynamics and chemotherapeutic control of Trichuris trichiura infection of children in Jamaica and St. Lucia. Trans R Soc Trop Med Hyg. 1985;79: 759–764. pmid:3832488
  59. 59. Anderson RM. The population dynamics and epidemiology of intestinal nematode infections. Trans R Soc Trop Med Hyg. 1986;80: 686–696. pmid:3299886
  60. 60. Moncayo AL, Vaca M, Amorim L, Rodriguez A, Erazo S, Oviedo G, et al. Impact of Long-Term Treatment with Ivermectin on the Prevalence and Intensity of Soil-Transmitted Helminth Infections. PLoS Negl Trop Dis. 2008;2: e293. pmid:18820741
  61. 61. Feikin DR, Audi A, Olack B, Bigogo GM, Polyak C, Burke H, et al. Evaluation of the Optimal Recall Period for Disease Symptoms in Home-Based Morbidity Surveillance in Rural and Urban Kenya. Int J Epidemiol. 2010;39: 450–458. pmid:20089695
  62. 62. Hutcheon JA, Chiolero A, Hanley JA. Random measurement error and regression dilution bias. BMJ. 2010;340: c2289–c2289. pmid:20573762
  63. 63. Barda B, Albonico M, Ianniello D, Ame SM, Keiser J, Speich B, et al. How Long Can Stool Samples Be Fixed for an Accurate Diagnosis of Soil-Transmitted Helminth Infection Using Mini-FLOTAC? PLoS Negl Trop Dis. 2015;9: e0003698. pmid:25848772
  64. 64. Nikolay B, Brooker SJ, Pullan RL. Sensitivity of diagnostic tests for human soil-transmitted helminth infections: a meta-analysis in the absence of a true gold standard. Int J Parasitol. 2014;44: 765–774. pmid:24992655
  65. 65. Greenland S. The Effect of Misclassification in the Presence of Covariates. Am J Epidemiol. 1980;112: 564–569. pmid:7424903