January 2013

Soil-transmitted Helminths, Deworming, and Reinfection in Chinas Guizhou

Province
Linxiu Zhang, Yingping Cai, Xiaobing Wang, Xiaochen Ma, Alexis Medina, D. Scott
Smith*

Linxiu Zhang, Deputy Director, Center for Chinese Agricultural Policy, Institute for Geographical
Sciences and Natural Resources Research, Chinese Academy of Sciences, Beijing, China. Email:
lxzhang.ccap@igsnrr.ac.cn; Yingping Cai, Center for Chinese Agricultural Policy, Institute for
Geographical Sciences and Natural Resources Research, Chinese Academy of Sciences, Beijing, China.
Email: caiyp.08b@igsnrr.ac.cn; Xiaobing Wang (corresponding author), Post-Doctoral Researcher, Center
for Chinese Agricultural Policy, Institute for Geographical Sciences and Natural Resources Research,
Chinese Academy of Sciences, Beijing, China. Email: xbwang.ccap@igsnrr.ac.cn; Xiaochen Ma, Graduate
Student, Department of Agricultural and Resource Economics, University of California, Davis. Email:
mxc4068@gmail.com; Alexis Medina, Project Manager, Rural Education Action Project, Stanford
University, Stanford. Email: amedina5@stanford.edu; D. Scott Smith, Professor, Freeman Spogli Institute
for International Studies, Stanford University, Stanford. Email: ssmith@stanford.edu.

Soil-transmitted Helminths, Deworming and Reinfections in Chinas

Guizhou Province
Abstract
There is little hard data on STH infections in China. To examine this issue in a more
systematic way, we conducted a survey of 1,724 children, aged 3-5 and aged 8-10, in 6
counties in poor, impoverished and minority areas of Guizhou province in Southwest
China. Anthropometric, demographic, parasitological and hygiene data were collected to
better understand the prevalence of STH infections and the effectiveness of deworming in
rural China. We find that 37.5% of the sample children were infected with one or more of
the three types of STH tested for in this study. However, only 50.4% of them reported
having taken deworming medicine during the 18 month period prior to the survey. Of
those who reported being dewormed, 34.6% tested positive for STH infections. We find
that poverty and the number of siblings are significantly and positively correlated with
infections and reinfections, and parental education is significantly and negatively
correlated with infections and reinfections. We find no statistically significant evidence
linking ethnicity with infection rate, decision to deworm, or reinfection status.
Key Words: intestinal roundworms, soil-transmitted helminths (STH), children, poor,
rural, China, deworming; reinfections
JEL Codes: I14, I15, I18, O10, O53

Soil-transmitted Helminths, Deworming, and Reinfection in Chinas Guizhou

Province
1. Introduction
Infections with soil-transmitted helminths (STH)including the roundworm
(Ascaris lumbricoides), two types of hookworm (Ancylostoma duodenale and Necator
americanus), and the whipworm (Trichuris trichiura)are amongst the most common
infections worldwide. More than 130 countries/territories are endemic with STHs (World
Health Organization, 2008). Children are disproportionately vulnerable to STH
infections. Globally, there are around 800 million children infected with STH, and most
them live in developing countries (Chan, 2011).
There is a well-established linkage between STH infection and anemia,
malnutrition and other adverse health and nutrition outcomes (Adams et al., 1994;
Corbett et al., 1992; Hotez and Pritchard, 1995; Pollitt, 1990). These findings point to the
importance of deworming programs; indeed, researchers have even shown a positive
educational impact of deworming in rural schools (Miguel and Kremer, 2004). Thus,
more effort in recent years has focused on deworming programs in the developing world
(see, for example, the Children Without Worms project).
Despite academic consensus in the international literature of the adverse
consequences of STH infections and the importance of deworming, as well as a growing
worldwide awareness about STHs, high infection rates continue to be observed in China.
While Chinas health officials were concerned with STH infections throughout the 1960s
and 1970s, and even had a comprehensive national deworming program admired by
many development practitioners (Wagstaff et al., 2009a), attention to the disease and its
treatment seems to have declined over the past two decades since then. This trend began

in the 1980s when public funding for rural health declined precipitously (Wagstaff et al.,
2009b). At the same time, rural Chinas barefoot doctor systemwhich effectively
offered free treatment of common diseases, including STH infectionscollapsed, leaving
rural residents to fend for themselves and forcing them to seek care from private
providers with little or no help from a national insurance system. A number of diseases
that had been well-controlled have re-emerged in rural China in recent years. STHs
perhaps due to their nearly invisible nature and prevalence that is highest in remote, rural
areasare now re-emerging (Wang et al., forthcoming).
The literature in Chinathough it is mostly anecdotal or focused on a single
regionseems to support the claim that there is a resurgence of STH infections in China.
High STH prevalence has been observed in various regions of the country from Yunnan
(Steinmann et al., 2008) to Fujian (Xu et al., 2000) to Hunan (Zhou et al., 2007), though
nearly all of these studies were small, limited to a single township or village. A
parasitological survey of 274 school children, aged 10-12 years, in five villages in Hunan
Province showed STH infection rates to be 37.3% in girls and 33.3% in boys (Zhou et al.,
2007). The prevalence of the three main STHs (Ascaris, hookworm, and Trichuris) in 215
individuals in one village in the southern part of Yunnan Province was above 85%.
Within our study areas, the national Ministry of Health (MOH) survey of helminth
prevalence from 2001-2004 included 356,629 individuals across China, and encompassed
several studies by local Centers for Disease Control and Prevention (CDC), including in
Guizhou Province (Coordinating Office of the National Survey on the Important Human
Parasitic Diseases, 2005). The Guizhou Provincial CDC found an STH prevalence of
53.8% among 4091 primary and middle school students, 48.3% among 1748 children of
preschool age (Wang et al., 2008). A more localized study in Guizhou found similar rates

of 50.0% in 2007 among children aged 2-12 years in a rural, mountainous county in
western Guizhou (Chen and Xia, 2010). These rates suggest that STH infections persist at
high levels among the rural population in Guizhou Province.
If there is little research on the prevalence of STH infections in China, there is
even less about recent efforts to control them. Ziegelbauer et al. (2010) looked at two
schools in Yunnan Province: one underwent annual deworming treatment for several
years, the other had no deworming treatment. They found a significant difference in the
prevalence and intensity of infections between the two schools, but no significant
difference in the self-rated quality of life or end-of-term grades among students with
infections compared to students without infections. They did find that students infected
with Ascaris had significantly lower measures of self-rated health. Besides the small
sample size, one of the main drawbacks of the study was that the two sample schools
were quite different. The school undergoing deworming was larger and located on the
outskirts of a large city, while the other school was a small boarding school located in a
remote rural area. These differences could lead to problems in identifying causality.
To our knowledge, there are no large statistical studies that have measured the
correlates of the decision to seek deworming treatment.
The overall goals of this paper are to report on the relationship between
deworming treatment and STH infection in rural China and to identify the individual and
household characteristics correlated with the decision to deworm. To achieve these goals,
we adopt a two-prong approach. First, we measure STH infection rates among schoolaged children in rural Guizhou, measure reinfection rates among children who have been
dewormed in the past, and analyze the effect of past deworming on current infection
status. Next, we identify the factors associated with STH infection, reinfection, and

participation in deworming activities, and compare the prevalence of STH infection rates
and effectiveness of deworming activities between Han Chinese and ethnic minorities.
The rest of the paper is organized as follows. Section 2 describes the data and
methodology. Section 3 documents the prevalence and intensity of STH infections, the
deworming efforts and the subsequent reinfection status in rural Guizhou. Using
multivariate analysis, we shed light on the effectiveness of deworming efforts, who
chooses to deworm and what other factors are associated with infection and reinfection
status. We also describe how the prevalence of STH infections, deworming, and
reinfection rates differ between communities of Han Chinese and communities of ethnic
minorities. Section 4 concludes.

2. Data and Methods

2.1 Study Setting
This survey of impoverished children in rural settings was carried out in June and
September of 2010 in Guizhou Province, located in Chinas Southwest region. Here, high
rates of poverty and the humid climate are conducive to STH infection. Guizhou is
demographically one of the most ethnically diverse provinces in China, with minority
groups representing 35.7% of the total population (National Bureau of Statistics of China,
2012).
Six rural counties were randomly selected from the bottom quartile of counties
based on average net per capita incomes (National Bureau of Statistics of China, 2010).
The average net per capita income in the study areas is 433 US$/year.

2.2 Study Population, Sample Size and Sampling Strategy

We studied two groups of children, aged 3-5 years old and 8-10 years old. Each
group differed with respect to socialization and exposure to school environments.
In each of the six selected counties, we ranked all towns according to net income
per capita, and then randomly chose four towns: two with income per capita above the
mean for the county and two with income per capita below the mean for the county. For
each of the four chosen towns, two sample schools were chosen: the central primary
school (which also serves as the local Bureau of Educations administrative
representative for all educational affairs in the town) and a randomly selected primary
school.
For each school, we obtained a list of all local villages that feed into the school.
We ranked this list of villages by the number of 8-10 year old students (henceforth called
school-aged children) enrolled at the school. We randomly selected two villages from the
list (henceforth called sample villages) that had 16 or more students enrolled at the
school. We then randomly chose 11 enrolled students from each sample village. In each
sample school, a total of 22 students were surveyed.
Next, we went to the sample villages to conduct the sampling of the 3-5 year old
children. We obtained a list of all the 3-5 year old children in each sample village from
the Registry of Child Immunization (which is recorded and stored in the towns health
center) and randomly chose 11 children from each of the sample villages (henceforth
called preschool-aged children).
Our power calculations indicated that for our primary variable of interest,
roundworm infection status, to estimate a 95% confidence interval with precision of 0.05
around a population prevalence of 40% and assuming a village-level intra-cluster

correlation (ICC) of 0.15, we required 8 children in each age group per village. We
increased this to 11 children to account for attrition.
In each sample village we randomly sampled 11 preschool-aged children and 11
school-aged children. This led to a total sample size of 844 pre-school aged children and
880 school-aged children (Table 1), for a total of 1,724 students in 48 schools and 94
villages. Because fecal samples were unattainable for some children, some sample
villages had fewer than 22 observations with fecal samples. In no case were there fewer
than 8 pre-school aged children and 8 school-aged children who provided samples. On
average, there were 9.36 school-aged children per sample village and 8.98 preschoolaged children per sample village. This level of attrition is considered to be low in the
existing literature (Steinmann et al., 2008; Xu et al., 2000; Zhou et al., 2004).
2.3 Data Collection and Survey
The primary variable of interest was stool parasite status of the selected child,
whether positive or negative for any of three types of STH, as determined by a single
stool sample. In addition, the intensity of every infection by egg density per gram was
measured using WHO standard protocol (Montresor et al., 2002). Other variables and
characteristics, such as household eating habits and sanitation information, were collected
using a socio-economic survey instrument. The survey contained questions about age,
gender, parental levels of education, health and sanitation behavior, household
characteristics, and whether the child had taken anti-helminthic medication within the 18
month period prior to the survey. The school-aged children completed the survey
themselves under the direct supervision of trained enumerators from the Chinese
Academy of Sciences. The preschool-aged childrens data was obtained by trained
enumerators who interviewed the childrens parents or caregivers.

Body height and weight were measured and recorded by trained nurses from
Xian Jiaotong University according to WHO recommendations (de Onis et al., 2004).
The children were measured in light clothing without shoes. Weight was measured with a
calibrated electronic scale recommended by the medical department at Xian Jiaotong
University. Body height was measured using a standard tape measure. The nursing team
was trained to make sure the weighing station was set up on level ground to ensure
accuracy of the equipment. Two nurses manned each measurement station, with one
responsible for preparing subjects for measurement (removing shoes, offering instruction,
reassuring parents, positioning children, etc.) and the other responsible for conducting
and recording the actual measurements.

2.4 Stool Sample Collection and Laboratory Testing

Stool samples of each of the children included in the study were collected once
and sent to the local county Center for Disease Control & Prevention (CDC). There was
one lab per county, for a total of six labs. The majority of the samples were tested the
same day that they were collected. Due to time and labor constraints, a small fraction of
samples were tested the day after collection. These samples were stored overnight in the
CDC laboratory refrigerator, which is kept at a constant 4 C. The Kato-Katz smear
method was used for species-specific identification of parasite eggs: Ascaris, hookworm,
and Trichuris. A single smear test was performed on each sample. Samples found to be
positive for any of these three parasites underwent egg burden counts to determine eggs
per gram (epg) of feces using standard WHO protocol.

CDC employees at the county level examined the samples and performed the
tests. As a quality control, ten percent of samples were also checked by a parasite expert
from the National Institute for Parasitic Disease to verify the initial diagnosis.
2.5 Ethical Considerations
This study was approved by the Stanford University Institutional Review Board
(IRB) on May 18, 2010 and was assigned study protocol number 18780. The legal
guardians (either parents or school principals) of all subjects provided informed oral
consent and the children themselves provided oral assent. The IRB approves the use of
oral consents in rural China to clarify understanding because many rural villagers are
illiterate and it is culturally unusual to sign in writing. Our study enumerators recorded
the consents on a list of names that is stored in a locked filing cabinet at the study center
in Beijing, China.
Stool sampling falls within the regular purview of the Chinese Center for Disease
Control & Prevention (CDC). CDC employees are professionally trained and perform
routine stool sampling in the study areas as part of their national responsibilities. The
stool sampling conducted as part of this study was approved and sanctioned by the
national Chinese CDC as well as the regional and local CDCs in the sample province.
2.6 Data Management and Statistical Analysis
In order to better understand the relationship between deworming treatment and
STH infection in rural China, we conduct a number of multivariate regression analyses.
First, using a probit estimator (since our dependent variable is binary in natureinfected
with STHs or not), we regress deworming history; household wealth measured by an
asset indexi (following Wang et al., 2012); individual and household characteristics; and

10

self-reported health and sanitation behavior on STH infection. The results of this analysis
will help shed light on understanding whether deworming is effective for reducing STH
infections among school-aged children in rural China, as well as what other factors are
most strongly associated with STH infection.
Second, we focus on the subsample of children who have undergone deworming
treatment in the past, and regress poverty level, individual and household characteristics,
and health and sanitation behavior on reinfection with STHs. We define reinfection to
include anyone who reported taking deworming medicine within the 18 month period
prior to the survey and who tested positive for STH infection. The results of this
regression will help us understand those behaviors and characteristics that are correlated
with reinfection.
Finally, we regress poverty level, individual and household characteristics, and
health and sanitation behavior on deworming history in order to identify factors
associated with the household / individual decision to undergo deworming treatment.

3. Results
3.1 Summary Statistics
We collected data on a variety of individual and household characteristics,
including age, gender, number of siblings, and parental education. In our sample, 45.1%
of children were female and 54.9% were male, a ratio typical in most poor areas in China
(National Bureau of Statistics of China, 2010). The average number of siblings is 1.6,
which means that, on average, each family has two or three children. Only 31.8% of
mothers and 48.7% of fathers have 9 years of education or more.

11

Health and sanitation habits include whether the children wash their hands before
eating; whether the children wash their hands after using the toilet; whether the children
consume undercooked or raw meat or vegetables; and whether the children drink unboiled water. The results are discouraging. Only 54.4% of children reported that they
wash their hands before eating, while just under half (49.3%) reported that they wash
their hands after using the toilet. Although we do not have data on why hand washing
behavior is so infrequent, our field visits and interviews suggest that one possible reason
may be a lack of water near the toilet and cooking facilities; indeed, many families in our
sample only have a single faucet in the courtyard of their homes, and no indoor
plumbing. Our data also show that 20.3% of children reported consuming uncooked
meat, while the rates for consumption of uncooked vegetables and un-boiled drinking
water were as high as 75.0% and 85.3%.
Infection rates in the study sites are high. Overall, 37.5% of the sampled children
were infected with one or more of the three types of STHs (Table 2). Infection rates with
Ascaris were the highest and infection with hookworm was the lowest. Infection rates
were higher in school-aged children (42.3%), and lower in preschool-aged children
(32.6%), and these differences were statistically significant (p<0.01). Statistically
significant differences in infection rates between the two age groups were also observed
between the three types of STHs when compared separately. The high infection rates and
the differences between the two age groups are consistent with prior recent observations.ii
The deworming rate in Guizhou was lower than we expected, given the low cost
of treatment and the past record of Chinas deworming initiatives. Only half of the
sample (50.4%) reported that they had taken deworming medicine within the past 18
months (Table 3). Preschool-aged children had significantly higher rates of deworming

12

than did school-aged children (p<0.01). The lower rate of deworming among school-aged
children seems counterintuitive, since it would seem that schools are a natural contact
place for an at-risk population, and that it would be easier to treat children when they are
together in a single place rather than when they are spread out and scattered as the
preschool-aged children are.
A full 34.6% of dewormed children had been reinfected with one or more of the
three types of STHs by the time of our survey; reinfection rates are significantly higher
(p<0.05) among school-aged children (39.8%) than among preschool-aged children
(31.6%, Table 4). Table 5, column 2 focuses on reinfection rates for Ascaris, since
prevalence was highest for this species. Although we do find significantly lower rates of
Ascaris infection among dewormed children (p<0.05), 30.0% of dewormed children had
been reinfected by the time of our survey. These high reinfection rates raise concerns that
the deworming efforts underway in Guizhou are not enough to curb the STH problem.
While prevalence data describe the proportion of individuals infected, they do not
provide an indication of the intensity of infection, which is also a central determinant of
STH transmission dynamics (Bundy and Silva 1998). Worm burden provides an estimate
of the number of roundworms in the intestines, a commonly used indicator of intensity of
infection. A successful anthelminthic treatment should reduce both the prevalence of
STHs in the population and the intensity of infection, and can typically be observed as a
change in the distribution of worm burden (i.e., shifts from heavy to light worm burden).
Table 5 shows data on worm burden separately for children who have been dewormed in
the past 18 months and for children who have not been dewormed in the past 18 months.
As prevalence rates were highest for Ascaris, we focus here on worm burden
measurements for children infected with Ascaris.

13

Our analysis of the worm burden data, like our analysis of the prevalence data,
casts some suspicion on the effectiveness of the deworming programs in the sample
villages/schools. In particular, the data show that there is no difference in mean Ascaris
burden between children who have been dewormed and those who have not. (p=0.77).
There is also no significant change in the distribution of Ascaris burden between children
who have been dewormed and those who have not. This holds true for both preschoolaged children and school-aged children (Table 5).
Why do the infection rates remain so high among children who have received
deworming treatment (Table 5)? And why are infection intensities unchanged among
children who have received treatment? One possibility is that individuals who take
deworming medicine may be taking an insufficient dose of the medication. National
guidelines released by the Chinese Center for Disease Control & Prevention (2010)
recommend the annual administration of only 200 mg of albendazole for children aged 212 years, only half of the 400 mg dose recommended by the World Health Organization
(2006).
A second reason may be related to the frequency of treatment. The WHO
treatment guidelines recommend that in communities with infection rates of 20% to 50%
(the rate observed in our study area), school-aged children should be treated once a year.
However, our survey asked individuals about their deworming history over the past 18
months. It is possible that many of them did not undergo deworming treatment once a
year as recommended, suggesting they are not receiving treatment with sufficient
regularity to be effective.
3.2 Multivariate Results

14

In order to more precisely estimate the relationship between deworming treatment

and infection rates among children in rural Guizhou, we use a probit estimator to regress
deworming history on infection status while controlling for different sets of control
variables. The results of the regression estimates are shown in columns 1 to 3 of Table 6.
The first column includes only controls for household wealth as measured by an asset
index. Columns 2 and 3 include additional control variables (individual characteristics
and healthy habits).
According to our analysis, there is no significant correlation between deworming
and infection status (Table 6, row 1). Although there appears to be a significant negative
correlation in column 1, the correlation fades in term of magnitude and statistical
significance as we refine the model by adding additional control variables (column 2 and
3).
What factors are associated with infection? In all three specifications, household
wealth is significantly correlated with infection. The poorer a household is, the more
likely it is that the children who live there will be infected. We also found that the number
of siblings was significantly and positively correlated with infectionliving in a family
with one additional sibling increases the probability of being infected by about 8%
columns 2 and 3). Parents educational background is also associated with the likelihood
of their children being infected. If a father has 9 years of education or more, the
probability of his children being infected with STHs is 20% lower than that of children
whose fathers have less than 9 years of education (columns 2 and 3). We find no
significant correlation between STH infection and hand washing behavior or the
consumption of uncooked meat or un-boiled water..
3.3 What Causes Reinfection?

15

When we limit our sample to only those children who reported having taken
deworming medication in the 18 months prior to our survey, we find similar results:
Household wealth, number of siblings, and parents education all predict risk of
reinfection. Household wealth is significantly correlated with the likelihood of
reinfection. The poorer one is, the higher the likelihood of reinfection (column 4). The
significant correlation disappears, however, when we control for additional individual
characteristics and healthy habits (columns 5 and 6). We also found that the number of
siblings was significantly and positively correlated with the likelihood of reinfection.
Living in a family with more than one sibling increases the probability of reinfection by
about 9% (columns 5 and 6). For the sub-sample of children who have undergone
deworming treatment, it is mothers education (rather than fathers) that is negatively and
significantly correlated with reinfection. When a mother has 9 years of education or
more, the probability of her already treated children becoming reinfected is about 22%
lower than that of children whose mothers have less than 9 years of education (columns 5
and 6). Consistent with the results of our other models, we found no statistically
significant results linking reinfection status with the childs health or sanitation habits.
This may be due to the low variance in sanitation habits across the study population.
3.4 Who Received Treatment?
Curiously, our analyses indicate that there are few correlates of the decision to
seek deworming treatment. According to our results (Table 7), treatment status is not
linked to household wealth. The coefficients on the asset variable (rows 1 to 3) are
statistically insignificant from zero (at the 10% level). In fact, none of the variables we
examine are significantly correlated with the decision to be treated. It seems that
individuals in our sample either do not have a clear strategy for being treated for STH

16

infections, or are not participating in a system that seeks out at-risk children and provides
them with treatment, or both.
This is consistent with the structure and organization of the rural Chinese health
care system, in which local health care providers have little or no incentive to provide or
even promote systematic community treatment of a low cost and low-profit disease.
Rather, the decision to seek care is left to the discretion of individuals and their families,
who are likely uninformed about both the health risks of STH infection and the benefits
of treatment.

3.5 Heterogeneity Between Han Chinese and Minority Chinese

In this section we examine if ethnicity is a determinant of infection, deworming
status, or reinfection after deworming. We collected data on minority status for three of
our six sample counties. In these areas, the share of minorities was 80.9%, 45.5% and
5.7%. We use this data to explore whether there are any systematic differences between
Han Chinese and other ethnic minorities in terms of STH prevalence, treatment status,
and reinfection rates.
Table 8 reports ethnicity-specific infection rates in the three sampled counties for
which we have data on minority status. Infection rates among the Han community are
11.9%, 46.0% and 48.9% in each of the three counties. In contrast, infection rates among
the non-Han community are 24.4%, 44.2%, and 62.5%, respectively. Overall, the
infection rates among ethnic minorities were around 13% higher than those among Han
Chinese (in two of three sampled counties). These differences may imply that STH
infections are more severe among Chinas minority groups. However, the difference is
statistically significant in only one county (at the 5% level).

17

The ethnicity-specific deworming rates in the three samples counties are shown in
Table 9. Deworming rates among the Han communities range from 43.2% to 59.9%. The
deworming activities of the non-Han communities range from 42.0% to 93.8%. In two of
the three counties, the deworming rates of minority Chinese were significantly higher
than those of Han Chinese (14% and 35% higher, respectively, both significant at the 1%
level). In other words, the evidence shows that deworming efforts are actually more
common in minority communities.
Do the higher treatment rates among minority populations result in lower
reinfection rates compared to Han communities? The results in Table 10 suggest the
opposite. The reinfection rates among Han Chinese are 13.3%, 35.5%, and 44.9%,
respectively, in each of the three counties. The reinfection rates among non-Han Chinese
are 26.7%, 38.1%, and 60.0%. In all three counties, the reinfection rates of minority
Chinese are higher than those of Han Chinese (13.3%, 2.6% and 15.1%). However, the
difference is not statistically significant in any of the three counties.
Overall, in the three counties for which we have information about minority
infection rates, the prevalence of STH infections is higher among minority groups than
among Han groups. The difference, however, is only statistically significant in one
county. Deworming efforts are also more common in minority communities. In two of the
three counties, the deworming rates of minorities are significantly higher than those of
Han communities. Unfortunately, the higher deworming rates among minority groups do
not translate to lower rates of reinfection. Instead, reinfection rates in minority
communities are higher than those of Han communities in all three counties. The
difference, however, is not statistically significant in any of the three counties.

18

In order to better explore the heterogeneity in the effectiveness of deworming

treatment between Han Chinese and other minority Chinese, and to identify the factors
correlated with deworming decisions and subsequent reinfection status, we conduct
multivariate regression analysis using the subsample of the three counties with minority
information (results not shown). We use the same models that were used in the main
results section, adding minority dummies to control for unobserved differences in
minority cultures and county fixed effects to control for large observed differences in
percent minority across the three counties. Importantly, we do not find any statistically
significant evidence that minority status is the determinant of infection. Minority status
also is not associated with deworming decisions or reinfection status.

4. Conclusion
In this paper we document the prevalence and intensity of STH infections among
1,724 children in rural, impoverished and minority areas of Guizhou Province in
Southwest China. According to our data, 37.5% of the sample children are infected with
one or more of the three types of STH. The high prevalence of STH infections found in
Guizhou is consistent with findings from other regions of China (Zhou et al., 2007; Wang
et al., forthcoming). Of the three types of STHs, infection with Ascaris was most
common. Hookworm was the least prevalent roundworm in the study areas. Overall
infection rates were higher in school-aged children (42.3%), and slightly lower in
preschool-aged children (32.6%).
Because only a single stool sample was collected on each child in this study,
infection rates are likely underestimated. A single stool sample can underreport infection
in an individual because of the temporal variation in egg excretion over hours and days

19

(Wang et al. forthcoming). Obtaining only a single sample for Ascaris and Trichuris can
lead to underestimated infection rates of up to 50% (Knopp, 2008). Therefore, the real
infection rate is likely to be even higher than what our data indicate.
We next investigated the relationship between deworming and infection. We
found low self-reported deworming rates (50.4%) and a high overall reinfection rate of
34.6%. In addition, we found no evidence of reduction in the worm burden. These
findings raise the concern that the deworming efforts currently employed in the study
areas are ineffective. This hypothesis is supported by the results of the multivariate
analysis, which show no significant correlation between deworming treatment and STH
infections when we control for multiple variables.
We find few correlates of deworming decisions. It seems that individuals in our
sample either do not have a clear treatment-seeking strategy, or are not in a system that
can correctly identify at-risk children and provide them with treatment. These findings
underscore the importance of a long-term, systematic and consistent deworming regimen.
It is important to note that a potential limitation of this study is its reliance on a selfreported variable: Did you (your child) take deworming medication within the past 18
months? Medication recall certainly involves some degree of error; however, other
researchers have observed agreement rates between patient recall and medical records of
69-87% (Hill & Ross, 1982). Given that no medical records exist of deworming activities
in our study areas, recall is the next best method of obtaining treatment history.
Our study also explores other factors associated with infection and reinfection. We
identify number of siblings as significantly and positively correlated with infection and
reinfection, and find household wealth and years of parental education to be significantly
and negatively correlated with infection and reinfection.

20

Finally, our study examines whether ethnicity is a determinant of infection,

treatment status, or reinfection after treatment. We find no statistically significant
evidence linking ethnicity with infection, deworming decisions or reinfection status.
One potential source of bias may have come from the administration of the
socioeconomic survey forms. School-aged children answered questions about their
individual, household, and health behavior characteristics on their own, while the same
information was collected about preschool children by asking their parents. It is possible
that this may have led to fundamental differences between the two age groups in the types
of answers we received, although there is no reason to think that either group would
systematically overestimate or underestimate in their responses. Regardless, any
differences between the age groups should not affect our results since we control for the
age of the child in our multivariate analyses.
It is our recommendation based on this studys findings that Chinas public health
system should consider launching a deworming treatment strategy that is consistent with
the recommendations of the World Health Organization (2005). WHO treatment
guidelines suggest that the implementation of STH control programs at the country level
requires strong links with existing interventions already in place. For example, in
Cambodia the government has developed a health package that delivers deworming drugs
together with vitamin A, which has reduced delivery costs and enabled remote
communities to be included in the program (World Health Organization, 2005). We
believe that it is entirely possible for China to use this hitching a lift strategy. Our
previous research has shown that as many as 39% of rural Chinese children suffer from
iron deficiency anemia, which in turn, can be caused by STH infections (Luo et al.,
2012). Hence, it may be of interest to consider the possible bundling of deworming

21

medication with iron and other micronutrient supplementation programs (World Health
Organization, 2005).
Certain aspects of Chinas current health system are not compatible with the
immediate launch of such deworming strategies. Chinas local officials have incentives
that are poorly aligned with ultimate social objectives like good population heath.
Hospitals have little incentive to deliver public health programs. The profit-motivated
health care providers in rural China have little to gain from implementing a low-cost STH
treatment program. Teachers and principals often avoid taking responsibility for basic
health interventions, such as deworming treatment. Because no single actor is
incentivized to take responsibility for STH infections, deworming efforts are condemned
before they even begin. For anthelminthic campaigns to actually succeed, China will need
to pay close attention to local-level incentives to improve child health.

22

References
Adams, Elizabeth J., Lani S. Stephenson, Michael C. Latham, Stephen N. Kinoti, 1994,
Physical Activity and Growth of Kenyan School Children with Hookworm,
Trichuris trichiura and Ascaris lumbricoides Infections Are Improved after
Treatment with Albendazole. Journal of Nutrition, Vol. 124, No.8, pp. 1199-1206.
Bundy, Donald and Nilanthi R. de Silva, 1998, Can We Deworm This Wormy World?
British Medical Bulletin, Vol. 54, No. 2.
Center for Disease Control & Prevention, 2010, Treatment guidelines for soil-transmitted
helminth infections. Official Disease Control Publication of the Administrative
Office of the Ministry of Health, Vol. 98. [in Chinese].
Chan, Margaret, 2011, WHO welcomes GlaxoSmithKline support to fight intestinal
worms in children. Statement by WHO Director-General, 9 September.
Chen, X., Xia, X., 2010, Survey of intestinal nematode infection among children visiting
a hospital in Puding County. Journal of Preventive Medicine, Vol. 4. [in Chinese].
Coordinating Office of the National Survey on the Important Human Parasitic Diseases,
2005, A National Survey on Current Status of the Important Parasitic Diseases in
Human Population, China Journal of Parasitology and Parasitic Diseases, Vol.
23, pp. 332-340.
Corbett, E.L. et al., 1992, Nutritional Status of Children with Schistosomiasis Mansoni in
Two Different areas of Machakos District, Kenya, Transactions of the Royal
Society of Tropical Medicine and Hygiene, Vol. 86, pp. 266-273.
Hill, A.P., Ross, R.K., 1982, Reliability of recall of drug usage and other health-related
information, Am. J. Epidemiol., Vol. 116, No. 1, pp. 114-122.
Hotez, Peter J., Pritchard, David I., 1995, Hookworm Infection. Scientific American, June
Vol., pp. 68-74.
Knopp, Stephanie et al., 2008, Diagnosis of Soil-Transmitted Helminths in the Era of
Preventive Chemotherapy: Effect of Multiple Stool Sampling and Use of
Different Diagnostic Techniques, PLoS Neglected Tropical Diseases, Vol. 2: e331.
Luo, Renfu et al., 2012, Nutrition And Educational Performance In Rural Chinas
Elementary Schools: Results of A Randomized Control Trial In Shaanxi Province,
Economic Development and Cultural Change, Vol. 60, No. 4, 735-772.
Miguel, Edward and Michael Kremer, 2004, Worms: Identifying Impacts On Education
And Health In The Presence Of Treatment Externalities. Econometrica, Vol. 72,
No. 1.

23

Montresor, Antonio et al., 2002, Helminth control in school-age children, Geneva: World
Health Organization.
National Bureau of Statistics of China, 2010, China Statistical Yearbook, Beijing: China
Statistics Press.
National Bureau of Statistics of China, 2012, Tabulation on the 2010 population census
of the Peoples Republic of China, Beijing: China Statistics Press.
de Onis, Mercedes, Adelheid W. Onyango, Jan Van den Broeck, Wm Cameron Chumlea,
Reynaldo Martorell, 2004, Measurement and standardization protocols for
anthropometry used in the construction of a new international growth reference,
Food and Nutrition Bulletin, Vol. 25. pp. S27-S36
Pollitt, Ernesto, 1990, Infection: Schistosomiasis. Malnutrition and Infection in the
Classroom, Paris, Unesco: pp. 151-162.
Steinmann, Peter et al., 2008, Extensive multiparasitism in a village of Yunnan province,
Peoples Republic of China, revealed by a suite of diagnostic methods, Am. J.
Trop. Med. Hyg., Vol. 78, pp. 760-769.
Wagstaff, Adam, Magnus Lindelow, Shiyong Wang, and Shuo Zhang, 2009a, Reforming
Chinas Rural Health System, Washington, DC: The World Bank.
Wagstaff, Adam, Winnie Yip, Magnus Lindelow, and William C. Hsiao, 2009b, Chinas
health system and its reform: A review of recent studies, Health Economics, Vol.
18, pp. S7-S23.
Wang, Shihai et al., 2008, Current status and analysis of important human parasitic
diseases in Guizhou Province, Journal of Pathogen Biology, Vol. 2, pp. 450-453.
[in Chinese]
Wang, Xiaobing et al., 2012, Are poor college students thriving? Financial aid and
college students from China's poor rural areas, Center for Chinese Agricultural
Policy, REAP Working Paper #226.
Wang, Xiaobing et al., 2012. Soil-Transmitted Helminth Infections and Correlated Risk
Factors in Preschool and School-aged Children in Rural Southwest China, PloS
One, Vol. 7, No. 9, e45939.
World Health Organization, 2005, Deworming for Health and Development. Geneva:
World Health Organization.
World Health Organization, 2006, Preventive Chemotherapy in Human Helminthiasis:
Coordinated Use of Anthelminthic Drugs, Geneva: World Health Organization.

24

World Health Organization, 2008, Soil-transmitted helminthiasis. Weekly

Epidemiological Record, Vol. 83, No. 27/28, pp. 237-252.
Xu, Longshan et al., 2000, Creating health-promoting schools in rural China: a project
started from deworming, Health Promotion International, Vol. 15, pp.197-206.
Zhou, Huan, Chiho Watanabe, Ryutaro Ohtsuka, 2007, Impacts of dietary intake and
helminth infection on diversity in growth among schoolchildren in rural south
China: A four-year longitudinal study, American Journal of Human Biology, Vol.
19, pp. 96-106.
Ziegelbauer, Kathrin. et al., 2010, Self-rated Quality Of Life And School Performance In
Relation To Helminth Infefctions: Case Study From Yunnan, Peoples Republic
Of China. Parasites & Vectors, Vol. 3, p. 61.

25

Table 1. Sample Description

County
Survey
Villages
Date
(No.)
County 1
County 2
County 3
County 4
County 5
County 6
Total

June, 2010
June, 2010
June, 2010
Sep, 2010
Sep, 2010
Sep, 2010

15
16
16
15
16
16
94

Schools
(No.)

Sampled
Children

8
8
7
8
9
8
48

261
267
283
309
323
281
1,724

Of which:
3-5 years
8-10 years
121
138
122
153
163
147
844

140
129
161
156
160
134
880

Note: Because multiple villages were merged into one administrative village in Counties 1
and 4, our sample there includes only 15 villages (instead of 16).

26

Table 2. Prevalence of STH Infections

Infected
-Ascaris
-Hookworm
-Trichuris
Not infected
Total
Sample size

All sampled
children
37.53%
32.71%
1.39%
10.85%
62.47%
100.00%
1,724

3-5
year-olds
32.58%
28.79%
0.47%
6.64%
67.42%
100.00%
844

8-10 yearolds
42.27%
36.48%
2.27%
14.89%
57.73%
100.00%
880

P value of
difference
0.0000
0.0007
0.0014
0.0000
----

Note:
1. Because some students have multiple infections, the sum of the by-species infection
rates is larger than the total infection rate.
2. The last column shows the P value of t-statistic of the difference in infection rates
between the two age groups.

27

Table 3. Deworming Rates

County 1
County 2
County 3
County 4
County 5
County 6
Total

All sampled children

No.
Percent
dewormed
261
59.77%
267
55.43%
283
34.63%
309
43.69%
323
49.54%
281
60.85%
1,724
50.35%

3-5 year-old children

No.
Percent
dewormed
121
76.03%
138
70.29%
122
49.18%
153
54.90%
163
68.71%
147
72.11%
844
65.28%

8-10 year-old children

No.
Percent
dewormed
140
45.71%
129
39.53%
161
23.60%
156
32.69%
160
30.00%
134
48.51%
880
36.02%

P value of
difference
0.0000
0.0000
0.0000
0.0001
0.0000
0.0000
0.0000

Note:
1. Percent dewormed indicates whether the child reported taking deworming medication within the past 18 months.
2. The last column shows the P value of t-statistic of the difference in deworming rates between the two age groups.

28

Table 4. Reinfection Rates (Children who tested positive for infection among those who reported taking deworming medication within the 18
months prior to the survey)

Infected
-Ascaris
-Hookworm
-Trichuris
Not infected
Total
Sample size

All sampled
children
34.56%
29.95%
0.81%
8.87%
65.44%
100.00%
868

3-5 year-old
children
31.58%
27.22%
0.54%
6.35%
68.42%
100.00%
551

8-10 year-old
children
39.75%
34.70%
1.26%
13.75%
60.25%
100.00%
317

P value
of difference
0.0148
0.0206
0.2557
0.0006
----

Note:
1. The last column shows the P value of t-statistic of the difference in reinfection rates between the two age groups.

29

Table 5. Ascaris Burden for Non-dewormed and Dewormed Children by Age Cohort

Sample Size

Ascaris
Infection
Rates
32.71%
35.51%
29.95%
0.0132

All sample children

Not dewormed
Dewormed
p-value of difference

1,724
856
868

3-5 year old children

Not dewormed
Dewormed
p-value of difference

844
293
551

28.79%
31.84%
27.22%
0.1586

8-10 year old children

Not dewormed
Dewormed
p-value of difference

880
563
317

36.48%
37.47%
34.70%
0.4118

If infected,
mean Ascaris
burden (epg)
11142
10882
11446
0.7672

If infected, % with worm burden that is:

Light

Moderate

Heavy

59.22%
59.87%
58.46%
0.7352

36.70%
36.51%
39.92%
0.9200

4.08%
3.62%
4.62%
0.5515

12538
14573
11276
0.3349

58.02%
54.84%
60.00%
0.4302

36.21%
37.63%
35.33%
0.7181

5.77%
7.52%
4.67%
0.3544

10085
9255
11677
0.2939

60.12%
62.08%
56.36%
0.3219

37.07%
36.02%
39.09%
0.5900

2.81%
1.90%
4.53%
0.1733

Note:
1. The severity of Ascaris burden is defined by the number of Ascaris eggs per gram (epg) of feces. Light infection is defined as < 4999 epg;
moderate infection is defined as 4999-49,999 epg, and heavy infection is defined as > 49,999 epg.

30

2. P value of t-statistics of difference is reported.

31

Table 6. Probit Estimates of Infection and Reinfection

Dependent
Variable

Full sample

Dewormed sample

(1 = Infected, 0 = Otherwise)

(1 = Infected, 0 = Otherwise)

(1)

(2)

(3)

Reported taking deworming

-0.151

(4)

(5)

(6)

-0.087

-0.080

medicine in past 18 months

(-2.44)

(-1.32)

(-1.21)

-0.157

-0.148

-0.146

-0.121

-0.101

-0.100

(-1.80)

(-1.66)

(-1.64)

(-0.98)

(-0.79)

(-0.77)

(1=yes, 0=no)
Household Wealth
2nd quartile
rd

3 quartile
th

4 quartile

**

-0.226

-0.170

-0.140

-0.272

-0.203

-0.195

(-2.67)

(-1.95)

(-1.59)

(-2.22)

(-1.59)

(-1.52)

-0.311

-0.171

-0.150

(-2.56)

(-1.31)

(-1.11)

-0.349

***

(-4.07)

**

-0.238

-0.192

(-2.62)

(-2.04)

Individual Characteristics
Student status (1=school-aged

-0.512

-0.582

-0.224

-0.310

child, 0=preschool-aged child)

(-0.81)

(-0.92)

(-0.21)

(-0.29)

Age

0.0802

0.0784

0.125

0.126

(1.38)

(1.35)

(1.69)

(1.70)

0.025

0.034

-0.034

-0.028

(0.29)

(0.40)

(-0.25)

(-0.21)

0.006

-0.001

-0.061

-0.062

(0.10)

(-0.01)

(-0.67)

(-0.68)

**

0.090*
(2.27)

Student*Age
Gender (1=male, 0=female)
Number of Siblings

**

0.085

0.082

0.094

(3.07)

(2.94)

(2.39)
*

-0.211*

Mother finished secondary

-0.088

-0.082

-0.222

school or above (1=yes, 0=no)

(-1.19)

(-1.09)

(-2.10)

(-1.98)

**

**

Father finished secondary

-0.202

-0.201

-0.174

-0.169

school or above (1=yes, 0=no)

(-2.96)

(-2.93)

(-1.76)

(-1.70)

Health and Sanitation Habits

Wash hands before dinner

-0.035

-0.093

(1=yes, 0=no)

(-0.45)

(-0.83)

Wash hands after using toilet

-0.125

0.056

(1=yes, 0=no)

(-1.59)

(0.48)

**

0.159

Eat uncooked vegetables

0.192

(1=never, 0 otherwise)

(2.61)

(1.42)

Eat uncooked meat (1=never, 0

-0.010

-0.040

otherwise)

(-0.13)

(-0.33)

Drink un-boiled water

-0.063

-0.158

(1=never, 0 otherwise)

(-0.68)

Constant
Sample Size
*

(-1.11)
-0.737*

(-1.03)

(-2.26)

(-2.00)

(-2.66)

(-2.48)

(-2.16)

1724

1724

1724

868

868

868

Note: t statistics in parentheses; p < 0.10,

p < 0.05,

-0.775

-0.550

***

-0.224

**

-0.0666

**

-0.517

p < 0.01

The first quartile (omitted) represents households with the fewest assets and the fourth quartile
represents households with the most assets.
The coefficients of the probit regressions are reported in the table.

33

Table 7. Probit Estimates of Treatment Status

Dependent

Full sample

Variable

(1 = Dewormed, 0 =Not)
(1)

(2)

(3)

0.107

0.152

0.163

(1.24)

(1.69)

(1.81)

0.052

0.119

0.123

(0.63)

(1.36)

(1.38)

0.119

0.139

0.156

(1.43)

(1.54)

(1.67)

Student status (1=school-aged child,

-0.646

-0.528

0=preschool-aged child)

(-1.02)

(-0.83)

Age

0.065

0.070

(1.13)

(1.22)

-0.050

-0.063

(-0.59)

(-0.74)

0.016

0.020

(0.25)

(0.31)

0.019

0.017

(0.67)

(0.61)

Mother finished secondary school or

0.049

0.060

above (1=yes, 0=no)

(0.67)

(0.82)

Father finished secondary school or

0.046

0.046

above (1=yes, 0=no)

(0.67)

(0.67)

Household Wealth
2nd quartile
rd

3 quartile
th

4 quartile
Individual Characteristics

Student*Age
Gender (1=male, 0=female)
Number of Siblings

Health and Sanitation Habits

Wash hands before dinner (1=yes,

-0.046

0=no)

(-0.60)

Wash hands after using toilet

0.054

(1=yes, 0=no)

(0.69)

Eat uncooked vegetables (1=never,

-0.127

0 otherwise)

(-1.72)

Eat uncooked meat (1=never, 0

0.068

otherwise)

(0.84)

Drink un-boiled water (1=never, 0

-0.125

otherwise)

(-1.38)

Constant
Sample Size

-0.058

-0.023

-0.074

(-1.01)

(-0.10)

(-0.29)

1707

1707

1724
;*

Note: t statistics in parentheses p < 0.10,

**

p < 0.05,

***

p < 0.01

The first quartile (omitted) represents households with the fewest assets and the fourth quartile

34

represents households with the most assets.

The coefficients of the probit regressions are reported in the table.

35

Table 8. Infection Rate by Minority Status and County

All
-Han
-Non-Han
P value

County 4
Sample size
Infection rate
309
22.01%
59
11.86%
250
24.40%
0.0367

Note:
1. P value of t-statistics of difference is reported.

County 5
Sample size
Infection rate
323
45.20%
176
46.02%
147
44.22%
0.7464

County 6
Sample size
Infection rate
281
49.64%
265
48.86%
16
62.50%
0.2911

Table 9. Deworming Rate by Minority Status and County

All
-Han
-Non-Han
P value

County 4
Sample size
Infection rate
309
43.69%
59
50.85%
250
42.00%
0.2191

Note:
1. P value of t-statistics of difference is reported.

County 5
Sample size
Infection rate
323
49.54%
176
43.18%
147
57.14%
0.0124

County 6
Sample size
Infection rate
281
60.85%
265
59.09%
16
93.75%
0.0056

Table 10. Reinfection Rate by Minority Status and County (Children who tested positive for infection among those who reported taking
deworming medication within the 18 months prior to the survey)

All
-Han
-Non-Han
P value

County 4
Sample size
Infection rate
135
23.70%
30
13.33%
195
26.67%
0.1319

Note:
1. P value of t-statistics of difference is reported.

County 5
Sample size
Infection rate
160
36.88%
76
35.53%
84
38.09%
0.7386

County 6
Sample size
Infection rate
171
46.20%
156
44.87%
15
60.00%
0.2643

Endnotes

The asset index is broken into four quartiles. The first quartile represents households with the fewest
assets and and the fourth quartile represents households with the most assets.
ii
Wang et al. (2012) surveyed three counties in Sichuan and three counties in Guizhou. During the
study they found that the overall prevalence of intestinal infections were 40.1% for school-aged
children and 33.9% for preschool-aged children. The sample studied in our paper included the three
Guizhou counties studied by Wang et al. (2012). In this study, however, we have added three additional
counties with minority populations in Guizhou.