Potential impact of climate change and water resources development on the epidemiology of schistosomiasis in China

Schistosomiasis japonica, caused by the blood fluke Schistosoma japonicum, has been
endemic in China since ancient times. An estimated 11 million people were infected in the
mid-1950s. Recognizing the huge public health significance and the economic impact of the
disease, the central government of China implemented a large-scale control programme, which
has been sustained and constantly adapted over the past half century. Today, the endemic areas
are mainly confined to the lake and marshland regions along the Yangtze River in five
provinces, namely Jiangsu, Anhui, Jiangxi, Hunan and Hubei. It is estimated that currently
about 800,000 people are infected and that 40 million people are at risk of infection.
Historically, the northern geographical limit where schistosomiasis transmission occurred was
around the 33°15’ N latitude (e.g. in Baoying county, Jiangsu province), governed by low
temperature thresholds.
Based on various climate models, the Intergovernmental Panel of Climate Change (IPCC)
recently concluded that the Earth has warmed by approximately 0.6°C over the past 100 years.
This unusual warming has been particularly pronounced during the last three decades. There is
growing consensus that the global trend of climate warming will continue in the 21st century. It
has been suggested that climate change could impact on the distribution of the intermediate
host snail of S. japonicum, i.e. Oncomelania hupensis.
The frequency and transmission dynamics of schistosomiasis can also be affected by waterresource
development and management. Among others, the South-to-North Water Transfer
(SNWT) project” is currently under construction in China, which intends to divert water from
South (the snail-infested Yangtze River) to North (Beijing and Tianjing) via the lakes of
Gaoyou, Hongze and others. The implementation and operation of this project could further
amplify the negative effects of climate change and facilitate the northward spread of
O. hupensis.
The main objective of this PhD thesis was to explore the potential impact of climate change
and the SNWT project on the future distribution of schistosomiasis japonica, particularly in
eastern China. The techniques used were geographic information system (GIS) and remote
sensing (RS), coupled with Bayesian spatial statistics, which have become key tools for disease
mapping and prediction.
First, we reviewed the application of GIS/RS techniques for the epidemiology and control
of schistosomiasis in China. The applications included mapping prevalence and intensity data
of S. japonicum at a large scale, and identifying and predicting suitable habitats for O. hupensis
at a small scale. Other prominent applications were the prediction of infection risk due to
ecological transformations, particularly those induced by floods and water-resource
development projects, and the potential impact of climate change. We discussed the limitations
of the previous work, and outlined potential new applications of GIS/RS techniques, namely
quantitative GIS, WebGIS and the utilization of emerging satellite-derived data, as they hold
promise to further enhance infection risk mapping and disease prediction. We also stressed
current research needs to overcome some of the remaining challenges of GIS/RS applications
for schistosomiasis, so that further and sustained progress can be made towards the ultimate
goal to eliminate the disease from China.
Second, recognizing the advantages of combining GIS/RS techniques with advanced spatial
statistical approaches, we developed Bayesian spatio-temporal models to analyze the
relationship between key climatic factors and the risk of schistosomiasis infection. We used
parasitological data collected annually from 1990 to 1998 by means of cross-sectional surveys
carried out in 47 counties of Jiangsu province. Climatic factors, namely land surface
temperature (LST) and normalized difference vegetation index (NDVI), were obtained from
satellite sensors. Our analysis suggested a negative association between NDVI and the risk of
S. japonicum infection, whereas an increase in LST contributed to a significant increase in
S. japonicum infection prevalence.
Third, in order to better understand the changes in the frequency and transmission
dynamics of schistosomiasis in a warmer future China, a series of laboratory experiments were
conducted to assess the effect of temperature on the parasite-intermediate host snail interaction.
We found a positive linear relationship between the development of. S. japonicum larvae
harboured in O. hupensis and temperature. In snails kept at 15.3°C, S. japonicum larvae tend to
halt their development, while peak development occurs at 30°C. The temperature at which half
of the snails were in hibernation is 6.4°C. A statistically significant positive association was
observed between temperature and oxygen intake of O. hupensis at temperatures below 13.0°C.
We also detected a logistic relationship between snails’ oxygen intake and their hibernation
rate. Our results underscored the important role temperature plays both for the activity of
O. hupensis and the development of S. japonicum larvae harboured in the intermediate host
snail.
Fourth, to substantiate the claim that global warming might alter the frequency and
transmission dynamics of S. japonicum in China, we conducted a time-series analysis from
1972-2002, using temperature data from 39 counties of Jiangsu province. Using annual
growing degree days (AGDDs) with a temperature threshold of 15.3°C, we forecasted changes
in S. japonicum transmission. The final model included a temporal and a spatial component.
The temporal trend consisted of second order polynomials in time plus a seasonality
component, while the spatial trend was formed by second order polynomials of the coordinates
plus the thin plate smoothing splines. The AGDDs of S. japonicum in 2003 and 2006 and their
difference were calculated. The temperatures at the 39 locations showed an increasing temporal
trend and seasonality with periodicities of 12, 6 and 3 months. The predicted AGDDs increased
gradually from north to south in both 2003 and 2006. The increase in AGDD was particularly
pronounced in the southern part of the study area. Our results suggest that alterations in the
transmission intensity of S. japonicum in south Jiangsu will be more pronounced than in the
northern part of the province.
Fifth, we further assessed the potential impact of climate change on the distribution of
O.hupensis via a spatially-explicit analytical approach. We employed two 30-year composite
datasets comprising average monthly temperatures collected at 623 meteorological stations
throughout China, spanning the periods 1961-1990 and 1971-2000. Temperature changes were
assessed spatially between the 1960s and the 1990s for January, as this is the critical month for
survival of O. hupensis. Our results show that the mean January temperatures increased at 590
stations (94.7%), and that China’s average January temperature in the 1990s was 0.96°C higher
than 30 years earlier. The historical 0-1°C January isotherm, which has been considered the
approximate northern limit of S. japonicum transmission, has shifted from 33°15’ N to 33°41’
N, expanding the potential transmission area by 41,335 km2. This translates to an estimated
additional 21 million people at risk of schistosomiasis. Two lakes that form part of the SNWT
project are located in this new potential transmission area, namely Hongze and Baima.
Finally, we applied GIS/RS techniques to predict potentially new snail habitats around
the lakes of Hongze and Baima, as well as Gaoyou lake, which is considered as a habitat where
O. hupensis could re-emerge. A model based on flooding areas, NDVI and a wetness index
extracted from Landsat images was developed to predict the snail habitats at a small scale. A
total of 163.6 km2 of potential O. hupensis habitats were predicted around the three study lakes.
In conclusion, our work suggests that global warming and a major water-resource
development project could impact on the distribution of S. japonicum and its intermediate host
snail in China and demonstrates that the combination of GIS, RS and Bayesian spatial
statistical methods is a powerful approach in estimating their extent. The predictions can serve
as a basis for health policy makers and disease control managers, and can be of use in the
establishment and running of schistosomiasis surveillance systems. It is further suggested that
an efficient early warning system should be set up in potential new endemic areas to monitor
subtle changes in snail habitats due to climate change and major ecological transformations,
and to assure the early detection of emerging and re-emerging schistosomiasis.