Association between malaria incidence and meteorological factors: a multi-location study in China, 2005-2012.

This study aims to investigate the climate-malaria associations in nine cities selected from malaria high-risk areas in China. Daily reports of malaria cases in Anhui, Henan, and Yunnan Provinces for 2005-2012 were obtained from the Chinese Center for Disease Control and Prevention. Generalized estimating equation models were used to quantify the city-specific climate-malaria associations. Multivariate random-effects meta-regression analyses were used to pool the city-specific effects. An inverted-U-shaped curve relationship was observed between temperatures, average relative humidity, and malaria. A 1 °C increase of maximum temperature (T max) resulted in 6·7% (95% CI 4·6-8·8%) to 15·8% (95% CI 14·1-17·4%) increase of malaria, with corresponding lags ranging from 7 to 45 days. For minimum temperature (T min), the effect estimates peaked at lag 0 to 40 days, ranging from 5·3% (95% CI 4·4-6·2%) to 17·9% (95% CI 15·6-20·1%). Malaria is more sensitive to T min in cool climates and T max in warm climates. The duration of lag effect in a cool climate zone is longer than that in a warm climate zone. Lagged effects did not vanish after an epidemic season but waned gradually in the following 2-3 warm seasons. A warming climate may potentially increase the risk of malaria resurgence in China.

Experts' Perceptions on China's Capacity to Manage Emerging and Re-emerging Zoonotic Diseases in an Era of Climate Change.

Zoonotic diseases transmitted by arthropods and rodents are a major public health concern in China. However, interventions in recent decades have helped lower the incidence of several diseases despite the country's large, frequently mobile population and socio-economic challenges. Increasing globalization, rapid urbanization and a warming climate now add to the complexity of disease control and prevention and could challenge China's capacity to respond to threats of emerging and re-emerging zoonoses. To investigate this notion, face-to-face interviews were conducted with 30 infectious disease experts in four cities in China. The case study diseases under discussion were malaria, dengue fever and haemorrhagic fever with renal syndrome, all of which may be influenced by changing meteorological conditions. Data were analysed using standard qualitative techniques. The study participants viewed the current disease prevention and control system favourably and were optimistic about China's capacity to manage climate-sensitive diseases in the future. Several recommendations emerged from the data including the need to improve health literacy in the population regarding the transmission of infectious diseases and raising awareness of the health impacts of climate change amongst policymakers and health professionals. Participants thought that research capacity could be strengthened and human resources issues for front-line staff should be addressed. It was considered important that authorities are well prepared in advance for outbreaks such as dengue fever in populous subtropical areas, and a prompt and coordinated response is required when outbreaks occur. Furthermore, health professionals need to remain skilled in the identification of diseases for which incidence is declining, so that re-emerging or emerging trends can be rapidly identified. Recommendations such as these may be useful in formulating adaptation plans and capacity building for the future control and prevention of climate-sensitive zoonotic diseases in China and neighbouring countries.

Child mortality patterns in rural areas of Anhui and Henan provinces in China, 1990.

County-based IMR and U5MR in Anhui and Henan provinces in China were estimated and analyzed by using the 1990 Census Data. Census was conducted on July 1, 1990, the number of deaths only occurred in the first half year of 1990 was collected. In order to obtain the total population and total number of deaths in the same year, the total number of deaths in each age-sex group for the whole 1990 was then estimated by taking the death number in the first half of 1990 as the base and multiplying a coefficient, which varied in different age-sex-region groups. Two major adjustments for some possible under-reporting cases in female birth and infant death were made. If the sex ratio at age 0 in some counties was beyond 1.2, then it was taken as 1.15 for rural counties and 1.10 for urban cities, which were the estimates of sex ratios for the children at age 5 in the national 1% Population Sampling Survey in 1995. The adjustment for IMR were made by comparing the segment of the county lift table from age 15 through 59 with that from the same age groups in the international and Chinese Model Life Tables. The IMR in the county life table would be substituted by the one in the closest Model Life Talbe, if it was less than in the latter. The findings of the analysis may be summarized as follows: (i) Total county-based IMR and U5MR were 33.4 per 1,000 and 41.4 per 1,000 respectively, with great variations between urban cities (25.4 per 1,000 for IMR and 31.4 per 1,000 for U5MR) and rural counties (35.1 per 1,000 for IMR and 43.6 per 1,000 for U5MR). There were also significant differences in child mortality between nationally identified poor counties and other counties in rural areas. In the poor counties the total IMR was 40.7 per 1,000 living births in average while in non-poor counties it was only 33.2 per 1,000 in average (P < 0.05). The U5MR in poor counties was 25 percent higher than in non-poor counties (51.5 vs 40.9 per 1,000 living births). (ii) Statistically significant correlation between child mortality and socio-economic variables was revealed from the data set, among which gross social economic products per capita was found to have the strongest relationship with child mortality. The negative correlation was found between child mortality and a set of so-called 'rich' variables including the gross social products, gross agricultural products, gross industrial products and the proportions of high-educated population at county level, whereas the positive correlation was found between child mortality and a set of 'poor' variables, such as proportions of residents with lower level of education and illiteracy rate. (iii) Differences in child mortality between these two provinces were found, which were identical to the trends of differences in socio-economic indicators between them. Lower child mortality proved to be associated with better socio-economic conditions (higher per capita products, higher proportions of residents with higher level of education, lower proportion of less educated people and illiteracy) in province Henan. (iv) A simple linear regression model was developed separately for Henan and Anhui to predict the IMR and U5MRs in each stage of economic development, where the dependent variables were the logarithm of IMR and U5MR, and the independent variables were the quintiles of the output value of gross products (GOP). It was found that at the first quintile, which was equivalent to 800 yuan of GOP in average, the predicted IMR and U5MR would reach 40 per 1,000 and 51 per 1,000 respectively. It would decline to 38 per 1,000 for IMR and 47 per 1,000 for U5MR in the second lowest quintile. Dramatic drop of child mortality was found between the second quintile and the third quintile, where 6 per 1,000 decline would occur for both IMR and U5MR. The decline would continue subsequently, but slower. The prediction of child mortality in rural counties could be used as a reference to assess counties at different stages of socio-