- Email: [email protected]
Temperature, hospital admissions and emergency room visits in Lhasa, Tibet: A time-series analysis
Science of the Total Environment 490 (2014) 838–848
Contents lists available at ScienceDirect
Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv
Temperature, hospital admissions and emergency room visits in Lhasa, Tibet: A time-series analysis Li Bai a, Cirendunzhu b, Alistair Woodward c, Dawa b, Zhaxisangmu b, Bin Chen a, Qiyong Liu a,d,e,⁎ a State Key Laboratory for Infectious Disease Prevention and Control, National Institute for Communicable Disease Control and Prevention, Chinese Center for Disease Control and Prevention, 155 Changbai Road, Changping District, Beijing 102206, PR China b Tibet Center for Disease Control and Prevention, 21 Linkuo North Road, Lhasa, Tibet 850000, PR China c School of Population Health, University of Auckland, Private Bag 92019, Auckland 1142, New Zealand d Shandong University Climate Change and Health Center, 44 Wenhua Road, Jinan, Shangdong 250012, PR China e Collaborative Innovation Center for Diagnosis and Treatment of Infectious Diseases, Hangzhou 310003, PR China
H I G H L I G H T S • • • •
This is the first to examine temperature effects on morbidities in Tibet. Heat were more strongly associated with morbidity than were cold. Heat were associated with increases of total emergency room visits. Temperature effects on hospitalizations were cause- and population-specific.
a r t i c l e
i n f o
Article history: Received 16 January 2014 Received in revised form 30 April 2014 Accepted 6 May 2014 Available online xxxx Editor: Lidia Morawska Keywords: Emergency room visits Hospital admissions Morbidity Temperature Tibet Vulnerability
a b s t r a c t Background: Tibet of China, with an average altitude of over 4000 m, has experienced noticeable changes in its climate over the last 50 years. The association between temperature and morbidity (most commonly represented by hospital admissions) has been documented mainly in developed countries. Little is known about patterns in China; nor have the health effects of temperature variations been closely studied in highland areas, worldwide. Objective: We investigated the temperature–morbidity association in Lhasa, the capital city of Tibet, using sexand age-specific hospitalizations, excluding those due to external causes. Methods: A distributed lag non-linear model (DLNM) was applied to assess the nonlinear and delayed effects of temperature on morbidity (including total emergency room visits, total and cause-specific hospital admissions, sex- and age-specific non-external admissions). Results: High temperatures are associated with increases in morbidity, to a greater extent than low temperatures. Lag effects of high and low temperatures were cause-specific. The relative risks (RR) of high temperature for total emergency room visits and non-external hospitalizations were 1.162 (95% CI: 1.002–1.349) and 1.161 (95% CI: 1.007–1.339) respectively, for lag 0–14 days. The strongest cumulative effect of heat for lag 0–27 days was on admissions for infectious diseases (RR: 2.067, 95% CI: 1.026–4.027). Acute heat effects at lag 0 were related with increases of renal (RR: 1.478, 95% CI: 1.005–2.174) and respiratory diseases (RR: 1.119, 95% CI: 1.010–1.240), whereas immediate cold effects increased admission for digestive diseases (RR: 1.132, 95% CI: 1.002–1.282). Those ≥65 years of age and males were more vulnerable to high temperatures. Conclusion: We provide a first look at the temperature–morbidity relationship in Tibet. Exposure to both hot and cold temperatures resulted in increased admissions to hospital, but the immediate causes varied. We suggest that initiatives should be taken to reduce the adverse effects of temperature extremes in Tibet. © 2014 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/).
⁎ Corresponding author at: State Key Laboratory for Infectious Disease Prevention and Control, National Institute for Communicable Disease Control and Prevention, Chinese Center for Disease Control and Prevention, 155 Changbai Road, Changping District, Beijing 102206, PR China. E-mail addresses: [email protected] (L. Bai), [email protected] (Cirendunzhu), [email protected] (A. Woodward), [email protected] (Dawa), [email protected] (Zhaxisangmu), [email protected] (B. Chen), [email protected] (Q. Liu).
http://dx.doi.org/10.1016/j.scitotenv.2014.05.024 0048-9697/© 2014 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/).
L. Bai et al. / Science of the Total Environment 490 (2014) 838–848
1. Introduction
2. Methods
Variations in temperature affect human health in many ways (IPCC, 2014). The effects of both high and low temperatures on short-term increases in daily mortality and morbidity have been reported worldwide, particularly in developed countries, including the US (Barnett, 2007; Basu and Ostro, 2008; Braga et al., 2002), Canada (Pengelly et al., 2007), Europe (Aström et al., 2013; Vandentorren et al., 2006), and Australia (Vaneckova and Bambrick, 2013; Vaneckova et al., 2008). Temperature-sensitive health outcomes include chronic conditions such as cardiovascular diseases, respiratory diseases and renal diseases, as well as infectious diseases (vector-borne and water-borne) (IPCC, 2014). Importantly, factors that influence susceptibility to temperature tend to vary from one location to another and we note that little is known about how temperature and other aspects of weather affect human health in highland areas. Tibet of China, lying at an average altitude of more than 4000 m, is called “the third pole of the world”. Temperatures on the “roof of the world” have been increasing by 0.32 °C every decade in winters from 1955 to 1996, a much faster rate of change than has been observed in China, or in Asia generally (Liu and Chen, 2000). Accordingly Tibet is regarded as one of the most vulnerable areas to climate change in the world (Du et al., 2011; Yu et al., 2012a). However, in common with many other developing regions, nothing has been documented to date about the potential effects on health of long-term changes in climate, with particular reference to vulnerable subgroups. The capital city of Tibet, Lhasa, has experienced considerable changes in climate over the past 50 years. Maximum temperatures in Lhasa reached 30.4 °C in 2009 summer. Before this, the highest record is 29.9 °C in 1971. A cross-sectional survey of 619 respondents from urban Lhasa in 2012 summer found that many local residents are aware of the warming that has occurred in recent years (Bai et al., 2013). Over 78% reported that rising temperature is either a “very” or “somewhat” serious threat to their own health, and nearly 40% reported that they had experienced heat-related symptoms. The next step is to relate temperature variations to local measures of health and identify those in the population who are most affected, in order to target interventions and develop Tibetan-specific public health programs. The objective of this study was to investigate the relationship between temperature and morbidity in Lhasa, using sex- and agespecific hospital admissions excluding those due to external causes.
2.1. Data Collection
839
Lhasa, with an elevation of about 3680 m, is located in a flat river valley surrounded by the Himalaya Mountains rising to 5500 m (Picture 1). In 2010 the population of the city numbered 559,400 (287,400 men, 272,000 women). We obtained records of hospital admissions (1 January 2005 to 31 December 2012) and emergency room visits (1 March 2005 to 31 December 2012) from the People's Hospital of Tibet Autonomous Regions, which is located at the center of the city and close to the Potala Palace (Picture 1). This is the biggest public hospital in Tibet, with advanced facilities, more than 500 beds and over 900 staff. It is the only hospital in Tibet with a well-managed electronic medical record system, which can provide daily data from 2005 to the present, and the People's Hospital is also the first choice of most of the local residents whenever they need to see a doctor or have acute health problems. Hospitalization data included principal diagnosis, admission and discharge dates, total charges, date of birth, sex, ethnicity, and residential street address. Based on the International Classification of Disease, 10th Revision, Clinical Modification (ICD-10) we classified the data into five cause-specific categories: non-external hospital admission (A00-R99); infectious, (A00-B99) cardiovascular (I00-99), respiratory (J00-99), renal (N00-28) and digestive (K20–31, 50–52, 55–63, 92). For emergency room visit data, we obtained only the total daily counts with neither causes nor patients' information. Daily meteorological data for Lhasa were provided by the National Climate Center from 2005 to 2012. The variables included daily maximum, mean and minimum temperature and relative humidity.
2.2. Data Analysis Previous studies have shown that temperature effects on human health are frequently delayed in time. In this study, a distributed lag non-linear model (DLNM) was fitted to examine the relationship between temperature and daily numbers of hospital admissions and emergency room visits. The main advantage of this method is that it is flexible enough to simultaneously describe a non-linear exposure– response association and delayed effects or harvesting (Gasparrini, 2011; Gasparrini et al., 2010). Most recently, DLNM has been applied in studies to quantify the effects of temperature (Guo et al., 2011; Kim et al., 2012; Lin et al., 2013) and air pollution (Goldberg et al., 2013) on mortality.
Picture 1. An aerial photograph of Lhasa City, Tibet. (The red triangle represents the location of Potala Palace; the blue one represents the location of the People's Hospital of Tibet Autonomous Regions). (For interpretation of the references to color in this picture legend, the reader is referred to the web version of this article).
840
L. Bai et al. / Science of the Total Environment 490 (2014) 838–848
We controlled for seasonality and long-term trend using a natural cubic spline with 7 df per year for time. To control any confounding weekly pattern, the day of week was also included as an indicator in our analysis. Public holiday was controlled as a binary variable. We did not control for relative humidity, as levels of water vapor in the air in Tibet are low across the year. We evaluated the model fit using Akaike's Information Criterion for quasi-Poisson (Q-AIC). The results showed that models that controlled humidity had higher Q-AIC values, indicating that models excluding humidity were a better fit (results not shown). We used maximum temperature to assess temperature effects in this analysis since it was found to be a better predictor (i.e. had lower Q-AIC values) than the mean and minimum temperatures. We used a DLNM with 5 degrees of freedom natural cubic for temperature (knots at equally-spaced percentiles by default) and with 4 degrees of freedom natural cubic for lags (knots at equally-spaced values in the log scale of lags by default) (Guo et al., 2011; Tian et al., 2012). The median value of the maximum temperature (17.7 °C) was used as the reference value (centering value) to calculate the relative risks. A maximum lag of 27 was used to completely capture the overall temperature effect and adjust for a possible harvesting effect (Guo et al., 2011). We examined and plotted cumulative relative risks with temperature for lag 0, lags 0–2, 0–14 and 0–27. We also examined the hot and cold effects separately on relative risks of total emergency room visits, cause-specific hospital admissions, and sex- and age-specific non-accidental hospitalizations. For hot effects, we calculated relative risks associated with the 99th percentile of temperature (high temperature) relative to the 75th percentile of temperature. For cold effects, we calculated relative risks associated with the first percentile of temperature (cold temperature) relative to the 25th percentile of temperature (Guo et al., 2012). The cumulative effects of hot and cold temperatures along the lags were then estimated separately. Sensitivity analysis was conducted by changing the df for time from 6 to 12 per year, the maximum lag days from 14 to 30 days, and the df for temperature and lags from 3 to 6. All statistical tests and modeling were performed using the R software (version 3.0.1). Distributed lag nonlinear models were fitted through “dlnm” package (Gasparrini, 2011).
3. Results In all, there were 82,491 hospital admissions from 1 January 2005 to 31 December 2012 and 159,139 emergency room visits from 1 March 2005 to 31 December 2012. Table 1 shows the statistical summary for Table 1 Summary statistics of daily weather conditions, hospital admissions and emergency room visits in Lhasa, Tibet during 2005 to 2012.
Mean temperature (°C) Maximum temperature (°C) Minimum temperature (°C) Relative humidity (%) Total emergency room visits Total hospital admissions Non-external Infectious Cardiovascular Respiratory Renal Digestive Demographica Male Female 0–15 years 16–44 years 45–64 years ≥65 years
Mean
SD
Min
25th
Median
75th
Max
9.6 17.4 3.3 33.7 56 28 22 1 2 4 1 2
6.6 6.1 7.4 17.2 19.0 12.6 10.6 1.3 1.8 2.8 1.1 1.4
−7.3 −0.9 −16.1 5.0 3 1 1 0 0 0 0 0
4.0 12.6 −3.2 19.0 43 18 13 0 1 2 0 0
9.95 17.7 3.4 30.0 53 27 21 1 2 4 1 1
15.4 22.5 10.1 46.0 67 37 30 2 3 6 1 2
22.6 30.4 18.2 84.0 167 76 65 8 10 24 8 10
10 11 4 9 5 2
14 15 6 13 8 4
10 11 4 10 5 3
5.9 5.7 2.5 5.0 3.8 2.1
0 0 0 0 0 0
6 7 2 6 2 1
36 33 14 29 20 13
a Demographic variables are only applicable to the hospital admissions for non-external diseases.
hospital data and weather condition. The average daily maximum temperature was 17.4 °C; the mean temperature was 9.6 °C; the minimum temperature was 3.3 °C. Fig. 1 displays the full data series for temperature, total emergency room admissions and hospitalizations. The mean, maximum and minimum temperatures varied strongly with season. There was also a clear seasonal trend of total emergency room visits, with more patients at high temperature days. The daily number of hospitalizations has increased slightly since 2008 and we observed a weak association between temperature and hospitalizations after 2009, with more cases in summer than in winter. The 3-D plots show the non-linear relationships between maximum temperature and total emergency room admissions, as well as causespecific hospitalizations along 27 lag days (Fig. 2). For total hospital admissions, there was no evident effect of temperature, but both low and high temperatures were associated with increased numbers of total emergency room visits. For cause-specific hospitalizations, admissions for non-external, respiratory and renal diseases had higher relative risks at hot temperature, but relative risks of admissions for cardiovascular and digestive diseases were higher at cold temperatures. A delayed heat effect was observed for infectious disease admissions. Fig. 3 shows the temperature effects of current day (lag 0) and cumulative effects on total emergency room admissions, total and causespecific hospital admissions at lags 0–2, 0–14, and 0–27. For the effects of extreme hot temperature, the cumulative risks on total emergency room visits, non-external and infectious disease hospitalizations at lag 0–27 were higher than those at lag 0, lags 0–2 and 0–14, while the cumulative risk on admissions for renal diseases and respiratory diseases was highest at lag 0 and lag 0–2 respectively. For the effects of extreme cold temperature, the cumulative risks on total emergency room admissions and cause-specific hospitalizations were highest at lag 0–27, except for renal and respiratory admissions. We plotted the hot and cold effects of temperature separately on total emergency room visits and cause-specific hospitalizations along the lag days (Figs. 4, 5). Lag effects for different admission categories were quite different. For example, heat effects occurred immediately for total emergency room visits and hospitalizations for respiratory and renal diseases, whereas heat effects appeared after a 5-day lag for infectious diseases and a 7-day lag for cardiovascular diseases. Acute cold effects were also observed for total emergency room visits and digestive diseases, whereas cold effects occurred after a 2-day lag for total hospital admissions, cardiovascular and renal admissions. Fig. 6 shows the hot and cold effects on sex- and age-specific non-accidental hospitalizations along the lags. Immediate heat effects were found in all sub-populations except for those aged 16–44 years old. In addition, heat effects lasted longer in those aged 45–64 years and older than 65 years. Cold effects increased at long lag days for all sub-groups except for those aged 0–15 years old. The cumulative hot effects on total emergency room visits, causespecific hospitalizations, and sex- and age-specific non-external hospitalizations were calculated along the lags (Table 2). For the short lags (lag 0 and lag 0–2), heat effects were apparent for renal and respiratory diseases, whereas for the long lags (lag 0–14 and lag 0–27), high temperatures were associated with the risk of total emergency room visits, and hospitalizations for non-external and infectious diseases. The highest cumulative risk of heat effects was observed on infectious diseases at lag 0–27 (RR 2.067, 95% CI: 1.026–4.027). For sex- and agespecific non-external hospitalizations, males were more affected than females, and the elderly were more sensitive to hot temperatures than other age groups. The cumulative cold effects on total emergency room visits, causespecific hospitalizations, and sex- and age-specific non-external hospitalization were calculated along the lags (Table 3). The relative risk of admissions for digestive diseases at lag 0 was the only statistically significant result (RR: 1.132, 95% CI: 1.002–1.282). The highest risk of cold effect on demographic-specific non-external diseases was
L. Bai et al. / Science of the Total Environment 490 (2014) 838–848
841
Fig. 1. Temperature, total emergency room visits and hospital admissions in Lhasa, Tibet during 2005 to 2012.
observed in those older than 65 years at lag 0–27, although the result is not statistically significant. 3.1. Sensitivity Analysis We changed the df (6–12 per year) for time to control seasonality, which produced similar results (data not shown). We also altered the maximum lag days from 14 to 31 days, and found that the temperature effects were similar (data not shown). Additionally, the df for temperature and lags from 3 to 6 was changed, which also gave similar results. 4. Discussion In this study, we have analyzed the associations between daily maximum temperature and emergency room visits and hospital admissions in Lhasa, Tibet during 2005–2012. It appears that elevated temperatures affect morbidity to a greater degree than do low temperatures. High temperatures were associated with increases of total emergency room visits and hospitalizations for infectious, renal and respiratory diseases, while cold temperatures were associated with more frequent digestive disease admissions. Lag effects of both high and low temperatures were cause-specific. Acute temperature effects were observed with renal, respiratory diseases and digestive diseases, whereas delayed effects were associated with increased total emergency room visits, and hospitalizations for non-external and infectious diseases. We also observed some signs of increased vulnerability to high temperatures among those ≥65 years of age and males. This study is, to our best knowledge,
the first to examine temperature-related morbidity and vulnerable subpopulations in Tibet of China, or any other highland area. The Tibetan plateau, a part of the world that is particularly sensitive to environmental disruption (Yu et al., 2012a), has experienced rapid changes in its climate (Du et al., 2011; Liu and Chen, 2000). Others have suggested that heat exposure tends to be more troublesome in cold regions, whereas warm regions may suffer more from cold weather (Yang et al., 2013). Our findings are consistent with this since heat had a greater impact on human health than cold, and Lhasa experiences mild summer climates and extreme cold winter temperatures. It may be that people living in relatively cold regions have low awareness and limited coping resources (such as air-conditioning) to deal with the heat. As might be expected, cooling devices are not common in Lhasa: a survey of 619 respondents which we conducted in August 2012 found that 8.7% of participants had air-conditioning at home, 18.1% reported they had a fan and about 6% had both (Bai et al., 2013). Moreover, many of those who had air-conditioning at home reported that they seldom use the equipment, due to concerns about power costs or other unwanted effects of cooling. Others have reported immediate effects of ambient temperatures on total and cause-specific mortalities (Guo et al., 2012; Huang et al., 2014; Tian et al., 2012) and measures of morbidity (Chan et al., 2013; Wichmann et al., 2011; Wichmann et al., 2013; Zanobetti et al., 2013). In our analysis, we found both hot and cold effects with a range of lag periods. The risk of acute heat effects (lag 0) was highest for renal diseases (RR: 1.478, 95% CI: 1.005–2.174). This was consistent with previous studies which found that the largest hot effects on renal diseases occurred at lag 0 (Green et al., 2009) or lag 1 (Fletcher et al., 2012).
842
L. Bai et al. / Science of the Total Environment 490 (2014) 838–848
Total emergency room visits
Total hospital admissions
Non-external diseases*
Infectious diseases*
Cardiovascular diseases*
Respiratory diseases*
Renal diseases*
Digestive diseases*
Fig. 2. The three-dimensional plot of the association between maximum temperature and morbidity along 27 lags (*cause-specific hospital admissions).
Some studies have documented that the biologic mechanism underlying the relationship between high temperature and renal diseases, which disrupted fluid balance during heat exposure leads to acute renal failure and the development of renal calculi (Brikowski et al., 2008; Chen et al., 2008; Lo et al., 2010). Consistent with other studies (Chan et al., 2013; Lin et al., 2009), we also observed a short heat effect within 0–2 days on admissions for respiratory diseases (RR: 1.234, 95% CI: 1.034–1.472). Similarly, Lin et al. (2009) found that the greatest number of hospital admissions for respiratory diseases was 0–1 days after elevated temperatures.
Apart from short hot effects, delayed and longer lasting heat effects (lags 0–14 and 0–27) were observed with the increase of total emergency room visits, non-external hospitalizations and infectious disease admissions in this study. Most previous studies focused on short-term heat effects on cause-specific emergency department visits and hospital admissions such as total cardiovascular diseases (Ren and Tong, 2006; Michelozzi et al., 2009), total respiratory diseases (Lavigne et al., 2014; Lin et al., 2009), acute myocardial infarction (Ebi et al., 2004; Wichmann et al., 2013) and stroke (Dawson et al., 2008; Ohshige et al., 2006). However, a few confirmed the associations with total emergency room visits or total non-external hospitalizations. For instance, a study in UK found no relation between total emergency room visits and heat when using short-term temperatures (three-day moving average temperature) (Kovats et al., 2004). For infectious disease admissions, we also observed that the cumulative effect due to high temperature could last 27 days (RR: 2.067, 95% CI: 1.026–4.027). Previous studies have reported that the transmission and occurrence of some climate-sensitive infectious diseases such as malaria (Midekisa et al., 2012; Zhang et al., 2012), dengue (Lu et al., 2009), diarrhea (Alexander et al., 2013; Constantin de Magny et al., 2008), dysentery (Zhang et al., 2008) and salmonellosis (Grjibovski et al., 2012) were positively linked with one-month lag temperatures. In Tibet, rapid warming may introduce new threats, particularly those due to vector-borne diseases. With increased summer temperatures in Lhasa, we have already observed that mosquitoes in the Culex pipiens complex persist from one season to the next, and now appear to be established in what was previously recognized as a mosquitofree area due to its extreme altitude (Liu et al., 2013). Communicable diseases that may be transmitted by this mosquito species include Japanese encephalitis, meningitis, and urticaria. In 2009 Japanese encephalitis virus was isolated from Culex tritaeniorhynchus mosquitoes elsewhere in Tibet, indicating that JEV is already circulating in highaltitude regions (Li et al., 2011). On the basis of these changes we suggest targeted actions and programs such as specific disease monitoring, emergency preparedness, vector control and public education prior to the seasonal peak in mosquito numbers in Lhasa. In our analysis, acute cold effects at lag 0 were found to increase admissions for digestive diseases (RR: 1.132 95% CI: 1.002–1.282) which are ranked second in Lhasa in terms of both mortality and morbidity (Bai et al., 2008; Li, 2010). According to the literature, the seasonal pattern of digestive disorders is clear, with a significant increase found in the winter months for some severe digestive diseases such as gastric and duodenal ulcer and bleeding from the upper gastrointestinal tract (Du et al., 2010; Langma, 1964). It has been hypothesized that exposure to the cold tends to stimulate human digestive processes and worsen digestive symptoms by constricting the peripheral blood vessels, leading to congestion of the visceral vessels and increased secretions. However, few studies, to date, have explored the relationship between temperature and digestive diseases. Fernández-Raga et al. (2010) reported that the most comfortable temperatures for patients with digestive diseases (19.7 °C) are apparently higher than those for patients with cardiovascular diseases (16.8 °C) and respiratory diseases (18.1 °C) (Fernández-Raga et al., 2010). A study carried out in Beijing found that diurnal temperature range is an independent risk factor for emergency room admissions for digestive diseases among elderly persons (Wang et al., 2013), which implies that sudden modifications in the external temperature also react directly upon the digestive apparatus. Cardiovascular diseases are well-known to be sensitive to extreme temperatures (Basu, 2008). However, we found neither hot nor cold effects for this category of diseases. A review on recent temperature– morbidity studies concluded that sensitivity of cardiovascular diseases to temperatures varies by sub-types (Ye et al., 2012). Chronic and acute cardiovascular diseases were found to have opposite relationships to temperature increases in some studies (Green et al., 2009; Koken et al., 2003). The authors suggested that those with chronic longstanding conditions might avoid outdoor exposures during unpleasant
L. Bai et al. / Science of the Total Environment 490 (2014) 838–848
Lag0
Lag0-2
Lag0-14
843
Lag0-27
Total emergency room visits
Total hospital admissions
Non-external diseases*
Infectious diseases*
Cardiovascular diseases*
Respiratory diseases*
Renal diseases*
Digestive diseases*
Fig. 3. Exposure–response curves of maximum temperature and morbidities along the lags (*cause-specific hospital admissions).
844
L. Bai et al. / Science of the Total Environment 490 (2014) 838–848
Total emergency room visits
Total hospital admissions
Non-external diseases*
Infectious diseases*
Cardiovascular diseases*
Respiratory diseases*
Renal diseases*
Digestive diseases*
Fig. 4. The estimated hot effects of maximum temperature (99th relative to 75th percentile) on morbidities along the lags (*cause-specific hospital admissions).
weather, resulting in a null or negative correlation with temperature. Findings in our survey may also support this conclusion (Bai et al., 2013). In Lhasa, we found that residents with chronic diseases were more likely to perceive the risk on health of excessive heat and were more likely to alter their behaviors during hot summer days (e.g. drinking more fluids, staying indoors and avoiding the sun).
Vulnerability to temperature-related mortality and morbidity has been associated previously with socio-demographic characteristics such as age, sex, health, education, and occupation. We found that males and the elderly were at higher risk of morbidity associated with the heat. Some studies in other locations have reported that women had higher risks than men to extreme weather (Stafoggia et al., 2006;
L. Bai et al. / Science of the Total Environment 490 (2014) 838–848
Total emergency room visits
Total hospital admissions
Non-external diseases*
Infectious diseases*
Cardiovascular diseases*
Respiratory diseases*
Renal diseases*
Digestive diseases*
845
Fig. 5. The estimated cold effects of maximum temperature (1st relative to 25th percentile) on morbidities along the lags (*cause-specific hospital admissions).
Vaneckova et al., 2008; Yu et al., 2010). In Tibet, we note that men are more likely than women to suffer from chronic conditions including obesity (Yan et al., 2001), diabetes (Dawa et al., 2006; Yang et al., 2003), high blood pressure (Hu et al., 2003; Zheng et al., 2011), and
Alzheimer's disease (Zhao et al., 2002). In addition, Tibetan men are more likely to engage in outdoor physical activities. Consistent with others (Yu et al., 2012b), we observed a strong and significant effect of high temperatures on those older than 65 years. Nevertheless, we
846
L. Bai et al. / Science of the Total Environment 490 (2014) 838–848
Hot effects Male
Female
0-15 y
16-44 y
45-64 y
>=65 y
Male
Female
0-15 y
16-44 y
45-64 y
>=65 y
Cold effects
Fig. 6. The estimated hot and cold effects of maximum temperature on sex- and age-specific hospital admissions for non-external diseases along the lags.
suggest that special attention also should be paid to middle aged persons (45–64 years) since this age group in Tibet is commonly affected by chronic medical conditions (Bai et al., 2013; Hu et al., 2003). Also, unlike the elderly, most middle aged people are still at work and thus may be exposed to more unfavorable outdoor weather. Several limitations of the study should be noted. We did not have data on outdoor air pollution, which is independently associated with higher mortality, varies over short periods of time, and may modify the temperature–morbidity relationship. However, we note that air
quality in Lhasa is among the best in China. According to local environmental departments, the number of days with high air quality (Air Quality Index ≤ 50, which means that the daily average concentration of PM2.5 is 50 μg/m3 or lower) in 2012 was 361. Moreover, our results were based on analyzing one hospital data. This may result in bias in the estimated associations. Lastly, with this study design it was not possible to control other time-varying confounding variables such as intake of alcohol, smoking rates, use of medications and indoor air pollution due to residential heating in winters.
L. Bai et al. / Science of the Total Environment 490 (2014) 838–848
847
Table 2 The cumulative relative risks of hot effects on emergency room visits, cause-specific hospital admissions, and sex- and age-specific admissions for non-external diseases.
Total ERVa Total hospitalizations Non-external Infectious Cardiovascular Respiratory Renal Digestive Demographicb Male Female 0–15 16–44 45–64 ≥65
Lag 0
Lag 0–2
Lag 0–14
Lag 0–27
1.035 (0.956–1.122) 0.998 (0.897–1.111) 1.042 (0.996–1.091) 1.016 (0.871–1.184) 0.988 (0.867–1.127) 1.119 (1.010–1.240) 1.478 (1.005–2.174) 1.055 (0.921–1.208)
1.036 (0.951–1.128) 1.010 (0.886–1.151) 1.065 (0.984–1.153) 0.990 (0.757–1.293) 0.931 (0.730–1.189) 1.234 (1.034–1.472) 1.125 (0.791–1.597) 1.036 (0.818–1.313)
1.162 (1.002–1.349) 1.075 (0.836–1.382) 1.161 (1.007–1.339) 1.641 (1.026–2.207) 0.990 (0.563–1.740) 1.080 (0.808–1.444) 1.260 (0.704–2.253) 0.956 (0.632–1.449)
1.194 (0.951–1.500) 1.087 (0.728–1.623) 1.251 (1.011–1.549) 2.067 (1.026–4.027) 1.143 (0.449–2.900) 1.117 (0.696–1.792) 1.635 (0.655–2.760) 0.938 (0.472–1.864)
1.036 (0.975–1.102) 1.044 (0.986–1.105) 1.047 (0.962–1.140) 1.007 (0.950–1.068) 1.071 (0.988–1.162) 1.221 (1.023–1.457)
1.055 (0.949–1.172) 1.076 (0.975–1.187) 1.091 (0.942–1.263) 1.012 (0.914–1.119) 1.094 (0.951–1.260) 1.150 (1.002–1.321)
1.166 (1.002–1.460) 1.069 (0.910–1.256) 1.118 (0.876–1.426) 1.140 (0.964–1.347) 1.062 (0.842–1.340) 1.249 (0.910–1.712)
1.246 (0.939–1.652) 1.241 (0.956–1.612) 1.106 (0.742–1.647) 1.182 (0.901–1.551) 1.358 (0.931–1.978) 1.631 (0.984–2.703)
The bold-faced data means statistically significant (p b 0.05). a ERV: emergency room visits. b Demographic variables are only applicable to the hospital admissions for non-external diseases.
Table 3 The cumulative relative risks of cold effects on emergency room visits, cause-specific hospital admissions, sex- and age-specific admissions for non-external diseases.
a
Total ERV Total hospitalizations Non-external Infectious Cardiovascular Respiratory Renal Digestive Demographicb Male Female 0–15 16–44 45–64 ≥65
Lag 0
Lag 0–2
Lag 0–14
Lag 0–27
1.041 (0.975–1.110) 0.957 (0.912–1.005) 0.986 (0.994–1.030) 1.053 (0.905–1.226) 0.962 (0.851–1.087) 0.964 (0.893–1.040) 0.966 (0.690–1.350) 1.132 (1.002–1.282)
1.021 (0.943–1.105) 0.946 (0.891–1.004) 0.961 (0.890–1.038) 1.054 (0.810–1.373) 1.000 (0.802–1.248) 0.908 (0.792–1.040) 1.051 (0.737–1.498) 1.233 (0.998–1.531)
1.023 (0.877–1.194) 0.894 (0.792–1.008) 0.895 (0.788–1.017) 0.928 (0.592–1.455) 1.030 (0.656–1.618) 0.799 (0.616–1.037) 0.912 (0.498–1.670) 1.153 (0.776–1.713)
1.127 (0.890–1.428) 1.023 (0.823–1.287) 0.982 (0.810–1.191) 1.229 (0.627–2.407) 1.001 (0.462–2.166) 0.747 (0.499–1.119) 1.043 (0.405–2.448) 1.354 (0.733–2.503)
1.030 (0.972–1.091) 0.954 (0.904–1.007) 1.011 (0.936–1.092) 0.977 (0.923–1.035) 1.042 (0.964–1.126) 0.916 (0.792–1.059)
1.029 (0.931–1.138) 0.911 (0.830–1.002) 1.005 (0.879–1.150) 0.935 (0.846–1.034) 1.057 (0.922–1.210) 0.810 (0.675–0.973)
0.913 (0.771–1.081) 0.909 (0.772–1.070) 1.005 (0.800–1.262) 0.796 (0.673–0.940) 0.926 (0.738–1.163) 1.050 (0.773–1.427)
1.023 (0.793–1.319) 1.001 (0.780–1.286) 1.137 (0.801–1.613) 0.852 (0.662–1.095) 1.006 (0.715–1.415) 1.195 (0.757–1.887)
The bold-faced data means statistically significant (p b 0.05). a ERV: emergency room visits. b Demographic variables are only applicable to the hospital admissions for non-external diseases.
5. Conclusion Our study has added an important dimension to the growing body of work that documents the temperature–morbidity association, as there have been no reports previously from such a high-elevation setting. We found that hot temperatures were more strongly associated with morbidity in Lhasa than were cold temperatures. Heat effects were associated with increases of total emergency room visits, hospital admissions for non-external diseases, renal diseases and respiratory diseases, whereas cold temperatures were linked with digestive disease admissions. We suggest that initiatives are needed now to minimize the downside temperature effects in Tibet, especially in hot seasons, due to climate change and limited resources needed to cope. Particular attention should be paid to groups at greater risk, such as the elderly. Further studies are required to confirm our findings, to examine other population sensitivity factors and to include other locations in Tibet. Competing interests The authors declare that they have no competing interests. Acknowledgments This study was supported by the National Basic Research Program of China (973 Program) (Grant No. 2012CB955504). We are grateful to the
People's Hospital of Tibet Autonomous Regions for providing morbidity data. We also wish to thank Doctor Antonio Gasparrini from the London School of Hygiene and Tropical Medicine, UK for the guidance on statistical analysis. References Alexander KA, Carzolio M, Goodin D, Vance E. Climate change is likely to worsen the public health threat of diarrheal disease in Botswana. Int J Environ Res Public Health 2013;10:1202–30. Aström C, Orru H, Rocklöv J, Strandberg G, Ebi KL, Forsberg B. Heat-related respiratory hospital admissions in Europe in a changing climate: a health impact assessment. BMJ Open 2013;24:3. Bai G, Nimaqucuo, Gamacangjue, Guo S, Renqingyuzhen. Analysis of mortality in Tibet during 2004–2005. Chin J Prev Contr Chron Non-commun Dis 2008;16:525. Bai L, Cirendunzhu D, Pengcuociren D, Dawa D, Woodward A, Liu X, et al. Rapid warming in Tibet, China: public perception, response and coping resources in urban Lhasa. Environ Health 2013;12:71. Barnett AG. Temperature and cardiovascular deaths in the US elderly: changes over time. Epidemiology 2007;18:369–72. Basu R. High ambient temperature and mortality: a review of epidemiologic studies from 2001 to 2008. Environ Health 2008;8:40. Basu R, Ostro BD. A multicounty analysis identifying the populations vulnerable to mortality associated with high ambient temperature in California. Am J Epidemiol 2008;168:632–7. Braga AL, Zanobetti A, Schwartz J. The effect of weather on respiratory and cardiovascular deaths in 12 U.S. cities. Environ Health Perspect 2002;110:859–63. Brikowski TH, Lotan Y, Pearle MS. Climate-related increase in the prevalence of urolithiasis in the United States. Proc Natl Acad Sci U S A 2008;105:9841–6. Chan EY, Goggins WB, Yue JS, Lee P. Hospital admissions as a function of temperature, other weather phenomena and pollution levels in an urban setting in China. Bull World Health Organ 2013;91:576–84.
848
L. Bai et al. / Science of the Total Environment 490 (2014) 838–848
Chen YK, Lin HC, Chen CS, Yeh SD. Seasonal variations in urinary calculi attacks and their association with climate: a population based study. J Urol. 2008;179:564–9. Constantin de Magny G, Murtugudde R, Sapiano MR, Nizam A, Brown CW, Busalacchi AJ, et al. Environmental signatures associated with cholera epidemics. Proc Natl Acad Sci U S A 2008;105:17676–81. Dawa Ye XM, Dunzhuduoji, Gesang, Gamacangjue, Gou XQ. Study on the prevalence of major chronic diseases of adults in 2002 in Lhasa, Tibet. Food Sci Technol 2006;10: 19–21. [In Chinese]. Dawson J, Weir C, Wright F, Bryden C, Aslanyan S, Lees K, et al. Associations between meteorological variables and acute stroke hospital admissions in the west of Scotland. Acta Neurol Scand 2008;117:85–9. Du T, Lewin MR, Wang H, Ji X, Bohn HH, Xu T, et al. Circadian and seasonal rhythms of acute upper gastrointestinal bleeding in Beijing. Emerg Med J 2010;27: 504–7. Du J, Yang ZG, Shi L, Ma PF. Climate change characteristics of cold and warm winters in Tibet in 50 years. Acta Geograph Sin 2011;66:885–94. [In Chinese]. Ebi KL, Exuzides KA, Lau E, Kelsh M, Barnston A. Weather changes associated with hospitalizations for cardiovascular diseases and stroke in California, 1983–1998. Int J Biometeorol 2004;49:48–58. Fernández-Raga M, Tomás C, Fraile R. Human mortality seasonality in Castile-León, Spain, between 1980 and 1998: the influence of temperature, pressure and humidity. Int J Biometeorol 2010;54:379–92. Fletcher BA, Lin S, Fitzgerald EF, Hwang SA. Association of summer temperatures with hospital admissions for renal diseases in New York State: a case-crossover study. Am J Epidemiol 2012;175:907–16. Gasparrini A. Distributed lag linear and non-linear models in R: the package dlnm. J Stat Softw 2011;43:1–20. Gasparrini A, Armstrong B, Kenward MG. Distributed lag non-linear models. Stat Med 2010;29:2224–34. Goldberg MS, Burnett RT, Stieb DM, Brophy JM, Daskalopoulou SS, Valois MF, et al. Associations between ambient air pollution and daily mortality among elderly persons in Montreal, Quebec. Sci Total Environ 2013;463–464:931–42. Green RS, Basu R, Malig B, Broadwin R, Kim JJ, Ostro B. The effect of temperature on hospital admissions in nine California counties. Int J Public Health 2009;55:113–21. Grjibovski AM, Bushueva V, Boltenkov VP, Buzinov RV, Degteva GN, Yurasova ED, et al. Climate variations and salmonellosis in Northwest Russia: a time-series analysis. Epidemiol Infect 2012;4:1–8. Guo Y, Barnett AG, Pan X, Yu W, Tong S. The impact of temperature on mortality in Tianjin, China: a case-crossover design with a distributed lag nonlinear model. Environ Health Perspect 2011;119:1719–25. Guo Y, Punnasiri K, Tong S. Effects of temperature on mortality in Chiang Mai City, Thailand: a time series study. Environ Health 2012;11:36. Hu XJ, Zhao YH, Yang LH, Luosangdawa, Jiang L, Huang GW, et al. Health survey in Tibetan elders in Lhasa. Chin J Geriatr 2003;22:434–5. [In Chinese]. Huang J, Wang J, Yu W. The lag effects and vulnerabilities of temperature effects on cardiovascular disease mortality in a subtropical climate zone in China. Int J Environ Res Public Health. 2014;11:3982–94. IPCC. Fifth assessment report. Intergovernmental Panel on Climate Change; 2014. Kim YM, Park JW, Cheong HK. Estimated effect of climatic variables on the transmission of Plasmodium vivax malaria in the Republic of Korea. Environ Health Perspect 2012; 120:1314–9. Koken PJ, Piver WT, Ye F, Elixhauser A, Olsen LM, Portier CJ. Temperature, air pollution, and hospitalization for cardiovascular diseases among elderly people in Denver. Environ Health Perspect 2003;111:1312–7. Kovats RS, Hajat S, Wilkinson P. Contrasting patterns of mortality and hospital admissions during hot weather and heat waves in Greater London, UK. Occup Environ Med 2004; 61:893–8. Langma M. Seasonal incidence of upper gastrointestinal tract bleeding. Gut 1964;5: 142–4. Lavigne E, Gasparrini A, Wang X, Chen H, Yagouti A, Fleury MD, Cakmak S. Extreme ambient temperatures and cardiorespiratory emergency room visits: assessing risk by comorbid health conditions in a time series study. Environ Health 2014; 3. 13:5. Li J. Analysis of inpatient disease composition in People's Hospital of Tibet Autonomous Region. Chin Med Rec 2010;11:39–40. [In Chinese]. Li YX, Li MH, Fu SH, Chen WX, Liu QY, Zhang HL, et al. Japanese encephalitis, Tibet, China. Emerg Infect Dis 2011;17:934–6. Lin S, Luo M, Walker RJ, Liu X, Hwang SA, Chinery R. Extreme high temperatures and hospital admissions for respiratory and cardiovascular diseases. Epidemiology 2009;20: 738–46. Lin YK, Wang YC, Lin PL, Li MH, Ho TJ. Relationships between cold-temperature indices and all causes and cardiopulmonary morbidity and mortality in a subtropical island. Sci Total Environ 2013;461–462:627–35. Liu X, Chen B. Climatic warming in the Tibetan Plateau during recent decades. Int J Climatol 2000;20:1729–42. Liu Q, Liu X, Cirendunzhu C, Woodward A, Pengcuociren P, Bai L, et al. Mosquitoes established in Lhasa City, Tibet, China. Parasit Vectors 2013;6:224.
Lo SS, Johnston R, Al Sameraaii A, Metcalf PA, Rice ML, Masters JG. Seasonal variation in the acute presentation of urinary calculi over 8 years in Auckland, New Zealand. BJU Int 2010;106:96–101. Lu L, Lin H, Tian L, Yang W, Sun J, Liu Q. Time series analysis of dengue fever and weather in Guangzhou, China. BMC Public Health 2009;399:395. Michelozzi P, Accetta G, De Sario M, D'Ippoliti D, Marino C, Baccini M, et al. High temperature and hospitalizations for cardiovascular and respiratory causes in 12 European cities. Am J Respir Crit Care Med 2009;179:383–9. Midekisa A, Senay G, Henebry GM, Semuniguse P, Wimberly MC. Remote sensing-based time series models for malaria early warning in the highlands of Ethiopia. Malar J 2012;11:165. Ohshige K, Hori Y, Tochikubo O, Sugiyama M. Influence of weather on emergency transport events coded as stroke: population-based study in Japan. Int J Biometeorol 2006; 50:305–11. Pengelly LD, Campbell ME, Cheng CS, Fu C, Gingrich SE, Macfarlane R. Anatomy of heat waves and mortality in Toronto: lessons for public health protection. Can J Public Health 2007;98:364–8. Ren C, Tong S. Temperature modifies the health effects of particulate matter in Brisbane, Australia. Int J Biometeorol 2006;51:87–96. Stafoggia M, Forastiere F, Agostini D, Biggeri A, Bisanti L, Cadum E, et al. Vulnerability to heat-related mortality: a multicity, population-based, case-crossover analysis. Epidemiology 2006;17:315–23. Tian Z, Li S, Zhang J, Jaakkola JJ, Guo Y. Ambient temperature and coronary heart disease mortality in Beijing, China: a time series study. Environ Health 2012;11:56. Vandentorren S, Bretin P, Zeghnoun A, Mandereau-Bruno L, Croisier A, Cochet C, et al. August 2003 heat wave in France: risk factors for death of elderly people living at home. Eur J Public Health 2006;16:583–91. Vaneckova P, Bambrick H. Cause-specific hospital admissions on hot days in Sydney, Australia. PLoS ONE 2013;8:e55459. Vaneckova P, Beggs PJ, de Dear RJ, McCracken KW. Effect of temperature on mortality during the six warmer months in Sydney, Australia, between 1993 and 2004. Environ Res 2008;108:361–9. Wang MZ, Zheng S, He SL, Li B, Teng HJ, Wang SG, et al. The association between diurnal temperature range and emergency room admissions for cardiovascular, respiratory, digestive and genitourinary disease among the elderly: a time series study. Sci Total Environ 2013;456–457:370–5. Wichmann J, Andersen Z, Ketzel M, Ellermann T, Loft S. Apparent temperature and causespecific emergency hospital admissions in Greater Copenhagen, Denmark. PLoS ONE 2011;6:e22904. Wichmann J, Rosengren A, Sjöberg K, Barregard L, Sallsten G. Association between ambient temperature and acute myocardial infarction hospitalisations in Gothenburg, Sweden: 1985–2010. PLoS ONE 2013;8:e62059. Yan M, Chen Y, Guan ZF, Yang R. The prevalence and distribution of obesity among 696 Tibetan cadres. Chin J Prev Contr Chron Non-commun Dis 2001;15:62–3. [In Chinese]. Yang LH, Hu XJ, Zhao YH, Huang GW, Liu XQ, Jiang L, et al. A screening survey of diabetes in the middle-aged and elderly Tibetan population of Lhasa City. Chin J Endocrinol Metab 2003;19:358–60. [In Chinese]. Yang J, Liu HZ, Ou CQ, Lin GZ, Zhou Q, Shen GC, et al. Global climate change: impact of diurnal temperature range on mortality in Guangzhou, China. Environ Pollut 2013;175: 131–6. Ye X, Wolff R, Yu W, Vaneckova P, Pan X, Tong S. Ambient temperature and morbidity: a review of epidemiological evidence. Environ Health Perspect 2012;120:19–28. Yu W, Vaneckova P, Mengersen K, Pan X, Tong S. Is the association between temperature and mortality modified by age, gender and socio-economic status? Sci Total Environ 2010;408:3513–8. Yu C, Zhang Y, Claus H, Zeng R, Zhang X, Wang J. Ecological and environmental issues faced by a developing Tibet. Environ Sci Technol 2012a;46:1979–80. Yu W, Mengersen K, Wang X, Ye X, Guo Y, Pan X, et al. Daily average temperature and mortality among the elderly: a meta-analysis and systematic review of epidemiological evidence. Int J Biometeorol 2012b;56:569–81. Zanobetti A, O'Neill MS, Gronlund CJ, Schwartz JD. Susceptibility to mortality in weather extremes: effect modification by personal and small-area characteristics. Epidemiology 2013;24:809–19. Zhang Y, Bi P, Hiller JE. Weather and the transmission of bacillary dysentery in Jinan, northern China: a time-series analysis. Public Health Rep 2008;123:61–6. Zhang Y, Liu QY, Luan RS, Liu XB, Zhou GC, Jiang JY, et al. Spatial–temporal analysis of malaria and the effect of environmental factors on its incidence in Yongcheng, China, 2006–2010. BMC Public Health 2012;12:544. Zhao YH, Hu XJ, Yang LH, Luosangdawa, Jiang L, Huang GW, et al. Distribution of the minimental-status examination in the Tibetan resident population aged 55 years and over living in urban and rural areas of Lhasa City. Chin J Neurol 2002;35:336–8. [In Chinese]. Zheng X, Wang YD, Yao DK, Cirenzhuoma, Tang J, He Y. Investigation on prevalence and influence factors of hypertension in Lhasa City. Chin. Med. 2011;6:390–2. [In Chinese].
Contents lists available at ScienceDirect
Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv
Temperature, hospital admissions and emergency room visits in Lhasa, Tibet: A time-series analysis Li Bai a, Cirendunzhu b, Alistair Woodward c, Dawa b, Zhaxisangmu b, Bin Chen a, Qiyong Liu a,d,e,⁎ a State Key Laboratory for Infectious Disease Prevention and Control, National Institute for Communicable Disease Control and Prevention, Chinese Center for Disease Control and Prevention, 155 Changbai Road, Changping District, Beijing 102206, PR China b Tibet Center for Disease Control and Prevention, 21 Linkuo North Road, Lhasa, Tibet 850000, PR China c School of Population Health, University of Auckland, Private Bag 92019, Auckland 1142, New Zealand d Shandong University Climate Change and Health Center, 44 Wenhua Road, Jinan, Shangdong 250012, PR China e Collaborative Innovation Center for Diagnosis and Treatment of Infectious Diseases, Hangzhou 310003, PR China
H I G H L I G H T S • • • •
This is the first to examine temperature effects on morbidities in Tibet. Heat were more strongly associated with morbidity than were cold. Heat were associated with increases of total emergency room visits. Temperature effects on hospitalizations were cause- and population-specific.
a r t i c l e
i n f o
Article history: Received 16 January 2014 Received in revised form 30 April 2014 Accepted 6 May 2014 Available online xxxx Editor: Lidia Morawska Keywords: Emergency room visits Hospital admissions Morbidity Temperature Tibet Vulnerability
a b s t r a c t Background: Tibet of China, with an average altitude of over 4000 m, has experienced noticeable changes in its climate over the last 50 years. The association between temperature and morbidity (most commonly represented by hospital admissions) has been documented mainly in developed countries. Little is known about patterns in China; nor have the health effects of temperature variations been closely studied in highland areas, worldwide. Objective: We investigated the temperature–morbidity association in Lhasa, the capital city of Tibet, using sexand age-specific hospitalizations, excluding those due to external causes. Methods: A distributed lag non-linear model (DLNM) was applied to assess the nonlinear and delayed effects of temperature on morbidity (including total emergency room visits, total and cause-specific hospital admissions, sex- and age-specific non-external admissions). Results: High temperatures are associated with increases in morbidity, to a greater extent than low temperatures. Lag effects of high and low temperatures were cause-specific. The relative risks (RR) of high temperature for total emergency room visits and non-external hospitalizations were 1.162 (95% CI: 1.002–1.349) and 1.161 (95% CI: 1.007–1.339) respectively, for lag 0–14 days. The strongest cumulative effect of heat for lag 0–27 days was on admissions for infectious diseases (RR: 2.067, 95% CI: 1.026–4.027). Acute heat effects at lag 0 were related with increases of renal (RR: 1.478, 95% CI: 1.005–2.174) and respiratory diseases (RR: 1.119, 95% CI: 1.010–1.240), whereas immediate cold effects increased admission for digestive diseases (RR: 1.132, 95% CI: 1.002–1.282). Those ≥65 years of age and males were more vulnerable to high temperatures. Conclusion: We provide a first look at the temperature–morbidity relationship in Tibet. Exposure to both hot and cold temperatures resulted in increased admissions to hospital, but the immediate causes varied. We suggest that initiatives should be taken to reduce the adverse effects of temperature extremes in Tibet. © 2014 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/).
⁎ Corresponding author at: State Key Laboratory for Infectious Disease Prevention and Control, National Institute for Communicable Disease Control and Prevention, Chinese Center for Disease Control and Prevention, 155 Changbai Road, Changping District, Beijing 102206, PR China. E-mail addresses: [email protected] (L. Bai), [email protected] (Cirendunzhu), [email protected] (A. Woodward), [email protected] (Dawa), [email protected] (Zhaxisangmu), [email protected] (B. Chen), [email protected] (Q. Liu).
http://dx.doi.org/10.1016/j.scitotenv.2014.05.024 0048-9697/© 2014 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/).
L. Bai et al. / Science of the Total Environment 490 (2014) 838–848
1. Introduction
2. Methods
Variations in temperature affect human health in many ways (IPCC, 2014). The effects of both high and low temperatures on short-term increases in daily mortality and morbidity have been reported worldwide, particularly in developed countries, including the US (Barnett, 2007; Basu and Ostro, 2008; Braga et al., 2002), Canada (Pengelly et al., 2007), Europe (Aström et al., 2013; Vandentorren et al., 2006), and Australia (Vaneckova and Bambrick, 2013; Vaneckova et al., 2008). Temperature-sensitive health outcomes include chronic conditions such as cardiovascular diseases, respiratory diseases and renal diseases, as well as infectious diseases (vector-borne and water-borne) (IPCC, 2014). Importantly, factors that influence susceptibility to temperature tend to vary from one location to another and we note that little is known about how temperature and other aspects of weather affect human health in highland areas. Tibet of China, lying at an average altitude of more than 4000 m, is called “the third pole of the world”. Temperatures on the “roof of the world” have been increasing by 0.32 °C every decade in winters from 1955 to 1996, a much faster rate of change than has been observed in China, or in Asia generally (Liu and Chen, 2000). Accordingly Tibet is regarded as one of the most vulnerable areas to climate change in the world (Du et al., 2011; Yu et al., 2012a). However, in common with many other developing regions, nothing has been documented to date about the potential effects on health of long-term changes in climate, with particular reference to vulnerable subgroups. The capital city of Tibet, Lhasa, has experienced considerable changes in climate over the past 50 years. Maximum temperatures in Lhasa reached 30.4 °C in 2009 summer. Before this, the highest record is 29.9 °C in 1971. A cross-sectional survey of 619 respondents from urban Lhasa in 2012 summer found that many local residents are aware of the warming that has occurred in recent years (Bai et al., 2013). Over 78% reported that rising temperature is either a “very” or “somewhat” serious threat to their own health, and nearly 40% reported that they had experienced heat-related symptoms. The next step is to relate temperature variations to local measures of health and identify those in the population who are most affected, in order to target interventions and develop Tibetan-specific public health programs. The objective of this study was to investigate the relationship between temperature and morbidity in Lhasa, using sex- and agespecific hospital admissions excluding those due to external causes.
2.1. Data Collection
839
Lhasa, with an elevation of about 3680 m, is located in a flat river valley surrounded by the Himalaya Mountains rising to 5500 m (Picture 1). In 2010 the population of the city numbered 559,400 (287,400 men, 272,000 women). We obtained records of hospital admissions (1 January 2005 to 31 December 2012) and emergency room visits (1 March 2005 to 31 December 2012) from the People's Hospital of Tibet Autonomous Regions, which is located at the center of the city and close to the Potala Palace (Picture 1). This is the biggest public hospital in Tibet, with advanced facilities, more than 500 beds and over 900 staff. It is the only hospital in Tibet with a well-managed electronic medical record system, which can provide daily data from 2005 to the present, and the People's Hospital is also the first choice of most of the local residents whenever they need to see a doctor or have acute health problems. Hospitalization data included principal diagnosis, admission and discharge dates, total charges, date of birth, sex, ethnicity, and residential street address. Based on the International Classification of Disease, 10th Revision, Clinical Modification (ICD-10) we classified the data into five cause-specific categories: non-external hospital admission (A00-R99); infectious, (A00-B99) cardiovascular (I00-99), respiratory (J00-99), renal (N00-28) and digestive (K20–31, 50–52, 55–63, 92). For emergency room visit data, we obtained only the total daily counts with neither causes nor patients' information. Daily meteorological data for Lhasa were provided by the National Climate Center from 2005 to 2012. The variables included daily maximum, mean and minimum temperature and relative humidity.
2.2. Data Analysis Previous studies have shown that temperature effects on human health are frequently delayed in time. In this study, a distributed lag non-linear model (DLNM) was fitted to examine the relationship between temperature and daily numbers of hospital admissions and emergency room visits. The main advantage of this method is that it is flexible enough to simultaneously describe a non-linear exposure– response association and delayed effects or harvesting (Gasparrini, 2011; Gasparrini et al., 2010). Most recently, DLNM has been applied in studies to quantify the effects of temperature (Guo et al., 2011; Kim et al., 2012; Lin et al., 2013) and air pollution (Goldberg et al., 2013) on mortality.
Picture 1. An aerial photograph of Lhasa City, Tibet. (The red triangle represents the location of Potala Palace; the blue one represents the location of the People's Hospital of Tibet Autonomous Regions). (For interpretation of the references to color in this picture legend, the reader is referred to the web version of this article).
840
L. Bai et al. / Science of the Total Environment 490 (2014) 838–848
We controlled for seasonality and long-term trend using a natural cubic spline with 7 df per year for time. To control any confounding weekly pattern, the day of week was also included as an indicator in our analysis. Public holiday was controlled as a binary variable. We did not control for relative humidity, as levels of water vapor in the air in Tibet are low across the year. We evaluated the model fit using Akaike's Information Criterion for quasi-Poisson (Q-AIC). The results showed that models that controlled humidity had higher Q-AIC values, indicating that models excluding humidity were a better fit (results not shown). We used maximum temperature to assess temperature effects in this analysis since it was found to be a better predictor (i.e. had lower Q-AIC values) than the mean and minimum temperatures. We used a DLNM with 5 degrees of freedom natural cubic for temperature (knots at equally-spaced percentiles by default) and with 4 degrees of freedom natural cubic for lags (knots at equally-spaced values in the log scale of lags by default) (Guo et al., 2011; Tian et al., 2012). The median value of the maximum temperature (17.7 °C) was used as the reference value (centering value) to calculate the relative risks. A maximum lag of 27 was used to completely capture the overall temperature effect and adjust for a possible harvesting effect (Guo et al., 2011). We examined and plotted cumulative relative risks with temperature for lag 0, lags 0–2, 0–14 and 0–27. We also examined the hot and cold effects separately on relative risks of total emergency room visits, cause-specific hospital admissions, and sex- and age-specific non-accidental hospitalizations. For hot effects, we calculated relative risks associated with the 99th percentile of temperature (high temperature) relative to the 75th percentile of temperature. For cold effects, we calculated relative risks associated with the first percentile of temperature (cold temperature) relative to the 25th percentile of temperature (Guo et al., 2012). The cumulative effects of hot and cold temperatures along the lags were then estimated separately. Sensitivity analysis was conducted by changing the df for time from 6 to 12 per year, the maximum lag days from 14 to 30 days, and the df for temperature and lags from 3 to 6. All statistical tests and modeling were performed using the R software (version 3.0.1). Distributed lag nonlinear models were fitted through “dlnm” package (Gasparrini, 2011).
3. Results In all, there were 82,491 hospital admissions from 1 January 2005 to 31 December 2012 and 159,139 emergency room visits from 1 March 2005 to 31 December 2012. Table 1 shows the statistical summary for Table 1 Summary statistics of daily weather conditions, hospital admissions and emergency room visits in Lhasa, Tibet during 2005 to 2012.
Mean temperature (°C) Maximum temperature (°C) Minimum temperature (°C) Relative humidity (%) Total emergency room visits Total hospital admissions Non-external Infectious Cardiovascular Respiratory Renal Digestive Demographica Male Female 0–15 years 16–44 years 45–64 years ≥65 years
Mean
SD
Min
25th
Median
75th
Max
9.6 17.4 3.3 33.7 56 28 22 1 2 4 1 2
6.6 6.1 7.4 17.2 19.0 12.6 10.6 1.3 1.8 2.8 1.1 1.4
−7.3 −0.9 −16.1 5.0 3 1 1 0 0 0 0 0
4.0 12.6 −3.2 19.0 43 18 13 0 1 2 0 0
9.95 17.7 3.4 30.0 53 27 21 1 2 4 1 1
15.4 22.5 10.1 46.0 67 37 30 2 3 6 1 2
22.6 30.4 18.2 84.0 167 76 65 8 10 24 8 10
10 11 4 9 5 2
14 15 6 13 8 4
10 11 4 10 5 3
5.9 5.7 2.5 5.0 3.8 2.1
0 0 0 0 0 0
6 7 2 6 2 1
36 33 14 29 20 13
a Demographic variables are only applicable to the hospital admissions for non-external diseases.
hospital data and weather condition. The average daily maximum temperature was 17.4 °C; the mean temperature was 9.6 °C; the minimum temperature was 3.3 °C. Fig. 1 displays the full data series for temperature, total emergency room admissions and hospitalizations. The mean, maximum and minimum temperatures varied strongly with season. There was also a clear seasonal trend of total emergency room visits, with more patients at high temperature days. The daily number of hospitalizations has increased slightly since 2008 and we observed a weak association between temperature and hospitalizations after 2009, with more cases in summer than in winter. The 3-D plots show the non-linear relationships between maximum temperature and total emergency room admissions, as well as causespecific hospitalizations along 27 lag days (Fig. 2). For total hospital admissions, there was no evident effect of temperature, but both low and high temperatures were associated with increased numbers of total emergency room visits. For cause-specific hospitalizations, admissions for non-external, respiratory and renal diseases had higher relative risks at hot temperature, but relative risks of admissions for cardiovascular and digestive diseases were higher at cold temperatures. A delayed heat effect was observed for infectious disease admissions. Fig. 3 shows the temperature effects of current day (lag 0) and cumulative effects on total emergency room admissions, total and causespecific hospital admissions at lags 0–2, 0–14, and 0–27. For the effects of extreme hot temperature, the cumulative risks on total emergency room visits, non-external and infectious disease hospitalizations at lag 0–27 were higher than those at lag 0, lags 0–2 and 0–14, while the cumulative risk on admissions for renal diseases and respiratory diseases was highest at lag 0 and lag 0–2 respectively. For the effects of extreme cold temperature, the cumulative risks on total emergency room admissions and cause-specific hospitalizations were highest at lag 0–27, except for renal and respiratory admissions. We plotted the hot and cold effects of temperature separately on total emergency room visits and cause-specific hospitalizations along the lag days (Figs. 4, 5). Lag effects for different admission categories were quite different. For example, heat effects occurred immediately for total emergency room visits and hospitalizations for respiratory and renal diseases, whereas heat effects appeared after a 5-day lag for infectious diseases and a 7-day lag for cardiovascular diseases. Acute cold effects were also observed for total emergency room visits and digestive diseases, whereas cold effects occurred after a 2-day lag for total hospital admissions, cardiovascular and renal admissions. Fig. 6 shows the hot and cold effects on sex- and age-specific non-accidental hospitalizations along the lags. Immediate heat effects were found in all sub-populations except for those aged 16–44 years old. In addition, heat effects lasted longer in those aged 45–64 years and older than 65 years. Cold effects increased at long lag days for all sub-groups except for those aged 0–15 years old. The cumulative hot effects on total emergency room visits, causespecific hospitalizations, and sex- and age-specific non-external hospitalizations were calculated along the lags (Table 2). For the short lags (lag 0 and lag 0–2), heat effects were apparent for renal and respiratory diseases, whereas for the long lags (lag 0–14 and lag 0–27), high temperatures were associated with the risk of total emergency room visits, and hospitalizations for non-external and infectious diseases. The highest cumulative risk of heat effects was observed on infectious diseases at lag 0–27 (RR 2.067, 95% CI: 1.026–4.027). For sex- and agespecific non-external hospitalizations, males were more affected than females, and the elderly were more sensitive to hot temperatures than other age groups. The cumulative cold effects on total emergency room visits, causespecific hospitalizations, and sex- and age-specific non-external hospitalization were calculated along the lags (Table 3). The relative risk of admissions for digestive diseases at lag 0 was the only statistically significant result (RR: 1.132, 95% CI: 1.002–1.282). The highest risk of cold effect on demographic-specific non-external diseases was
L. Bai et al. / Science of the Total Environment 490 (2014) 838–848
841
Fig. 1. Temperature, total emergency room visits and hospital admissions in Lhasa, Tibet during 2005 to 2012.
observed in those older than 65 years at lag 0–27, although the result is not statistically significant. 3.1. Sensitivity Analysis We changed the df (6–12 per year) for time to control seasonality, which produced similar results (data not shown). We also altered the maximum lag days from 14 to 31 days, and found that the temperature effects were similar (data not shown). Additionally, the df for temperature and lags from 3 to 6 was changed, which also gave similar results. 4. Discussion In this study, we have analyzed the associations between daily maximum temperature and emergency room visits and hospital admissions in Lhasa, Tibet during 2005–2012. It appears that elevated temperatures affect morbidity to a greater degree than do low temperatures. High temperatures were associated with increases of total emergency room visits and hospitalizations for infectious, renal and respiratory diseases, while cold temperatures were associated with more frequent digestive disease admissions. Lag effects of both high and low temperatures were cause-specific. Acute temperature effects were observed with renal, respiratory diseases and digestive diseases, whereas delayed effects were associated with increased total emergency room visits, and hospitalizations for non-external and infectious diseases. We also observed some signs of increased vulnerability to high temperatures among those ≥65 years of age and males. This study is, to our best knowledge,
the first to examine temperature-related morbidity and vulnerable subpopulations in Tibet of China, or any other highland area. The Tibetan plateau, a part of the world that is particularly sensitive to environmental disruption (Yu et al., 2012a), has experienced rapid changes in its climate (Du et al., 2011; Liu and Chen, 2000). Others have suggested that heat exposure tends to be more troublesome in cold regions, whereas warm regions may suffer more from cold weather (Yang et al., 2013). Our findings are consistent with this since heat had a greater impact on human health than cold, and Lhasa experiences mild summer climates and extreme cold winter temperatures. It may be that people living in relatively cold regions have low awareness and limited coping resources (such as air-conditioning) to deal with the heat. As might be expected, cooling devices are not common in Lhasa: a survey of 619 respondents which we conducted in August 2012 found that 8.7% of participants had air-conditioning at home, 18.1% reported they had a fan and about 6% had both (Bai et al., 2013). Moreover, many of those who had air-conditioning at home reported that they seldom use the equipment, due to concerns about power costs or other unwanted effects of cooling. Others have reported immediate effects of ambient temperatures on total and cause-specific mortalities (Guo et al., 2012; Huang et al., 2014; Tian et al., 2012) and measures of morbidity (Chan et al., 2013; Wichmann et al., 2011; Wichmann et al., 2013; Zanobetti et al., 2013). In our analysis, we found both hot and cold effects with a range of lag periods. The risk of acute heat effects (lag 0) was highest for renal diseases (RR: 1.478, 95% CI: 1.005–2.174). This was consistent with previous studies which found that the largest hot effects on renal diseases occurred at lag 0 (Green et al., 2009) or lag 1 (Fletcher et al., 2012).
842
L. Bai et al. / Science of the Total Environment 490 (2014) 838–848
Total emergency room visits
Total hospital admissions
Non-external diseases*
Infectious diseases*
Cardiovascular diseases*
Respiratory diseases*
Renal diseases*
Digestive diseases*
Fig. 2. The three-dimensional plot of the association between maximum temperature and morbidity along 27 lags (*cause-specific hospital admissions).
Some studies have documented that the biologic mechanism underlying the relationship between high temperature and renal diseases, which disrupted fluid balance during heat exposure leads to acute renal failure and the development of renal calculi (Brikowski et al., 2008; Chen et al., 2008; Lo et al., 2010). Consistent with other studies (Chan et al., 2013; Lin et al., 2009), we also observed a short heat effect within 0–2 days on admissions for respiratory diseases (RR: 1.234, 95% CI: 1.034–1.472). Similarly, Lin et al. (2009) found that the greatest number of hospital admissions for respiratory diseases was 0–1 days after elevated temperatures.
Apart from short hot effects, delayed and longer lasting heat effects (lags 0–14 and 0–27) were observed with the increase of total emergency room visits, non-external hospitalizations and infectious disease admissions in this study. Most previous studies focused on short-term heat effects on cause-specific emergency department visits and hospital admissions such as total cardiovascular diseases (Ren and Tong, 2006; Michelozzi et al., 2009), total respiratory diseases (Lavigne et al., 2014; Lin et al., 2009), acute myocardial infarction (Ebi et al., 2004; Wichmann et al., 2013) and stroke (Dawson et al., 2008; Ohshige et al., 2006). However, a few confirmed the associations with total emergency room visits or total non-external hospitalizations. For instance, a study in UK found no relation between total emergency room visits and heat when using short-term temperatures (three-day moving average temperature) (Kovats et al., 2004). For infectious disease admissions, we also observed that the cumulative effect due to high temperature could last 27 days (RR: 2.067, 95% CI: 1.026–4.027). Previous studies have reported that the transmission and occurrence of some climate-sensitive infectious diseases such as malaria (Midekisa et al., 2012; Zhang et al., 2012), dengue (Lu et al., 2009), diarrhea (Alexander et al., 2013; Constantin de Magny et al., 2008), dysentery (Zhang et al., 2008) and salmonellosis (Grjibovski et al., 2012) were positively linked with one-month lag temperatures. In Tibet, rapid warming may introduce new threats, particularly those due to vector-borne diseases. With increased summer temperatures in Lhasa, we have already observed that mosquitoes in the Culex pipiens complex persist from one season to the next, and now appear to be established in what was previously recognized as a mosquitofree area due to its extreme altitude (Liu et al., 2013). Communicable diseases that may be transmitted by this mosquito species include Japanese encephalitis, meningitis, and urticaria. In 2009 Japanese encephalitis virus was isolated from Culex tritaeniorhynchus mosquitoes elsewhere in Tibet, indicating that JEV is already circulating in highaltitude regions (Li et al., 2011). On the basis of these changes we suggest targeted actions and programs such as specific disease monitoring, emergency preparedness, vector control and public education prior to the seasonal peak in mosquito numbers in Lhasa. In our analysis, acute cold effects at lag 0 were found to increase admissions for digestive diseases (RR: 1.132 95% CI: 1.002–1.282) which are ranked second in Lhasa in terms of both mortality and morbidity (Bai et al., 2008; Li, 2010). According to the literature, the seasonal pattern of digestive disorders is clear, with a significant increase found in the winter months for some severe digestive diseases such as gastric and duodenal ulcer and bleeding from the upper gastrointestinal tract (Du et al., 2010; Langma, 1964). It has been hypothesized that exposure to the cold tends to stimulate human digestive processes and worsen digestive symptoms by constricting the peripheral blood vessels, leading to congestion of the visceral vessels and increased secretions. However, few studies, to date, have explored the relationship between temperature and digestive diseases. Fernández-Raga et al. (2010) reported that the most comfortable temperatures for patients with digestive diseases (19.7 °C) are apparently higher than those for patients with cardiovascular diseases (16.8 °C) and respiratory diseases (18.1 °C) (Fernández-Raga et al., 2010). A study carried out in Beijing found that diurnal temperature range is an independent risk factor for emergency room admissions for digestive diseases among elderly persons (Wang et al., 2013), which implies that sudden modifications in the external temperature also react directly upon the digestive apparatus. Cardiovascular diseases are well-known to be sensitive to extreme temperatures (Basu, 2008). However, we found neither hot nor cold effects for this category of diseases. A review on recent temperature– morbidity studies concluded that sensitivity of cardiovascular diseases to temperatures varies by sub-types (Ye et al., 2012). Chronic and acute cardiovascular diseases were found to have opposite relationships to temperature increases in some studies (Green et al., 2009; Koken et al., 2003). The authors suggested that those with chronic longstanding conditions might avoid outdoor exposures during unpleasant
L. Bai et al. / Science of the Total Environment 490 (2014) 838–848
Lag0
Lag0-2
Lag0-14
843
Lag0-27
Total emergency room visits
Total hospital admissions
Non-external diseases*
Infectious diseases*
Cardiovascular diseases*
Respiratory diseases*
Renal diseases*
Digestive diseases*
Fig. 3. Exposure–response curves of maximum temperature and morbidities along the lags (*cause-specific hospital admissions).
844
L. Bai et al. / Science of the Total Environment 490 (2014) 838–848
Total emergency room visits
Total hospital admissions
Non-external diseases*
Infectious diseases*
Cardiovascular diseases*
Respiratory diseases*
Renal diseases*
Digestive diseases*
Fig. 4. The estimated hot effects of maximum temperature (99th relative to 75th percentile) on morbidities along the lags (*cause-specific hospital admissions).
weather, resulting in a null or negative correlation with temperature. Findings in our survey may also support this conclusion (Bai et al., 2013). In Lhasa, we found that residents with chronic diseases were more likely to perceive the risk on health of excessive heat and were more likely to alter their behaviors during hot summer days (e.g. drinking more fluids, staying indoors and avoiding the sun).
Vulnerability to temperature-related mortality and morbidity has been associated previously with socio-demographic characteristics such as age, sex, health, education, and occupation. We found that males and the elderly were at higher risk of morbidity associated with the heat. Some studies in other locations have reported that women had higher risks than men to extreme weather (Stafoggia et al., 2006;
L. Bai et al. / Science of the Total Environment 490 (2014) 838–848
Total emergency room visits
Total hospital admissions
Non-external diseases*
Infectious diseases*
Cardiovascular diseases*
Respiratory diseases*
Renal diseases*
Digestive diseases*
845
Fig. 5. The estimated cold effects of maximum temperature (1st relative to 25th percentile) on morbidities along the lags (*cause-specific hospital admissions).
Vaneckova et al., 2008; Yu et al., 2010). In Tibet, we note that men are more likely than women to suffer from chronic conditions including obesity (Yan et al., 2001), diabetes (Dawa et al., 2006; Yang et al., 2003), high blood pressure (Hu et al., 2003; Zheng et al., 2011), and
Alzheimer's disease (Zhao et al., 2002). In addition, Tibetan men are more likely to engage in outdoor physical activities. Consistent with others (Yu et al., 2012b), we observed a strong and significant effect of high temperatures on those older than 65 years. Nevertheless, we
846
L. Bai et al. / Science of the Total Environment 490 (2014) 838–848
Hot effects Male
Female
0-15 y
16-44 y
45-64 y
>=65 y
Male
Female
0-15 y
16-44 y
45-64 y
>=65 y
Cold effects
Fig. 6. The estimated hot and cold effects of maximum temperature on sex- and age-specific hospital admissions for non-external diseases along the lags.
suggest that special attention also should be paid to middle aged persons (45–64 years) since this age group in Tibet is commonly affected by chronic medical conditions (Bai et al., 2013; Hu et al., 2003). Also, unlike the elderly, most middle aged people are still at work and thus may be exposed to more unfavorable outdoor weather. Several limitations of the study should be noted. We did not have data on outdoor air pollution, which is independently associated with higher mortality, varies over short periods of time, and may modify the temperature–morbidity relationship. However, we note that air
quality in Lhasa is among the best in China. According to local environmental departments, the number of days with high air quality (Air Quality Index ≤ 50, which means that the daily average concentration of PM2.5 is 50 μg/m3 or lower) in 2012 was 361. Moreover, our results were based on analyzing one hospital data. This may result in bias in the estimated associations. Lastly, with this study design it was not possible to control other time-varying confounding variables such as intake of alcohol, smoking rates, use of medications and indoor air pollution due to residential heating in winters.
L. Bai et al. / Science of the Total Environment 490 (2014) 838–848
847
Table 2 The cumulative relative risks of hot effects on emergency room visits, cause-specific hospital admissions, and sex- and age-specific admissions for non-external diseases.
Total ERVa Total hospitalizations Non-external Infectious Cardiovascular Respiratory Renal Digestive Demographicb Male Female 0–15 16–44 45–64 ≥65
Lag 0
Lag 0–2
Lag 0–14
Lag 0–27
1.035 (0.956–1.122) 0.998 (0.897–1.111) 1.042 (0.996–1.091) 1.016 (0.871–1.184) 0.988 (0.867–1.127) 1.119 (1.010–1.240) 1.478 (1.005–2.174) 1.055 (0.921–1.208)
1.036 (0.951–1.128) 1.010 (0.886–1.151) 1.065 (0.984–1.153) 0.990 (0.757–1.293) 0.931 (0.730–1.189) 1.234 (1.034–1.472) 1.125 (0.791–1.597) 1.036 (0.818–1.313)
1.162 (1.002–1.349) 1.075 (0.836–1.382) 1.161 (1.007–1.339) 1.641 (1.026–2.207) 0.990 (0.563–1.740) 1.080 (0.808–1.444) 1.260 (0.704–2.253) 0.956 (0.632–1.449)
1.194 (0.951–1.500) 1.087 (0.728–1.623) 1.251 (1.011–1.549) 2.067 (1.026–4.027) 1.143 (0.449–2.900) 1.117 (0.696–1.792) 1.635 (0.655–2.760) 0.938 (0.472–1.864)
1.036 (0.975–1.102) 1.044 (0.986–1.105) 1.047 (0.962–1.140) 1.007 (0.950–1.068) 1.071 (0.988–1.162) 1.221 (1.023–1.457)
1.055 (0.949–1.172) 1.076 (0.975–1.187) 1.091 (0.942–1.263) 1.012 (0.914–1.119) 1.094 (0.951–1.260) 1.150 (1.002–1.321)
1.166 (1.002–1.460) 1.069 (0.910–1.256) 1.118 (0.876–1.426) 1.140 (0.964–1.347) 1.062 (0.842–1.340) 1.249 (0.910–1.712)
1.246 (0.939–1.652) 1.241 (0.956–1.612) 1.106 (0.742–1.647) 1.182 (0.901–1.551) 1.358 (0.931–1.978) 1.631 (0.984–2.703)
The bold-faced data means statistically significant (p b 0.05). a ERV: emergency room visits. b Demographic variables are only applicable to the hospital admissions for non-external diseases.
Table 3 The cumulative relative risks of cold effects on emergency room visits, cause-specific hospital admissions, sex- and age-specific admissions for non-external diseases.
a
Total ERV Total hospitalizations Non-external Infectious Cardiovascular Respiratory Renal Digestive Demographicb Male Female 0–15 16–44 45–64 ≥65
Lag 0
Lag 0–2
Lag 0–14
Lag 0–27
1.041 (0.975–1.110) 0.957 (0.912–1.005) 0.986 (0.994–1.030) 1.053 (0.905–1.226) 0.962 (0.851–1.087) 0.964 (0.893–1.040) 0.966 (0.690–1.350) 1.132 (1.002–1.282)
1.021 (0.943–1.105) 0.946 (0.891–1.004) 0.961 (0.890–1.038) 1.054 (0.810–1.373) 1.000 (0.802–1.248) 0.908 (0.792–1.040) 1.051 (0.737–1.498) 1.233 (0.998–1.531)
1.023 (0.877–1.194) 0.894 (0.792–1.008) 0.895 (0.788–1.017) 0.928 (0.592–1.455) 1.030 (0.656–1.618) 0.799 (0.616–1.037) 0.912 (0.498–1.670) 1.153 (0.776–1.713)
1.127 (0.890–1.428) 1.023 (0.823–1.287) 0.982 (0.810–1.191) 1.229 (0.627–2.407) 1.001 (0.462–2.166) 0.747 (0.499–1.119) 1.043 (0.405–2.448) 1.354 (0.733–2.503)
1.030 (0.972–1.091) 0.954 (0.904–1.007) 1.011 (0.936–1.092) 0.977 (0.923–1.035) 1.042 (0.964–1.126) 0.916 (0.792–1.059)
1.029 (0.931–1.138) 0.911 (0.830–1.002) 1.005 (0.879–1.150) 0.935 (0.846–1.034) 1.057 (0.922–1.210) 0.810 (0.675–0.973)
0.913 (0.771–1.081) 0.909 (0.772–1.070) 1.005 (0.800–1.262) 0.796 (0.673–0.940) 0.926 (0.738–1.163) 1.050 (0.773–1.427)
1.023 (0.793–1.319) 1.001 (0.780–1.286) 1.137 (0.801–1.613) 0.852 (0.662–1.095) 1.006 (0.715–1.415) 1.195 (0.757–1.887)
The bold-faced data means statistically significant (p b 0.05). a ERV: emergency room visits. b Demographic variables are only applicable to the hospital admissions for non-external diseases.
5. Conclusion Our study has added an important dimension to the growing body of work that documents the temperature–morbidity association, as there have been no reports previously from such a high-elevation setting. We found that hot temperatures were more strongly associated with morbidity in Lhasa than were cold temperatures. Heat effects were associated with increases of total emergency room visits, hospital admissions for non-external diseases, renal diseases and respiratory diseases, whereas cold temperatures were linked with digestive disease admissions. We suggest that initiatives are needed now to minimize the downside temperature effects in Tibet, especially in hot seasons, due to climate change and limited resources needed to cope. Particular attention should be paid to groups at greater risk, such as the elderly. Further studies are required to confirm our findings, to examine other population sensitivity factors and to include other locations in Tibet. Competing interests The authors declare that they have no competing interests. Acknowledgments This study was supported by the National Basic Research Program of China (973 Program) (Grant No. 2012CB955504). We are grateful to the
People's Hospital of Tibet Autonomous Regions for providing morbidity data. We also wish to thank Doctor Antonio Gasparrini from the London School of Hygiene and Tropical Medicine, UK for the guidance on statistical analysis. References Alexander KA, Carzolio M, Goodin D, Vance E. Climate change is likely to worsen the public health threat of diarrheal disease in Botswana. Int J Environ Res Public Health 2013;10:1202–30. Aström C, Orru H, Rocklöv J, Strandberg G, Ebi KL, Forsberg B. Heat-related respiratory hospital admissions in Europe in a changing climate: a health impact assessment. BMJ Open 2013;24:3. Bai G, Nimaqucuo, Gamacangjue, Guo S, Renqingyuzhen. Analysis of mortality in Tibet during 2004–2005. Chin J Prev Contr Chron Non-commun Dis 2008;16:525. Bai L, Cirendunzhu D, Pengcuociren D, Dawa D, Woodward A, Liu X, et al. Rapid warming in Tibet, China: public perception, response and coping resources in urban Lhasa. Environ Health 2013;12:71. Barnett AG. Temperature and cardiovascular deaths in the US elderly: changes over time. Epidemiology 2007;18:369–72. Basu R. High ambient temperature and mortality: a review of epidemiologic studies from 2001 to 2008. Environ Health 2008;8:40. Basu R, Ostro BD. A multicounty analysis identifying the populations vulnerable to mortality associated with high ambient temperature in California. Am J Epidemiol 2008;168:632–7. Braga AL, Zanobetti A, Schwartz J. The effect of weather on respiratory and cardiovascular deaths in 12 U.S. cities. Environ Health Perspect 2002;110:859–63. Brikowski TH, Lotan Y, Pearle MS. Climate-related increase in the prevalence of urolithiasis in the United States. Proc Natl Acad Sci U S A 2008;105:9841–6. Chan EY, Goggins WB, Yue JS, Lee P. Hospital admissions as a function of temperature, other weather phenomena and pollution levels in an urban setting in China. Bull World Health Organ 2013;91:576–84.
848
L. Bai et al. / Science of the Total Environment 490 (2014) 838–848
Chen YK, Lin HC, Chen CS, Yeh SD. Seasonal variations in urinary calculi attacks and their association with climate: a population based study. J Urol. 2008;179:564–9. Constantin de Magny G, Murtugudde R, Sapiano MR, Nizam A, Brown CW, Busalacchi AJ, et al. Environmental signatures associated with cholera epidemics. Proc Natl Acad Sci U S A 2008;105:17676–81. Dawa Ye XM, Dunzhuduoji, Gesang, Gamacangjue, Gou XQ. Study on the prevalence of major chronic diseases of adults in 2002 in Lhasa, Tibet. Food Sci Technol 2006;10: 19–21. [In Chinese]. Dawson J, Weir C, Wright F, Bryden C, Aslanyan S, Lees K, et al. Associations between meteorological variables and acute stroke hospital admissions in the west of Scotland. Acta Neurol Scand 2008;117:85–9. Du T, Lewin MR, Wang H, Ji X, Bohn HH, Xu T, et al. Circadian and seasonal rhythms of acute upper gastrointestinal bleeding in Beijing. Emerg Med J 2010;27: 504–7. Du J, Yang ZG, Shi L, Ma PF. Climate change characteristics of cold and warm winters in Tibet in 50 years. Acta Geograph Sin 2011;66:885–94. [In Chinese]. Ebi KL, Exuzides KA, Lau E, Kelsh M, Barnston A. Weather changes associated with hospitalizations for cardiovascular diseases and stroke in California, 1983–1998. Int J Biometeorol 2004;49:48–58. Fernández-Raga M, Tomás C, Fraile R. Human mortality seasonality in Castile-León, Spain, between 1980 and 1998: the influence of temperature, pressure and humidity. Int J Biometeorol 2010;54:379–92. Fletcher BA, Lin S, Fitzgerald EF, Hwang SA. Association of summer temperatures with hospital admissions for renal diseases in New York State: a case-crossover study. Am J Epidemiol 2012;175:907–16. Gasparrini A. Distributed lag linear and non-linear models in R: the package dlnm. J Stat Softw 2011;43:1–20. Gasparrini A, Armstrong B, Kenward MG. Distributed lag non-linear models. Stat Med 2010;29:2224–34. Goldberg MS, Burnett RT, Stieb DM, Brophy JM, Daskalopoulou SS, Valois MF, et al. Associations between ambient air pollution and daily mortality among elderly persons in Montreal, Quebec. Sci Total Environ 2013;463–464:931–42. Green RS, Basu R, Malig B, Broadwin R, Kim JJ, Ostro B. The effect of temperature on hospital admissions in nine California counties. Int J Public Health 2009;55:113–21. Grjibovski AM, Bushueva V, Boltenkov VP, Buzinov RV, Degteva GN, Yurasova ED, et al. Climate variations and salmonellosis in Northwest Russia: a time-series analysis. Epidemiol Infect 2012;4:1–8. Guo Y, Barnett AG, Pan X, Yu W, Tong S. The impact of temperature on mortality in Tianjin, China: a case-crossover design with a distributed lag nonlinear model. Environ Health Perspect 2011;119:1719–25. Guo Y, Punnasiri K, Tong S. Effects of temperature on mortality in Chiang Mai City, Thailand: a time series study. Environ Health 2012;11:36. Hu XJ, Zhao YH, Yang LH, Luosangdawa, Jiang L, Huang GW, et al. Health survey in Tibetan elders in Lhasa. Chin J Geriatr 2003;22:434–5. [In Chinese]. Huang J, Wang J, Yu W. The lag effects and vulnerabilities of temperature effects on cardiovascular disease mortality in a subtropical climate zone in China. Int J Environ Res Public Health. 2014;11:3982–94. IPCC. Fifth assessment report. Intergovernmental Panel on Climate Change; 2014. Kim YM, Park JW, Cheong HK. Estimated effect of climatic variables on the transmission of Plasmodium vivax malaria in the Republic of Korea. Environ Health Perspect 2012; 120:1314–9. Koken PJ, Piver WT, Ye F, Elixhauser A, Olsen LM, Portier CJ. Temperature, air pollution, and hospitalization for cardiovascular diseases among elderly people in Denver. Environ Health Perspect 2003;111:1312–7. Kovats RS, Hajat S, Wilkinson P. Contrasting patterns of mortality and hospital admissions during hot weather and heat waves in Greater London, UK. Occup Environ Med 2004; 61:893–8. Langma M. Seasonal incidence of upper gastrointestinal tract bleeding. Gut 1964;5: 142–4. Lavigne E, Gasparrini A, Wang X, Chen H, Yagouti A, Fleury MD, Cakmak S. Extreme ambient temperatures and cardiorespiratory emergency room visits: assessing risk by comorbid health conditions in a time series study. Environ Health 2014; 3. 13:5. Li J. Analysis of inpatient disease composition in People's Hospital of Tibet Autonomous Region. Chin Med Rec 2010;11:39–40. [In Chinese]. Li YX, Li MH, Fu SH, Chen WX, Liu QY, Zhang HL, et al. Japanese encephalitis, Tibet, China. Emerg Infect Dis 2011;17:934–6. Lin S, Luo M, Walker RJ, Liu X, Hwang SA, Chinery R. Extreme high temperatures and hospital admissions for respiratory and cardiovascular diseases. Epidemiology 2009;20: 738–46. Lin YK, Wang YC, Lin PL, Li MH, Ho TJ. Relationships between cold-temperature indices and all causes and cardiopulmonary morbidity and mortality in a subtropical island. Sci Total Environ 2013;461–462:627–35. Liu X, Chen B. Climatic warming in the Tibetan Plateau during recent decades. Int J Climatol 2000;20:1729–42. Liu Q, Liu X, Cirendunzhu C, Woodward A, Pengcuociren P, Bai L, et al. Mosquitoes established in Lhasa City, Tibet, China. Parasit Vectors 2013;6:224.
Lo SS, Johnston R, Al Sameraaii A, Metcalf PA, Rice ML, Masters JG. Seasonal variation in the acute presentation of urinary calculi over 8 years in Auckland, New Zealand. BJU Int 2010;106:96–101. Lu L, Lin H, Tian L, Yang W, Sun J, Liu Q. Time series analysis of dengue fever and weather in Guangzhou, China. BMC Public Health 2009;399:395. Michelozzi P, Accetta G, De Sario M, D'Ippoliti D, Marino C, Baccini M, et al. High temperature and hospitalizations for cardiovascular and respiratory causes in 12 European cities. Am J Respir Crit Care Med 2009;179:383–9. Midekisa A, Senay G, Henebry GM, Semuniguse P, Wimberly MC. Remote sensing-based time series models for malaria early warning in the highlands of Ethiopia. Malar J 2012;11:165. Ohshige K, Hori Y, Tochikubo O, Sugiyama M. Influence of weather on emergency transport events coded as stroke: population-based study in Japan. Int J Biometeorol 2006; 50:305–11. Pengelly LD, Campbell ME, Cheng CS, Fu C, Gingrich SE, Macfarlane R. Anatomy of heat waves and mortality in Toronto: lessons for public health protection. Can J Public Health 2007;98:364–8. Ren C, Tong S. Temperature modifies the health effects of particulate matter in Brisbane, Australia. Int J Biometeorol 2006;51:87–96. Stafoggia M, Forastiere F, Agostini D, Biggeri A, Bisanti L, Cadum E, et al. Vulnerability to heat-related mortality: a multicity, population-based, case-crossover analysis. Epidemiology 2006;17:315–23. Tian Z, Li S, Zhang J, Jaakkola JJ, Guo Y. Ambient temperature and coronary heart disease mortality in Beijing, China: a time series study. Environ Health 2012;11:56. Vandentorren S, Bretin P, Zeghnoun A, Mandereau-Bruno L, Croisier A, Cochet C, et al. August 2003 heat wave in France: risk factors for death of elderly people living at home. Eur J Public Health 2006;16:583–91. Vaneckova P, Bambrick H. Cause-specific hospital admissions on hot days in Sydney, Australia. PLoS ONE 2013;8:e55459. Vaneckova P, Beggs PJ, de Dear RJ, McCracken KW. Effect of temperature on mortality during the six warmer months in Sydney, Australia, between 1993 and 2004. Environ Res 2008;108:361–9. Wang MZ, Zheng S, He SL, Li B, Teng HJ, Wang SG, et al. The association between diurnal temperature range and emergency room admissions for cardiovascular, respiratory, digestive and genitourinary disease among the elderly: a time series study. Sci Total Environ 2013;456–457:370–5. Wichmann J, Andersen Z, Ketzel M, Ellermann T, Loft S. Apparent temperature and causespecific emergency hospital admissions in Greater Copenhagen, Denmark. PLoS ONE 2011;6:e22904. Wichmann J, Rosengren A, Sjöberg K, Barregard L, Sallsten G. Association between ambient temperature and acute myocardial infarction hospitalisations in Gothenburg, Sweden: 1985–2010. PLoS ONE 2013;8:e62059. Yan M, Chen Y, Guan ZF, Yang R. The prevalence and distribution of obesity among 696 Tibetan cadres. Chin J Prev Contr Chron Non-commun Dis 2001;15:62–3. [In Chinese]. Yang LH, Hu XJ, Zhao YH, Huang GW, Liu XQ, Jiang L, et al. A screening survey of diabetes in the middle-aged and elderly Tibetan population of Lhasa City. Chin J Endocrinol Metab 2003;19:358–60. [In Chinese]. Yang J, Liu HZ, Ou CQ, Lin GZ, Zhou Q, Shen GC, et al. Global climate change: impact of diurnal temperature range on mortality in Guangzhou, China. Environ Pollut 2013;175: 131–6. Ye X, Wolff R, Yu W, Vaneckova P, Pan X, Tong S. Ambient temperature and morbidity: a review of epidemiological evidence. Environ Health Perspect 2012;120:19–28. Yu W, Vaneckova P, Mengersen K, Pan X, Tong S. Is the association between temperature and mortality modified by age, gender and socio-economic status? Sci Total Environ 2010;408:3513–8. Yu C, Zhang Y, Claus H, Zeng R, Zhang X, Wang J. Ecological and environmental issues faced by a developing Tibet. Environ Sci Technol 2012a;46:1979–80. Yu W, Mengersen K, Wang X, Ye X, Guo Y, Pan X, et al. Daily average temperature and mortality among the elderly: a meta-analysis and systematic review of epidemiological evidence. Int J Biometeorol 2012b;56:569–81. Zanobetti A, O'Neill MS, Gronlund CJ, Schwartz JD. Susceptibility to mortality in weather extremes: effect modification by personal and small-area characteristics. Epidemiology 2013;24:809–19. Zhang Y, Bi P, Hiller JE. Weather and the transmission of bacillary dysentery in Jinan, northern China: a time-series analysis. Public Health Rep 2008;123:61–6. Zhang Y, Liu QY, Luan RS, Liu XB, Zhou GC, Jiang JY, et al. Spatial–temporal analysis of malaria and the effect of environmental factors on its incidence in Yongcheng, China, 2006–2010. BMC Public Health 2012;12:544. Zhao YH, Hu XJ, Yang LH, Luosangdawa, Jiang L, Huang GW, et al. Distribution of the minimental-status examination in the Tibetan resident population aged 55 years and over living in urban and rural areas of Lhasa City. Chin J Neurol 2002;35:336–8. [In Chinese]. Zheng X, Wang YD, Yao DK, Cirenzhuoma, Tang J, He Y. Investigation on prevalence and influence factors of hypertension in Lhasa City. Chin. Med. 2011;6:390–2. [In Chinese].