| Background:Previous studies mainly focused on the impacts of non-optimal temperature and extreme temperature events on human health.However,there is insufficient evidence showing the effects of the temperature variability(TV)on deaths from cardiovascular diseases(CVD).Meanwhile,there are very few studies focusing on the effect modification of TV-cardiovascular mortality association by air pollutants.Objectives:This study aims to explore the association between the TV and daily CVD mortality based on a TV indicator that simultaneously allows for the inter-and intra-day temperature change,to identify the corresponding susceptible population,and season pattern,and to determine the potential effect modification by air pollutants in three Chinese cities(i.e.Beijing,Nanjing,Chengdu).Methods:City-specific data of daily CVD(ICD-10,100-199)mortality from 2008 to 2011 were derived from the causes of Death Registry of Chinese Center for the Disease Control(China CDC).The data of air pollutants,including inhalable particulate matter(PMio),sulphur dioxide(SO2)and nitrogen dioxide(NO2),and meteorological factors,such as daily mean temperature,maximum temperature and minimum temperature,for the three cities in the same period,were collected from Chinese National Air Quality Monitoring Stations(NAQMS)and China Meteorological Administration,respectively.TV was calculated as the standard deviation of daily maximum temperature and minimum temperature over the defined cumulative exposure days.Specifically,we constructed a series of TV indicators with different cumulative exposure days,ranging from TV01 to TV08.TV01 represent the TV over current day and preceding one day.A Two stage time series analysis was adopted to assess the association between TV and daily cardiovascular mortality and to determine the corresponding exposure-response relationship as well as to explore the effect modification by air pollutants.In the first stage,a flexible Poisson regression model,which allows for over-dispersion within the framework of distributed lag nonlinear model,was fitted to estimate the city-specific TV-CVD mortality association.In the second stage,a multivariate meta-analysis was used to pool the city-specific estimates.Within this two stage analyses framework,stratified analyses by gender,age and season were conducted to explore their effect modification in terms of TV-CVD mortality association.Meanwhile,the concentration of each air pollutant was dividied into three strata(i.e.low,middle,high level)with the 25th and 75th percentiles of each pollutant's concentration as the cut-off points.An interaction term of TV and a three-level air pollutant stratum indicator was then included in the models within this two stage analyses framework.Effect modification by ambient air pollutants were assessed by comparing the estimates of TV's effects between compared pollutant stratums and calculating the corresponding 95%confidential interval(CI)of the differences.Results:(1)A total of 260,603 CVD deaths were observed in three cities during 2008-2011.Among the total CVD deaths,145,477 cases were registered in Beijing,61,274 cases were registered in Nanjing,and 83,852 cases were registered in Chengdu.Daily mean TV of Beijing,Nanjing and Chengdu during the study period was 5.67 ?,4.90 ? and 4.29 ? repectively.(2)For the pooled estimate,TV was significantly positively associated with the increased risk of cardiovascular mortality after adjusting for the potential confounders,especially for longer TV exposure days(i.e.after TV04).This result indicated the independent effects of TV for cardiovascular mortality.The highest estimate was observed at TV08,with the estimate of 1.37%(95%CI:0.78-1.97%)increase in daily cardiovascular mortality per 1 ? in TV.(3)Generally,the pooled estimates by gender and age suggested that female and elderly(age>65yrs)are the susceptible population.The pattern of changes in estimations of TV-cardiovascular mortality across TV exposure days was similar to that in the whole population.But the differences in estimates of TV's mortality effect between gender and age groups were insignificant.(4)Further stratified analyses by season indicated the existence of season variation of TV-cardiovascular mortality association.For longer TV exposure days,mortality effects of TV in cool season were stronger than that in warm season,while this trend was reversed at shorter TV exposure days(i.e.before TV04).But significant differences in estimates between two season were absent across all the TV exposure days.(5)Nonlinear exposure-relationships for the association between TV and cardiovascular mortality were observed,with the minimum mortality TV varied across cities and TV exposures days.(6)Among the three air pollutants,only PM10 can modify the association between TV and cardiovascular mortality,with apparent increasing trend of TV's mortality effects across the three PM10 stratums.Specifically,there was a significant difference between the TV's mortality effects in PM10 low stratum and that in PMio high stratum.This trend also persisted in subgroups,specifically in females and the elderly.Conclusion:The findings of present study confirmed the previous observed independent adverse effects of temperature variability on cardiovascular mortality,adding the evidence of adverse effects of climate change on human.The identified effect modification by PM10 indicates that it is imperative and meaningful to take the collective measures to protect the public health from the combined adverse effects brought by temperature variability and air pollution. |
PDF Full Text Request