Dose-Response Curves for PM10 and Non-accidental Mortality for 4 different periods

Analysis are done by group of city, according to the availability of the data

G1 - Three period considered. - City with data available from 2001 to 2015: Clermont Ferrand, Le Havre, Marseille, Nantes, Paris, Rouen, Strasbourg. - City with data available from 2002 to 2015: Bordeaux, Grenoble(only from 2003), Lille, Rennes, Toulouse. Three period considered:

G2 - Only two periods Lyon (from 2007), Nancy (from 2005), Nice (from 2006, without 2009)

Descriptive Statistics for PM10

Descriptive Statistics for pm10
Ville Minimum X1st.Qu Median Mean X3rd.Qu Max NA.s
bordeaux 0.0000000 17.00000 22.33333 24.68781 29.49141 101.66667 1589
clermont 2.0000000 12.66667 17.50000 19.80135 24.00000 105.75000 1102
grenoble 3.0000000 17.00000 24.00000 27.29601 34.00000 110.50000 1551
lehavre 5.0416667 16.79354 21.66667 24.75754 29.33333 123.66667 1145
lille 0.0000000 16.50000 22.00000 25.53733 30.50000 131.00000 1593
lyon 3.0729167 16.11979 22.69792 26.41374 32.00000 167.00000 3378
marseille 4.8229167 22.66667 30.33333 31.90142 39.00000 120.00000 1162
montpellier 1.0000000 13.67437 21.00000 22.40362 29.00000 131.41667 1481
nancy 0.0000000 14.00000 19.71875 22.13904 27.00000 178.00000 2473
nantes 4.5000000 14.72830 19.00000 21.44513 25.33333 112.00000 1153
nice -1.0000000 22.91667 28.45312 29.39850 34.85417 127.58333 3131
paris 7.0000000 22.50000 28.80000 31.49370 37.32765 153.50000 1125
rennes 0.0000000 12.50000 17.00000 19.34861 23.00000 101.50000 1528
rouen 0.3958333 17.67969 22.50000 25.34694 29.50000 116.50000 1137
strasbourg 2.0000000 16.25000 22.66667 25.45405 31.33333 167.00000 1204
toulouse 3.2000000 14.44066 19.60000 21.15854 26.00000 95.33333 1230

Descriptive Statistics for pm10 by period

Meta-Regression results

## Call:  mvmeta(formula = y0105 ~ 1, S = S0105, method = "ml")
## 
## Multivariate random-effects meta-analysis
## Dimension: 3
## Estimation method: ML
## 
## Fixed-effects coefficients
##     Estimate  Std. Error        z  Pr(>|z|)  95%ci.lb  95%ci.ub   
## y1   -0.0200      0.0148  -1.3472    0.1779   -0.0491    0.0091   
## y2    0.0046      0.0284   0.1612    0.8719   -0.0510    0.0602   
## y3    0.0242      0.0251   0.9657    0.3342   -0.0249    0.0733   
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1 
## 
## Between-study random-effects (co)variance components
## Structure: General positive-definite
##     Std. Dev  Corr    
## y1    0.0266    y1  y2
## y2    0.0211     1    
## y3    0.0009     1   1
## 
## Multivariate Cochran Q-test for heterogeneity:
## Q = 36.4469 (df = 36), p-value = 0.4479
## I-square statistic = 1.2%
## 
## 13 studies, 39 observations, 3 fixed and 6 random-effects parameters
##   logLik       AIC       BIC  
##  43.5105  -69.0210  -54.0489
## Call:  mvmeta(formula = y0610 ~ 1, S = S0610, method = "ml")
## 
## Multivariate random-effects meta-analysis
## Dimension: 3
## Estimation method: ML
## 
## Fixed-effects coefficients
##     Estimate  Std. Error       z  Pr(>|z|)  95%ci.lb  95%ci.ub   
## y1    0.0106      0.0107  0.9896    0.3224   -0.0104    0.0316   
## y2    0.0119      0.0248  0.4803    0.6310   -0.0367    0.0606   
## y3    0.0334      0.0262  1.2734    0.2029   -0.0180    0.0848   
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1 
## 
## Between-study random-effects (co)variance components
## Structure: General positive-definite
##     Std. Dev     Corr         
## y1    0.0008       y1       y2
## y2    0.0079  -0.9992         
## y3    0.0072  -0.9995   1.0000
## 
## Multivariate Cochran Q-test for heterogeneity:
## Q = 44.0096 (df = 45), p-value = 0.5138
## I-square statistic = 1.0%
## 
## 16 studies, 48 observations, 3 fixed and 6 random-effects parameters
##   logLik       AIC       BIC  
##  57.3472  -96.6944  -79.8536
## Call:  mvmeta(formula = y1115 ~ 1, S = S1115, method = "ml")
## 
## Multivariate random-effects meta-analysis
## Dimension: 3
## Estimation method: ML
## 
## Fixed-effects coefficients
##     Estimate  Std. Error        z  Pr(>|z|)  95%ci.lb  95%ci.ub   
## y1    0.0165      0.0107   1.5363    0.1245   -0.0045    0.0375   
## y2   -0.0161      0.0326  -0.4937    0.6215   -0.0801    0.0479   
## y3   -0.0448      0.0266  -1.6869    0.0916   -0.0969    0.0073  .
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1 
## 
## Between-study random-effects (co)variance components
## Structure: General positive-definite
##     Std. Dev  Corr    
## y1    0.0104    y1  y2
## y2    0.0711     1    
## y3    0.0448     1   1
## 
## Multivariate Cochran Q-test for heterogeneity:
## Q = 41.5808 (df = 45), p-value = 0.6176
## I-square statistic = 1.0%
## 
## 16 studies, 48 observations, 3 fixed and 6 random-effects parameters
##    logLik        AIC        BIC  
##   62.2302  -106.4605   -89.6196

Pooled Dose-Responses Curves

Doses-Response Curves by cities

Bordeaux

Clermont-Ferrand

Grenoble

Le Havre

Lille

Lyon

Marseille

Montpellier

Nancy

Nantes

Nice

Paris

Rennes

Rouen

Strasbourg

Toulouse