partial acf -0.10 -0.05 0.00 0.05 0.10...

33
Supplementary data Figure S1. Diagnostics of models: partial autocorrelation function of the residuals. 0 5 10 15 20 25 30 -0.10 -0.05 0.00 0.05 0.10 0.15 Lag Partial ACF ACF of response residuals

Upload: others

Post on 21-Jan-2020

22 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Partial ACF -0.10 -0.05 0.00 0.05 0.10 0download.lww.com/.../MD/A/MD_2015_12_03_HAGIHARA_MD...Supplementary data Figure S1. Diagnostics of models: partial autocorrelation function

Supplementary data Figure S1. Diagnostics of models: partial autocorrelation function of the

residuals.

0 5 10 15 20 25 30

-0.10

-0.05

0.00

0.05

0.10

0.15

Lag

Partial ACF

ACF of response residuals

Page 2: Partial ACF -0.10 -0.05 0.00 0.05 0.10 0download.lww.com/.../MD/A/MD_2015_12_03_HAGIHARA_MD...Supplementary data Figure S1. Diagnostics of models: partial autocorrelation function

Supplementary data Figure S2. Overall cumulative exposure-response associations between the

RRs (95% CI) of emergency transport for all-cause morbidity and temperatures in 47 prefectures,

Japan. Exposure-response associations as the best linear unbiased prediction (with 95% empirical

CI, shaded grey), with the related temperature distributions. The solid grey lines indicate the

minimum mortality temperatures, and the dashed lines indicate the 2.5th and 97.5th percentiles.

Page 3: Partial ACF -0.10 -0.05 0.00 0.05 0.10 0download.lww.com/.../MD/A/MD_2015_12_03_HAGIHARA_MD...Supplementary data Figure S1. Diagnostics of models: partial autocorrelation function
Page 4: Partial ACF -0.10 -0.05 0.00 0.05 0.10 0download.lww.com/.../MD/A/MD_2015_12_03_HAGIHARA_MD...Supplementary data Figure S1. Diagnostics of models: partial autocorrelation function
Page 5: Partial ACF -0.10 -0.05 0.00 0.05 0.10 0download.lww.com/.../MD/A/MD_2015_12_03_HAGIHARA_MD...Supplementary data Figure S1. Diagnostics of models: partial autocorrelation function
Page 6: Partial ACF -0.10 -0.05 0.00 0.05 0.10 0download.lww.com/.../MD/A/MD_2015_12_03_HAGIHARA_MD...Supplementary data Figure S1. Diagnostics of models: partial autocorrelation function
Page 7: Partial ACF -0.10 -0.05 0.00 0.05 0.10 0download.lww.com/.../MD/A/MD_2015_12_03_HAGIHARA_MD...Supplementary data Figure S1. Diagnostics of models: partial autocorrelation function

Supplementary data Figure S3. Overall cumulative exposure-response associations between the

RRs (95% CI) of emergency transport for cardiovascular diseases and temperatures in 47

prefectures, Japan. The exposure-response associations expressed as the best linear unbiased

prediction (with 95% empirical CI, shaded grey), with the related temperature distributions. The

solid grey lines indicate the minimum mortality temperatures, and the dashed lines indicate the

2.5th and 97.5th percentiles.

Page 8: Partial ACF -0.10 -0.05 0.00 0.05 0.10 0download.lww.com/.../MD/A/MD_2015_12_03_HAGIHARA_MD...Supplementary data Figure S1. Diagnostics of models: partial autocorrelation function
Page 9: Partial ACF -0.10 -0.05 0.00 0.05 0.10 0download.lww.com/.../MD/A/MD_2015_12_03_HAGIHARA_MD...Supplementary data Figure S1. Diagnostics of models: partial autocorrelation function
Page 10: Partial ACF -0.10 -0.05 0.00 0.05 0.10 0download.lww.com/.../MD/A/MD_2015_12_03_HAGIHARA_MD...Supplementary data Figure S1. Diagnostics of models: partial autocorrelation function
Page 11: Partial ACF -0.10 -0.05 0.00 0.05 0.10 0download.lww.com/.../MD/A/MD_2015_12_03_HAGIHARA_MD...Supplementary data Figure S1. Diagnostics of models: partial autocorrelation function
Page 12: Partial ACF -0.10 -0.05 0.00 0.05 0.10 0download.lww.com/.../MD/A/MD_2015_12_03_HAGIHARA_MD...Supplementary data Figure S1. Diagnostics of models: partial autocorrelation function
Page 13: Partial ACF -0.10 -0.05 0.00 0.05 0.10 0download.lww.com/.../MD/A/MD_2015_12_03_HAGIHARA_MD...Supplementary data Figure S1. Diagnostics of models: partial autocorrelation function
Page 14: Partial ACF -0.10 -0.05 0.00 0.05 0.10 0download.lww.com/.../MD/A/MD_2015_12_03_HAGIHARA_MD...Supplementary data Figure S1. Diagnostics of models: partial autocorrelation function
Page 15: Partial ACF -0.10 -0.05 0.00 0.05 0.10 0download.lww.com/.../MD/A/MD_2015_12_03_HAGIHARA_MD...Supplementary data Figure S1. Diagnostics of models: partial autocorrelation function
Page 16: Partial ACF -0.10 -0.05 0.00 0.05 0.10 0download.lww.com/.../MD/A/MD_2015_12_03_HAGIHARA_MD...Supplementary data Figure S1. Diagnostics of models: partial autocorrelation function
Page 17: Partial ACF -0.10 -0.05 0.00 0.05 0.10 0download.lww.com/.../MD/A/MD_2015_12_03_HAGIHARA_MD...Supplementary data Figure S1. Diagnostics of models: partial autocorrelation function
Page 18: Partial ACF -0.10 -0.05 0.00 0.05 0.10 0download.lww.com/.../MD/A/MD_2015_12_03_HAGIHARA_MD...Supplementary data Figure S1. Diagnostics of models: partial autocorrelation function

Supplementary data Figure S4. Overall cumulative exposure-response associations between the RRs (95% CI) of emergency

transport for respiratory diseases and temperatures in 47 prefectures, Japan. Exposure-response associations expressed as the best

linear unbiased prediction (with 95% empirical CI, shaded grey), with the related temperature distributions. The solid grey lines

indicate the minimum mortality temperatures, and the dashed lines indicate the 2.5th and 97.5th percentiles.

Page 19: Partial ACF -0.10 -0.05 0.00 0.05 0.10 0download.lww.com/.../MD/A/MD_2015_12_03_HAGIHARA_MD...Supplementary data Figure S1. Diagnostics of models: partial autocorrelation function
Page 20: Partial ACF -0.10 -0.05 0.00 0.05 0.10 0download.lww.com/.../MD/A/MD_2015_12_03_HAGIHARA_MD...Supplementary data Figure S1. Diagnostics of models: partial autocorrelation function
Page 21: Partial ACF -0.10 -0.05 0.00 0.05 0.10 0download.lww.com/.../MD/A/MD_2015_12_03_HAGIHARA_MD...Supplementary data Figure S1. Diagnostics of models: partial autocorrelation function
Page 22: Partial ACF -0.10 -0.05 0.00 0.05 0.10 0download.lww.com/.../MD/A/MD_2015_12_03_HAGIHARA_MD...Supplementary data Figure S1. Diagnostics of models: partial autocorrelation function
Page 23: Partial ACF -0.10 -0.05 0.00 0.05 0.10 0download.lww.com/.../MD/A/MD_2015_12_03_HAGIHARA_MD...Supplementary data Figure S1. Diagnostics of models: partial autocorrelation function
Page 24: Partial ACF -0.10 -0.05 0.00 0.05 0.10 0download.lww.com/.../MD/A/MD_2015_12_03_HAGIHARA_MD...Supplementary data Figure S1. Diagnostics of models: partial autocorrelation function
Page 25: Partial ACF -0.10 -0.05 0.00 0.05 0.10 0download.lww.com/.../MD/A/MD_2015_12_03_HAGIHARA_MD...Supplementary data Figure S1. Diagnostics of models: partial autocorrelation function
Page 26: Partial ACF -0.10 -0.05 0.00 0.05 0.10 0download.lww.com/.../MD/A/MD_2015_12_03_HAGIHARA_MD...Supplementary data Figure S1. Diagnostics of models: partial autocorrelation function
Page 27: Partial ACF -0.10 -0.05 0.00 0.05 0.10 0download.lww.com/.../MD/A/MD_2015_12_03_HAGIHARA_MD...Supplementary data Figure S1. Diagnostics of models: partial autocorrelation function
Page 28: Partial ACF -0.10 -0.05 0.00 0.05 0.10 0download.lww.com/.../MD/A/MD_2015_12_03_HAGIHARA_MD...Supplementary data Figure S1. Diagnostics of models: partial autocorrelation function
Page 29: Partial ACF -0.10 -0.05 0.00 0.05 0.10 0download.lww.com/.../MD/A/MD_2015_12_03_HAGIHARA_MD...Supplementary data Figure S1. Diagnostics of models: partial autocorrelation function

Supplementary data Table S1. Descriptive statistics for the 47 prefectures of Japan from 2007 to 2010: the total number of cases for

emergency transport; the mean daily temperature distribution (°C).

Mean daily temperature (°C) Prefecture Population (2010) No. of cases

Mean SD Min 2.5th Median 97.5th Max Hokkaido 5,506,419 765,010 9.6 9.4 -10.7 -5.6 9.7 24.6 29.3 Aomori 1,373,339 132,934 10.9 8.8 -6.8 -3.1 11.2 26.1 30.1 Iwate 1,330,147 149,111 10.8 9.2 -7.6 -3.3 10.8 26.3 29.3 Miyagi 2,348,165 300,401 13.0 8.1 -4.5 0.2 13.1 27.5 31.2 Akita 1,085,997 130,503 12.3 8.9 -4.9 -1.6 12.3 26.6 30.3 Yamagata 1,168,924 147,049 12.2 9.2 -5.8 -1.9 12.3 27.7 30.4 Fukushima 2,029,064 256,770 13.6 8.6 -4.2 0.3 13.5 28.9 31.4 Ibaraki 2,969,770 387,516 14.3 8.0 -0.1 1.8 14.5 28.2 31.0 Tochigi 2,007,683 250,492 14.6 8.3 -1.1 1.4 14.9 28.6 31.4 Gunma 2,008,068 268,005 15.2 8.4 0.2 2.1 15.3 29.5 32.2 Saitama 7,194,556 978,341 15.7 8.3 0.2 2.8 15.9 30.0 33.7 Chiba 6,216,289 881,542 16.5 7.6 1.4 4.3 16.9 29.5 30.9 Tokyo* 112,388 22,705 16.8 7.7 1.5 4.7 17.0 30.1 32.7 Kanagawa 9,048,331 1,279,141 16.4 7.5 1.2 4.6 16.8 29.1 30.9 Niigata 2,374,450 309,910 14.3 8.4 -1.6 1.1 14.1 29.0 31.1 Toyama 1,093,247 111,473 14.8 8.6 -1.3 0.9 14.9 29.5 31.6 Ishikawa 1,169,788 135,997 15.0 8.4 -0.5 1.6 15.1 29.4 31.1 Fukui 806,314 85,426 14.9 8.7 -0.5 1.3 15.0 29.8 31.9 Yamanashi 863,075 125,386 15.2 8.4 -0.8 1.6 15.2 28.8 30.3 Nagano 2,152,449 293,702 12.5 9.2 -4.3 -1.9 12.5 27.3 29.8 Gifu 2,080,773 276,925 16.3 8.4 0.0 2.7 16.5 30.4 32.7

Page 30: Partial ACF -0.10 -0.05 0.00 0.05 0.10 0download.lww.com/.../MD/A/MD_2015_12_03_HAGIHARA_MD...Supplementary data Figure S1. Diagnostics of models: partial autocorrelation function

Shizuoka 3,765,007 506,707 17.0 7.3 3.1 5.0 17.3 29.0 30.2 Aichi 7,410,719 1,010,255 16.5 8.3 0.4 3.0 16.6 30.3 32.7 Mie 1,854,724 281,273 16.5 8.0 0.6 4.1 16.6 29.8 32.2 Shiga 1,410,777 183,586 15.2 8.4 -0.2 2.3 15.1 29.4 30.7 Kyoto 2,636,092 430,243 16.3 8.5 0.2 2.8 16.4 30.4 31.5 Osaka 8,865,245 1,494,417 17.3 8.2 2.0 4.4 17.4 30.7 32.5 Hyogo 5,588,133 747,856 17.2 8.0 1.3 4.1 17.5 29.9 31.8 Nara 1,400,728 200,425 15.2 8.3 -0.1 2.3 15.1 29.0 30.4 Wakayama 1,002,198 171,964 17.0 7.9 1.8 4.5 17.2 29.5 31.5 Tottori 588,667 80,951 15.3 8.3 -0.4 1.8 15.1 29.6 32.0 Shimane 717,397 96,590 15.4 8.1 -0.3 2.1 15.4 29.4 32.2 Okayama 1,945,276 283,424 16.8 8.5 0.7 3.1 16.9 30.5 32.0 Hiroshima 2,860,750 416,944 16.6 8.3 0.3 3.4 16.8 30.3 31.6 Yamaguchi 1,451,338 229,032 16.0 8.3 -0.6 2.6 16.1 29.4 30.4 Tokushima 785,491 94,996 17.1 7.9 0.7 4.2 17.3 29.6 31.8 Kagawa 995,842 159,060 17.0 8.3 1.3 4.0 17.2 30.5 31.9 Ehime 1,431,493 219,779 17.0 7.9 1.6 4.4 17.1 29.9 31.7 Kochi 764,456 133,692 17.6 7.6 1.9 4.6 18.0 29.5 30.8 Fukuoka 5,071,968 551,048 17.5 7.8 0.7 4.7 17.8 30.4 31.8 Saga 849,788 116,380 17.0 8.2 0.0 3.6 17.2 29.9 32.0 Nagasaki 1,426,779 199,861 17.6 7.6 0.7 4.8 17.9 29.6 31.4 Kumamoto 1,817,426 256,704 17.5 8.2 0.4 3.9 18.0 30.2 31.3 Oita 1,196,529 171,588 17.1 7.7 0.9 4.6 17.2 29.6 31.2 Miyazaki 1,135,233 124,461 17.8 7.3 2.0 5.3 18.1 29.2 31.2 Kagoshima 1,706,242 215,291 19.0 7.4 0.7 6.1 19.4 30.0 30.9 Okinawa 1,392,818 203,220 23.4 4.6 11.5 14.6 23.3 29.9 31.1 *Tokyo Metropolitan area was excluded because of the lack of data on emergency ambulance dispatches.

Page 31: Partial ACF -0.10 -0.05 0.00 0.05 0.10 0download.lww.com/.../MD/A/MD_2015_12_03_HAGIHARA_MD...Supplementary data Figure S1. Diagnostics of models: partial autocorrelation function
Page 32: Partial ACF -0.10 -0.05 0.00 0.05 0.10 0download.lww.com/.../MD/A/MD_2015_12_03_HAGIHARA_MD...Supplementary data Figure S1. Diagnostics of models: partial autocorrelation function

Supplementary data Table S2. Sensitivity analysis. Using the fraction (%) attributable to temperature (total, high and low

temperature components), by varying the modeling choices, controlling for different degrees of freedom for the seasonal and long-term

trends (6 and 10 df per year) and relative humidity.

Total (%) Low temperature (%) High temperature (%)

All causes (A00-Z99)

Modelling choices (47 prefectrures)

Main model 7.95 (6.90‒ 8.84) 6.94 (5.93‒ 7.70) 1.01 (0.90‒ 1.11)

Df/year for seasonal control: 6 8.75 (8.12‒ 9.33) 6.81 (6.15‒ 7.37) 1.93 (1.67‒ 2.19)

Df/year for seasonal control: 10 8.82 (7.61‒ 9.93) 6.65 (5.88‒ 7.32) 2.17 (1.13‒ 3.14)

Control for relative himidity (47 prefectrures)

With relative humidity 7.95 (6.90‒ 8.84) 6.94 (5.93‒ 7.70) 1.01 (0.90‒ 1.11)

Without relative humidity 8.08 (6.99‒ 8.99) 6.92 (5.89‒ 7.79) 1.17 (1.04‒ 1.28)

Cardiovascular diseases (I00-I99)

Modelling choices (47 prefectrures)

Main model 18.02 (16.32‒ 19.32) 17.93 (16.10‒ 19.25) 0.10 (0.04‒ 0.14)

Df/year for seasonal control: 6 15.58 (13.21‒ 17.16) 15.31 (12.89‒ 16.89) 0.27 (-0.08‒ 0.56)

Df/year for seasonal control: 10 15.55 (13.22‒ 17.10) 15.32 (13.04‒ 16.92) 0.23 (0.12‒ 0.32)

Control for relative himidity (47 prefectrures)

Page 33: Partial ACF -0.10 -0.05 0.00 0.05 0.10 0download.lww.com/.../MD/A/MD_2015_12_03_HAGIHARA_MD...Supplementary data Figure S1. Diagnostics of models: partial autocorrelation function

With relative humidity 18.02 (16.32‒ 19.32) 17.93 (16.10‒ 19.25) 0.10 (0.04‒ 0.14)

Without relative humidity 17.26 (15.44‒ 18.53) 17.16 (15.33‒ 18.49) 0.10 (0.04‒ 0.15)

Respiratory diseases (J00-J99)

Modelling choices (47 prefectrures)

Main model 12.48 (10.30‒ 14.10) 12.19 (9.90‒ 13.66) 0.29 (0.07‒ 0.50)

Df/year for seasonal control: 6 9.60 (8.04‒ 10.81) 7.14 (5.52‒ 8.19) 2.46 (1.80‒ 3.00)

Df/year for seasonal control: 10 13.52 (12.01‒ 14.59) 11.63 (10.24‒ 12.70) 1.89 (1.49‒ 2.22)

Control for relative himidity (47 prefectrures)

With relative humidity 12.48 (10.30‒ 14.10) 12.19 (9.90‒ 13.66) 0.29 (0.07‒ 0.50)

Without relative humidity 12.58 (10.32‒ 14.25) 12.35 (10.27‒ 13.99) 0.23 (0.01‒ 0.43)