brfss-analysis/cntwts_asthnow.R at master · rebeccagreenblatt/brfss-analysis · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
library(survey)
library(car)
library(dplyr)

BRFSS.complete.allvars <- read.csv("brfss.complete.wcntywts.csv")
BRFSS.complete.allvars$CntyFIPS <- as.integer(gsub(",","", as.character(BRFSS.complete.allvars$CntyFIPS)))
BRFSS.complete.allvars$STSTR <- BRFSS.complete.allvars$CntyFIPS #think this is what STSTR was in Omar's file...
names(BRFSS.complete.allvars)[names(BRFSS.complete.allvars)=="INCOME2"] <- "INCOME"
names(BRFSS.complete.allvars)[names(BRFSS.complete.allvars)=="RACE2"] <- "RACE"
names(BRFSS.complete.allvars)[names(BRFSS.complete.allvars)=="BMICAT"] <- "BMI"

##I might be wrongly assigning CNTYFIPS to STSTR variable... don't know where STSTR came from

BRFSS.complete <- select(BRFSS.complete.allvars, id, ASTHNOW, AGE, EDUCA, INCOME, SEX, SMOKER, BMI, CNTYWT, RACE, CntyFIPS)

#--FOR ASTHNOW --
##need to confirm that 0 also means N/A... responses 1 and 2 roughly add up to people who responded yes to ever having asthma, so just sticking with these
BRFSS.complete$ASTHNOW <- recode(BRFSS.complete$ASTHNOW, recodes = "1='1'; 2='0'; else = NA", as.factor.result=T)
BRFSS.complete$ASTHNOW<-relevel(BRFSS.complete$ASTHNOW, ref='0')
#------------------------

#Race/ethnicity
BRFSS.complete$RACE <- recode(BRFSS.complete$RACE, recodes="1='white'; 2='black'; 3:4='asian/pacific islander'; 5='native'; 8='hispanic'; c(6,7,9,0)=NA", as.factor.result=T)
BRFSS.complete$RACE <- relevel(BRFSS.complete$RACE, ref='white')

#Income <$25K, $25-75K, >$75k
BRFSS.complete$INCOME <- recode(BRFSS.complete$INCOME, recodes="1:4='low'; 5:7='middle'; 8='high'; c(0,77,99)=NA", as.factor.result=T)
BRFSS.complete$INCOME <- relevel(BRFSS.complete$INCOME, ref='high')

#Age
BRFSS.complete$AGE <- recode(BRFSS.complete$AGE, recodes="0:9=NA", as.numeric.result=T)

#Education
BRFSS.complete$EDUCA <- recode(BRFSS.complete$EDUCA, recodes="1:2='noHighSchool'; 3:4='Some/highSchool'; 5:6='higherEducation';c(0,9)=NA", as.factor.result=T)
BRFSS.complete$EDUCA <- relevel(BRFSS.complete$EDUCA, ref='higherEducation')

#Sex
BRFSS.complete$SEX <- recode(BRFSS.complete$SEX, recodes="1='male'; 2='female'", as.factor.result=T)
BRFSS.complete$SEX <- relevel(BRFSS.complete$SEX, ref='male')

#Smoking
BRFSS.complete$SMOKER <- recode(BRFSS.complete$SMOKER, recodes="1:2='current smoker'; 3='former smoker'; 4='never smoked';9=NA", as.factor.result=T)
BRFSS.complete$SMOKER <- relevel(BRFSS.complete$SMOKER, ref='never smoked')

#BMI
BRFSS.complete$BMI <- recode(BRFSS.complete$BMI, recodes="1='neither overweight nor obese'; 2='overweight'; 3='obese';c(0,9)=NA", as.factor.result=T)
BRFSS.complete$BMI <- relevel(BRFSS.complete$BMI, ref='neither overweight nor obese')

#BRFSS.complete <- BRFSS.complete[ -c(1) ]

BRFSS.complete <- BRFSS.complete[ which(BRFSS.complete$ASTHNOW !='NA' ), ]
BRFSS.test <- BRFSS.complete[complete.cases(BRFSS.complete), ]

# BRFSS.complete.asthnow <- BRFSS.complete[ which(BRFSS.complete$ASTHNOW !='NA' ), ]
# BRFSS.test.asthnow <- BRFSS.complete.asthnow[complete.cases(BRFSS.complete.asthnow), ]

##joining with epa
epa <- read.csv("epa_final.csv")
ozone_dat <- filter(epa, Parameter_Name == "Ozone")
pm2.5_dat <- filter(epa, Parameter_Name == "PM2.5 - Local Conditions")

dat_pm2.5 <- inner_join(pm2.5_dat, BRFSS.test, by = c("id", "CntyFIPS"))
brfss.pm2.5 <- dat_pm2.5[complete.cases(dat_pm2.5),]

dat_ozone <- inner_join(ozone_dat, BRFSS.test, by = c("id", "CntyFIPS"))
brfss.ozone <- dat_ozone[complete.cases(dat_ozone),]

options(survey.lonely.psu = "adjust")

##ASTHNOW + pm2.5 - local
des.pm2.5 <- svydesign(ids=~1, strata=~CntyFIPS, weights=~as.numeric(CNTYWT),  data=brfss.pm2.5)
model.pm2.5 <- svyglm(ASTHNOW ~ AGE+EDUCA+INCOME+RACE+SEX+SMOKER+BMI+median+sd, des.pm2.5, family=quasibinomial)

summary(model.pm2.5)
exp(coefficients(model.pm2.5))
#
# > summary(model.pm2.5)
#
# Call:
#   svyglm(formula = ASTHNOW ~ AGE + EDUCA + INCOME + RACE + SEX +
#            SMOKER + BMI + median + sd, des.pm2.5, family = quasibinomial)
#
# Survey design:
#   svydesign(ids = ~1, strata = ~CntyFIPS, weights = ~as.numeric(CNTYWT),
#             data = brfss.pm2.5)
#
# Coefficients:
#   Estimate Std. Error t value Pr(>|t|)
# (Intercept)                -0.4405137  0.0669980  -6.575 4.88e-11 ***
#   AGE                         0.0106240  0.0007568  14.039  < 2e-16 ***
#   EDUCAnoHighSchool           0.1096698  0.0936850   1.171 0.241752
# EDUCASome/highSchool        0.1236069  0.0291337   4.243 2.21e-05 ***
#   INCOMElow                   0.4156302  0.0345153  12.042  < 2e-16 ***
#   INCOMEmiddle                0.1211291  0.0274836   4.407 1.05e-05 ***
#   RACEasian/pacific islander -0.3753240  0.0708074  -5.301 1.16e-07 ***
#   RACEblack                  -0.0184329  0.0367048  -0.502 0.615534
# RACEhispanic               -0.4249085  0.0418630 -10.150  < 2e-16 ***
#   RACEnative                  0.0221174  0.1264066   0.175 0.861103
# SEXfemale                   0.5945792  0.0238388  24.942  < 2e-16 ***
#   SMOKERcurrent smoker        0.0235573  0.0323997   0.727 0.467176
# SMOKERformer smoker        -0.0433790  0.0269873  -1.607 0.107972
# BMIobese                    0.3211626  0.0292133  10.994  < 2e-16 ***
#   BMIoverweight               0.1012464  0.0285512   3.546 0.000391 ***
#   median                     -0.0167379  0.0053512  -3.128 0.001761 **
#   sd                          0.0216145  0.0072667   2.974 0.002936 **
#   ---
#   Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
#
# (Dispersion parameter for quasibinomial family taken to be 1.000705)
#
# Number of Fisher Scoring iterations: 4
#
# > exp(coefficients(model.pm2.5))
# (Intercept)                        AGE          EDUCAnoHighSchool
# 0.6437057                  1.0106807                  1.1159095
# EDUCASome/highSchool                  INCOMElow               INCOMEmiddle
# 1.1315710                  1.5153254                  1.1287706
# RACEasian/pacific islander                  RACEblack               RACEhispanic
# 0.6870667                  0.9817360                  0.6538296
# RACEnative                  SEXfemale       SMOKERcurrent smoker
# 1.0223638                  1.8122681                  1.0238369
# SMOKERformer smoker                   BMIobese              BMIoverweight
# 0.9575484                  1.3787298                  1.1065493
# median                         sd
# 0.9834014                  1.0218497

##ASTHNOW + OZONE
options(survey.lonely.psu = "adjust")
des.ozone <- svydesign(ids=~1, strata=~CntyFIPS, weights=~as.numeric(CNTYWT),  data=brfss.ozone)
model.ozone <- svyglm(ASTHNOW ~ AGE+EDUCA+INCOME+RACE+SEX+SMOKER+BMI+F95th, des.ozone, family=quasibinomial)

summary(model.ozone)
exp(coefficients(model.ozone))
# > summary(model.ozone)
#
# Call:
#   svyglm(formula = ASTHNOW ~ AGE + EDUCA + INCOME + RACE + SEX +
#            SMOKER + BMI + F95th, des.ozone, family = quasibinomial)
#
# Survey design:
#   svydesign(ids = ~1, strata = ~CntyFIPS, weights = ~as.numeric(CNTYWT),
#             data = brfss.ozone)
#
# Coefficients:
#   Estimate Std. Error t value Pr(>|t|)
# (Intercept)                -0.7545420  0.0941188  -8.017 1.09e-15 ***
#   AGE                         0.0108163  0.0007597  14.238  < 2e-16 ***
#   EDUCAnoHighSchool           0.0991454  0.0950921   1.043 0.297124
# EDUCASome/highSchool        0.1283190  0.0294166   4.362 1.29e-05 ***
#   INCOMElow                   0.3997143  0.0348146  11.481  < 2e-16 ***
#   INCOMEmiddle                0.1031728  0.0277029   3.724 0.000196 ***
#   RACEasian/pacific islander -0.3591269  0.0720295  -4.986 6.18e-07 ***
#   RACEblack                  -0.0376827  0.0366266  -1.029 0.303559
# RACEhispanic               -0.4182218  0.0422150  -9.907  < 2e-16 ***
#   RACEnative                  0.0641860  0.1321953   0.486 0.627294
# SEXfemale                   0.5954936  0.0240630  24.747  < 2e-16 ***
#   SMOKERcurrent smoker        0.0214019  0.0328730   0.651 0.515016
# SMOKERformer smoker        -0.0387686  0.0271497  -1.428 0.153306
# BMIobese                    0.3006913  0.0294776  10.201  < 2e-16 ***
#   BMIoverweight               0.0978268  0.0288447   3.391 0.000695 ***
#   F95th                       4.0444382  1.1704683   3.455 0.000550 ***
#   ---
#   Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
#
# (Dispersion parameter for quasibinomial family taken to be 1.00016)
#
# Number of Fisher Scoring iterations: 4
#
# > summary(brfss.ozone$F95th)
# Min. 1st Qu.  Median    Mean 3rd Qu.    Max.
# 0.03100 0.06367 0.07017 0.06973 0.07600 0.10570
# > exp(coefficients(model.ozone))
# (Intercept)                        AGE          EDUCAnoHighSchool
# 0.4702259                  1.0108750                  1.1042268
# EDUCASome/highSchool                  INCOMElow               INCOMEmiddle
# 1.1369156                  1.4913985                  1.1086830
# RACEasian/pacific islander                  RACEblack               RACEhispanic
# 0.6982857                  0.9630185                  0.6582162
# RACEnative                  SEXfemale       SMOKERcurrent smoker
# 1.0662907                  1.8139261                  1.0216326
# SMOKERformer smoker                   BMIobese              BMIoverweight
# 0.9619733                  1.3507923                  1.1027718
# F95th
# 57.0791090