Modelos de regresion logistica para erodentina sin encisales

Aca se analiza si la prevalencia de severidad ajustada se asocia con las variables de riesgo

library(survey)
## Attaching package: 'survey'
## The following object(s) are masked from 'package:graphics':
## 
## dotchart
options(OutDec = ",")
library(car)
## Loading required package: MASS
## Loading required package: nnet
load("~/Dropbox/odontologia/maestria licet/octubre 2013/datos_licet_03112013.RData")
rm(list = ls(pattern = "modelo"))

diseniopost1$variables$Erosinbord <- as.factor(diseniopost1$variables$Erosinbord)

Erosinbord

Modelo_ero_1.logit <- svyglm(Erosinbord ~ Sexo, design = diseniopost1, family = quasibinomial())

summary(Modelo_ero_1.logit)
## 
## Call:
## svyglm(formula = Erosinbord ~ Sexo, design = diseniopost1, family = quasibinomial())
## 
## Survey design:
## postStratify(disenio1, ~categor.rec + Sexo, tabla.pob)
## 
## Coefficients:
##             Estimate Std. Error t value Pr(>|t|)    
## (Intercept)   -3,731      0,369  -10,12  2,5e-12 ***
## SexoM          1,045      0,480    2,18    0,036 *  
## ---
## Signif. codes:  0 '***' 0,001 '**' 0,01 '*' 0,05 '.' 0,1 ' ' 1 
## 
## (Dispersion parameter for quasibinomial family taken to be 0,9818)
## 
## Number of Fisher Scoring iterations: 6
confint(Modelo_ero_1.logit)
##               2,5 % 97,5 %
## (Intercept) -4,4543 -3,009
## SexoM        0,1036  1,987

Modelo_ero_2.logit <- svyglm(Erosinbord ~ UsoDentifrico3.rec, design = diseniopost1, 
    family = quasibinomial())
summary(Modelo_ero_2.logit)
## 
## Call:
## svyglm(formula = Erosinbord ~ UsoDentifrico3.rec, design = diseniopost1, 
##     family = quasibinomial())
## 
## Survey design:
## postStratify(disenio1, ~categor.rec + Sexo, tabla.pob)
## 
## Coefficients:
##                        Estimate Std. Error t value Pr(>|t|)    
## (Intercept)              -3,076      0,231  -13,31  6,9e-16 ***
## UsoDentifrico3.rec2-No    0,272      0,882    0,31     0,76    
## ---
## Signif. codes:  0 '***' 0,001 '**' 0,01 '*' 0,05 '.' 0,1 ' ' 1 
## 
## (Dispersion parameter for quasibinomial family taken to be 0,9781)
## 
## Number of Fisher Scoring iterations: 5
confint(Modelo_ero_2.logit)
##                         2,5 % 97,5 %
## (Intercept)            -3,528 -2,623
## UsoDentifrico3.rec2-No -1,456  2,001

Modelo_ero_3.logit <- svyglm(Erosinbord ~ Nivel.Educativo.de.la.Madre1.rec, 
    design = diseniopost1, family = quasibinomial())
summary(Modelo_ero_3.logit)
## 
## Call:
## svyglm(formula = Erosinbord ~ Nivel.Educativo.de.la.Madre1.rec, 
##     design = diseniopost1, family = quasibinomial())
## 
## Survey design:
## postStratify(disenio1, ~categor.rec + Sexo, tabla.pob)
## 
## Coefficients:
##                                          Estimate Std. Error t value
## (Intercept)                                -1,923      0,480   -4,00
## Nivel.Educativo.de.la.Madre1.rec2-Basic    -1,303      0,543   -2,40
## Nivel.Educativo.de.la.Madre1.rec3-Medium   -1,445      0,640   -2,26
## Nivel.Educativo.de.la.Madre1.rec4-High     -1,082      0,652   -1,66
##                                          Pr(>|t|)    
## (Intercept)                                0,0003 ***
## Nivel.Educativo.de.la.Madre1.rec2-Basic    0,0218 *  
## Nivel.Educativo.de.la.Madre1.rec3-Medium   0,0301 *  
## Nivel.Educativo.de.la.Madre1.rec4-High     0,1058    
## ---
## Signif. codes:  0 '***' 0,001 '**' 0,01 '*' 0,05 '.' 0,1 ' ' 1 
## 
## (Dispersion parameter for quasibinomial family taken to be 0,9559)
## 
## Number of Fisher Scoring iterations: 6
confint(Modelo_ero_3.logit)
##                                           2,5 %  97,5 %
## (Intercept)                              -2,865 -0,9817
## Nivel.Educativo.de.la.Madre1.rec2-Basic  -2,368 -0,2380
## Nivel.Educativo.de.la.Madre1.rec3-Medium -2,699 -0,1912
## Nivel.Educativo.de.la.Madre1.rec4-High   -2,360  0,1963

Modelo_ero_3a.logit <- svyglm(Erosinbord ~ Nive.Educativo.de.la.Madre2, design = diseniopost1, 
    family = quasibinomial())
summary(Modelo_ero_3a.logit)
## 
## Call:
## svyglm(formula = Erosinbord ~ Nive.Educativo.de.la.Madre2, design = diseniopost1, 
##     family = quasibinomial())
## 
## Survey design:
## postStratify(disenio1, ~categor.rec + Sexo, tabla.pob)
## 
## Coefficients:
##                                               Estimate Std. Error t value
## (Intercept)                                     -3,508      0,771   -4,55
## Nive.Educativo.de.la.Madre2COLLEGE-UNIVERSITY    0,632      0,660    0,96
## Nive.Educativo.de.la.Madre2ELEMENTARY SCHOOL     0,709      0,871    0,81
## Nive.Educativo.de.la.Madre2HIGH SCHOOL           0,172      0,826    0,21
##                                               Pr(>|t|)    
## (Intercept)                                    5,8e-05 ***
## Nive.Educativo.de.la.Madre2COLLEGE-UNIVERSITY     0,34    
## Nive.Educativo.de.la.Madre2ELEMENTARY SCHOOL      0,42    
## Nive.Educativo.de.la.Madre2HIGH SCHOOL            0,84    
## ---
## Signif. codes:  0 '***' 0,001 '**' 0,01 '*' 0,05 '.' 0,1 ' ' 1 
## 
## (Dispersion parameter for quasibinomial family taken to be 0,9818)
## 
## Number of Fisher Scoring iterations: 6
confint(Modelo_ero_3a.logit)
##                                                 2,5 % 97,5 %
## (Intercept)                                   -5,0185 -1,998
## Nive.Educativo.de.la.Madre2COLLEGE-UNIVERSITY -0,6610  1,925
## Nive.Educativo.de.la.Madre2ELEMENTARY SCHOOL  -0,9976  2,416
## Nive.Educativo.de.la.Madre2HIGH SCHOOL        -1,4479  1,791


Modelo_ero_4.logit <- svyglm(Erosinbord ~ FrCepDenti.rec, design = diseniopost1, 
    family = quasibinomial())
summary(Modelo_ero_4.logit)
## 
## Call:
## svyglm(formula = Erosinbord ~ FrCepDenti.rec, design = diseniopost1, 
##     family = quasibinomial())
## 
## Survey design:
## postStratify(disenio1, ~categor.rec + Sexo, tabla.pob)
## 
## Coefficients:
##                                      Estimate Std. Error t value Pr(>|t|)
## (Intercept)                           -3,2219     0,3824   -8,43  3,9e-10
## FrCepDenti.rec2- 2 veces al dia        0,0495     0,6087    0,08     0,94
## FrCepDenti.rec3-3 o mas veces al dia   0,3132     0,4705    0,67     0,51
##                                         
## (Intercept)                          ***
## FrCepDenti.rec2- 2 veces al dia         
## FrCepDenti.rec3-3 o mas veces al dia    
## ---
## Signif. codes:  0 '***' 0,001 '**' 0,01 '*' 0,05 '.' 0,1 ' ' 1 
## 
## (Dispersion parameter for quasibinomial family taken to be 0,956)
## 
## Number of Fisher Scoring iterations: 5
confint(Modelo_ero_4.logit)
##                                       2,5 % 97,5 %
## (Intercept)                          -3,971 -2,472
## FrCepDenti.rec2- 2 veces al dia      -1,144  1,243
## FrCepDenti.rec3-3 o mas veces al dia -0,609  1,235

levels(diseniopost1$variables$FrCepDenti.rec)
## [1] "1-1 vez al dia o menos" "2- 2 veces al dia"     
## [3] "3-3 o mas veces al dia"
diseniopost1$variables$FRCEPDENTI.rec <- recode(diseniopost1$variables$FrCepDenti.rec, 
    "'1-1 vez al dia o menos'='3 1 vez al dia o menos';'2- 2 veces al dia'='2  2 veces al dia';'3-3 o mas veces al dia'='1 3 o mas veces al dia'")

Modelo_ero_4a.logit <- svyglm(Erosinbord ~ FRCEPDENTI.rec, design = diseniopost1, 
    family = quasibinomial())
summary(Modelo_ero_4a.logit)
## 
## Call:
## svyglm(formula = Erosinbord ~ FRCEPDENTI.rec, design = diseniopost1, 
##     family = quasibinomial())
## 
## Survey design:
## postStratify(disenio1, ~categor.rec + Sexo, tabla.pob)
## 
## Coefficients:
##                                      Estimate Std. Error t value Pr(>|t|)
## (Intercept)                            -2,909      0,268  -10,84  4,8e-13
## FRCEPDENTI.rec2  2 veces al dia        -0,264      0,489   -0,54     0,59
## FRCEPDENTI.rec3 1 vez al dia o menos   -0,313      0,471   -0,67     0,51
##                                         
## (Intercept)                          ***
## FRCEPDENTI.rec2  2 veces al dia         
## FRCEPDENTI.rec3 1 vez al dia o menos    
## ---
## Signif. codes:  0 '***' 0,001 '**' 0,01 '*' 0,05 '.' 0,1 ' ' 1 
## 
## (Dispersion parameter for quasibinomial family taken to be 0,956)
## 
## Number of Fisher Scoring iterations: 5
confint(Modelo_ero_4a.logit)
##                                       2,5 %  97,5 %
## (Intercept)                          -3,435 -2,3828
## FRCEPDENTI.rec2  2 veces al dia      -1,223  0,6955
## FRCEPDENTI.rec3 1 vez al dia o menos -1,235  0,6090

table(diseniopost1$variables$FrCepDenti.rec, diseniopost1$variables$FRCEPDENTI.rec)
##                         
##                          1 3 o mas veces al dia 2  2 veces al dia
##   1-1 vez al dia o menos                      0                 0
##   2- 2 veces al dia                           0               377
##   3-3 o mas veces al dia                    496                 0
##                         
##                          3 1 vez al dia o menos
##   1-1 vez al dia o menos                    238
##   2- 2 veces al dia                           0
##   3-3 o mas veces al dia                      0

Modelo_ero_5.logit <- svyglm(Erosinbord ~ IGS.rec, design = diseniopost1, family = quasibinomial())
summary(Modelo_ero_5.logit)
## 
## Call:
## svyglm(formula = Erosinbord ~ IGS.rec, design = diseniopost1, 
##     family = quasibinomial())
## 
## Survey design:
## postStratify(disenio1, ~categor.rec + Sexo, tabla.pob)
## 
## Coefficients:
##                              Estimate Std. Error t value Pr(>|t|)    
## (Intercept)                     -4,80       1,06   -4,54  5,9e-05 ***
## IGS.rec2-De 45 a 60              1,74       1,16    1,49    0,144    
## IGS.rec3 -Menos o igual a 45     1,80       1,04    1,74    0,091 .  
## ---
## Signif. codes:  0 '***' 0,001 '**' 0,01 '*' 0,05 '.' 0,1 ' ' 1 
## 
## (Dispersion parameter for quasibinomial family taken to be 0,9818)
## 
## Number of Fisher Scoring iterations: 7
confint(Modelo_ero_5.logit)
##                                2,5 % 97,5 %
## (Intercept)                  -6,8772 -2,727
## IGS.rec2-De 45 a 60          -0,5432  4,019
## IGS.rec3 -Menos o igual a 45 -0,2307  3,833


Modelo_ero_6.logit <- svyglm(Erosinbord ~ MedResp.rec, design = diseniopost1, 
    family = quasibinomial())
summary(Modelo_ero_6.logit)
## 
## Call:
## svyglm(formula = Erosinbord ~ MedResp.rec, design = diseniopost1, 
##     family = quasibinomial())
## 
## Survey design:
## postStratify(disenio1, ~categor.rec + Sexo, tabla.pob)
## 
## Coefficients:
##                  Estimate Std. Error t value Pr(>|t|)    
## (Intercept)        -3,013      0,231  -13,05  1,3e-15 ***
## MedResp.rec2-Yes   -0,462      0,596   -0,78     0,44    
## ---
## Signif. codes:  0 '***' 0,001 '**' 0,01 '*' 0,05 '.' 0,1 ' ' 1 
## 
## (Dispersion parameter for quasibinomial family taken to be 0,9635)
## 
## Number of Fisher Scoring iterations: 6
confint(Modelo_ero_6.logit)
##                   2,5 %  97,5 %
## (Intercept)      -3,466 -2,5608
## MedResp.rec2-Yes -1,629  0,7056

Modelo_ero_7.logit <- svyglm(Erosinbord ~ AlterGastrica.rec, design = diseniopost1, 
    family = quasibinomial())
summary(Modelo_ero_7.logit)
## 
## Call:
## svyglm(formula = Erosinbord ~ AlterGastrica.rec, design = diseniopost1, 
##     family = quasibinomial())
## 
## Survey design:
## postStratify(disenio1, ~categor.rec + Sexo, tabla.pob)
## 
## Coefficients:
##                        Estimate Std. Error t value Pr(>|t|)    
## (Intercept)              -3,012      0,225   -13,4  6,2e-16 ***
## AlterGastrica.rec2-Yes  -15,680      0,212   -74,0  < 2e-16 ***
## ---
## Signif. codes:  0 '***' 0,001 '**' 0,01 '*' 0,05 '.' 0,1 ' ' 1 
## 
## (Dispersion parameter for quasibinomial family taken to be 0,9127)
## 
## Number of Fisher Scoring iterations: 17
confint(Modelo_ero_7.logit)
##                          2,5 % 97,5 %
## (Intercept)             -3,454  -2,57
## AlterGastrica.rec2-Yes -16,095 -15,26


Modelo_ero_8.logit <- svyglm(Erosinbord ~ RefrCola., design = diseniopost1, 
    family = quasibinomial())
summary(Modelo_ero_8.logit)
## 
## Call:
## svyglm(formula = Erosinbord ~ RefrCola., design = diseniopost1, 
##     family = quasibinomial())
## 
## Survey design:
## postStratify(disenio1, ~categor.rec + Sexo, tabla.pob)
## 
## Coefficients:
##             Estimate Std. Error t value Pr(>|t|)    
## (Intercept) -3,15275    0,25087  -12,57  4,2e-15 ***
## RefrCola.    0,01729    0,00971    1,78    0,083 .  
## ---
## Signif. codes:  0 '***' 0,001 '**' 0,01 '*' 0,05 '.' 0,1 ' ' 1 
## 
## (Dispersion parameter for quasibinomial family taken to be 0,9802)
## 
## Number of Fisher Scoring iterations: 5
confint(Modelo_ero_8.logit)
##                 2,5 %   97,5 %
## (Intercept) -3,644455 -2,66105
## RefrCola.   -0,001746  0,03632

Modelo_ero_9.logit <- svyglm(Erosinbord ~ Consitencia_Cepillo.rec, design = diseniopost1, 
    family = quasibinomial())
summary(Modelo_ero_9.logit)
## 
## Call:
## svyglm(formula = Erosinbord ~ Consitencia_Cepillo.rec, design = diseniopost1, 
##     family = quasibinomial())
## 
## Survey design:
## postStratify(disenio1, ~categor.rec + Sexo, tabla.pob)
## 
## Coefficients:
##                                 Estimate Std. Error t value Pr(>|t|)    
## (Intercept)                      -2,4473     0,2857   -8,57  8,7e-10 ***
## Consitencia_Cepillo.rec2-Medium  -1,0137     0,2924   -3,47   0,0015 ** 
## Consitencia_Cepillo.rec3-Hard    -0,0785     0,7204   -0,11   0,9139    
## ---
## Signif. codes:  0 '***' 0,001 '**' 0,01 '*' 0,05 '.' 0,1 ' ' 1 
## 
## (Dispersion parameter for quasibinomial family taken to be 0,7872)
## 
## Number of Fisher Scoring iterations: 6
confint(Modelo_ero_9.logit)
##                                  2,5 %  97,5 %
## (Intercept)                     -3,007 -1,8873
## Consitencia_Cepillo.rec2-Medium -1,587 -0,4406
## Consitencia_Cepillo.rec3-Hard   -1,490  1,3335

Modelo_ero_10.logit <- svyglm(Erosinbord ~ Natac2vec.rec, design = diseniopost1, 
    family = quasibinomial())
summary(Modelo_ero_10.logit)
## 
## Call:
## svyglm(formula = Erosinbord ~ Natac2vec.rec, design = diseniopost1, 
##     family = quasibinomial())
## 
## Survey design:
## postStratify(disenio1, ~categor.rec + Sexo, tabla.pob)
## 
## Coefficients:
##                    Estimate Std. Error t value Pr(>|t|)    
## (Intercept)         -3,0415     0,2868  -10,61  3,6e-12 ***
## Natac2vec.rec2-Yes   0,0789     0,4669    0,17     0,87    
## ---
## Signif. codes:  0 '***' 0,001 '**' 0,01 '*' 0,05 '.' 0,1 ' ' 1 
## 
## (Dispersion parameter for quasibinomial family taken to be 0,8127)
## 
## Number of Fisher Scoring iterations: 5
confint(Modelo_ero_10.logit)
##                      2,5 %  97,5 %
## (Intercept)        -3,6035 -2,4794
## Natac2vec.rec2-Yes -0,8361  0,9939

Modelo_ero_11.logit <- svyglm(Erosinbord ~ BuchTragar.rec, design = diseniopost1, 
    family = quasibinomial())
summary(Modelo_ero_11.logit)
## 
## Call:
## svyglm(formula = Erosinbord ~ BuchTragar.rec, design = diseniopost1, 
##     family = quasibinomial())
## 
## Survey design:
## postStratify(disenio1, ~categor.rec + Sexo, tabla.pob)
## 
## Coefficients:
##                     Estimate Std. Error t value Pr(>|t|)    
## (Intercept)           -3,134      0,294  -10,66  3,2e-12 ***
## BuchTragar.rec2-Yes    0,770      0,397    1,94    0,061 .  
## ---
## Signif. codes:  0 '***' 0,001 '**' 0,01 '*' 0,05 '.' 0,1 ' ' 1 
## 
## (Dispersion parameter for quasibinomial family taken to be 0,7982)
## 
## Number of Fisher Scoring iterations: 5
confint(Modelo_ero_11.logit)
##                         2,5 % 97,5 %
## (Intercept)         -3,710152 -2,557
## BuchTragar.rec2-Yes -0,007202  1,547

Modelo_ero_12.logit <- svyglm(Erosinbord ~ FormBeber.rec, design = diseniopost1, 
    family = quasibinomial())
summary(Modelo_ero_12.logit)
## 
## Call:
## svyglm(formula = Erosinbord ~ FormBeber.rec, design = diseniopost1, 
##     family = quasibinomial())
## 
## Survey design:
## postStratify(disenio1, ~categor.rec + Sexo, tabla.pob)
## 
## Coefficients:
##                            Estimate Std. Error t value Pr(>|t|)    
## (Intercept)                 -3,0562     0,6398   -4,78  2,8e-05 ***
## FormBeber.rec2-Por el pico  -0,7785     0,7794   -1,00     0,32    
## FormBeber.rec3-Con vaso      0,0866     0,7446    0,12     0,91    
## ---
## Signif. codes:  0 '***' 0,001 '**' 0,01 '*' 0,05 '.' 0,1 ' ' 1 
## 
## (Dispersion parameter for quasibinomial family taken to be 0,9754)
## 
## Number of Fisher Scoring iterations: 6
confint(Modelo_ero_12.logit)
##                             2,5 %  97,5 %
## (Intercept)                -4,310 -1,8022
## FormBeber.rec2-Por el pico -2,306  0,7491
## FormBeber.rec3-Con vaso    -1,373  1,5459


Modelo_ero_13.logit <- svyglm(Erosinbord ~ Nivel.Socieconomico.rec, design = diseniopost1, 
    family = quasibinomial())
summary(Modelo_ero_13.logit)
## 
## Call:
## svyglm(formula = Erosinbord ~ Nivel.Socieconomico.rec, design = diseniopost1, 
##     family = quasibinomial())
## 
## Survey design:
## postStratify(disenio1, ~categor.rec + Sexo, tabla.pob)
## 
## Coefficients:
##                                Estimate Std. Error t value Pr(>|t|)    
## (Intercept)                     -3,1878     0,3146  -10,13  3,2e-12 ***
## Nivel.Socieconomico.rec2-MEDIO   0,0235     0,4332    0,05     0,96    
## Nivel.Socieconomico.rec3-ALTO    0,3825     0,4219    0,91     0,37    
## ---
## Signif. codes:  0 '***' 0,001 '**' 0,01 '*' 0,05 '.' 0,1 ' ' 1 
## 
## (Dispersion parameter for quasibinomial family taken to be 0,9713)
## 
## Number of Fisher Scoring iterations: 5
confint(Modelo_ero_13.logit)
##                                  2,5 %  97,5 %
## (Intercept)                    -3,8044 -2,5712
## Nivel.Socieconomico.rec2-MEDIO -0,8254  0,8725
## Nivel.Socieconomico.rec3-ALTO  -0,4445  1,2095

Modelo_ero_14.logit <- svyglm(Erosinbord ~ Tipo.de.Escuela.rec, design = diseniopost1, 
    family = quasibinomial())
summary(Modelo_ero_14.logit)
## 
## Call:
## svyglm(formula = Erosinbord ~ Tipo.de.Escuela.rec, design = diseniopost1, 
##     family = quasibinomial())
## 
## Survey design:
## postStratify(disenio1, ~categor.rec + Sexo, tabla.pob)
## 
## Coefficients:
##                              Estimate Std. Error t value Pr(>|t|)    
## (Intercept)                    -3,734      0,599   -6,24  2,7e-07 ***
## Tipo.de.Escuela.rec2-Private    0,819      0,647    1,26     0,21    
## ---
## Signif. codes:  0 '***' 0,001 '**' 0,01 '*' 0,05 '.' 0,1 ' ' 1 
## 
## (Dispersion parameter for quasibinomial family taken to be 0,9818)
## 
## Number of Fisher Scoring iterations: 6
confint(Modelo_ero_14.logit)
##                                2,5 % 97,5 %
## (Intercept)                  -4,9072 -2,560
## Tipo.de.Escuela.rec2-Private -0,4502  2,087

levels(diseniopost1$variables$Yogurt.rec)
## [1] "1-Mas de 3 veces al dia" "2-Todos los dias"       
## [3] "3-Nunca o raramente"

Modelo_ero_15.logit <- svyglm(Erosinbord ~ Yogurt.rec, design = diseniopost1, 
    family = quasibinomial())
summary(Modelo_ero_15.logit)
## 
## Call:
## svyglm(formula = Erosinbord ~ Yogurt.rec, design = diseniopost1, 
##     family = quasibinomial())
## 
## Survey design:
## postStratify(disenio1, ~categor.rec + Sexo, tabla.pob)
## 
## Coefficients:
##                               Estimate Std. Error t value Pr(>|t|)   
## (Intercept)                     -2,177      0,614   -3,55   0,0011 **
## Yogurt.rec2-Todos los dias      -0,985      0,622   -1,58   0,1220   
## Yogurt.rec3-Nunca o raramente   -0,942      0,490   -1,92   0,0624 . 
## ---
## Signif. codes:  0 '***' 0,001 '**' 0,01 '*' 0,05 '.' 0,1 ' ' 1 
## 
## (Dispersion parameter for quasibinomial family taken to be 0,9633)
## 
## Number of Fisher Scoring iterations: 5
confint(Modelo_ero_15.logit)
##                                2,5 %   97,5 %
## (Intercept)                   -3,380 -0,97338
## Yogurt.rec2-Todos los dias    -2,204  0,23472
## Yogurt.rec3-Nunca o raramente -1,903  0,01886


Modelo_ero_16.logit <- svyglm(Erosinbord ~ Bruxismo.rec, design = diseniopost1, 
    family = quasibinomial())
summary(Modelo_ero_16.logit)
## 
## Call:
## svyglm(formula = Erosinbord ~ Bruxismo.rec, design = diseniopost1, 
##     family = quasibinomial())
## 
## Survey design:
## postStratify(disenio1, ~categor.rec + Sexo, tabla.pob)
## 
## Coefficients:
##                   Estimate Std. Error t value Pr(>|t|)    
## (Intercept)         -3,057      0,247   -12,4  6,8e-15 ***
## Bruxismo.rec2-Yes   -0,085      0,427    -0,2     0,84    
## ---
## Signif. codes:  0 '***' 0,001 '**' 0,01 '*' 0,05 '.' 0,1 ' ' 1 
## 
## (Dispersion parameter for quasibinomial family taken to be 0,964)
## 
## Number of Fisher Scoring iterations: 5
confint(Modelo_ero_16.logit)
##                     2,5 %  97,5 %
## (Intercept)       -3,5415 -2,5725
## Bruxismo.rec2-Yes -0,9215  0,7515

Modelo_ero_17.logit <- svyglm(Erosinbord ~ bedeportediario.rec, design = diseniopost1, 
    family = quasibinomial())
summary(Modelo_ero_17.logit)
## 
## Call:
## svyglm(formula = Erosinbord ~ bedeportediario.rec, design = diseniopost1, 
##     family = quasibinomial())
## 
## Survey design:
## postStratify(disenio1, ~categor.rec + Sexo, tabla.pob)
## 
## Coefficients:
##                               Estimate Std. Error t value Pr(>|t|)    
## (Intercept)                     -3,121      0,334   -9,34  1,2e-10 ***
## bedeportediario.rec2-Gatorade    0,892      0,864    1,03     0,31    
## bedeportediario.rec3-other      -1,354      1,114   -1,22     0,23    
## ---
## Signif. codes:  0 '***' 0,001 '**' 0,01 '*' 0,05 '.' 0,1 ' ' 1 
## 
## (Dispersion parameter for quasibinomial family taken to be 0,8143)
## 
## Number of Fisher Scoring iterations: 6
confint(Modelo_ero_17.logit)
##                                 2,5 %  97,5 %
## (Intercept)                   -3,7755 -2,4656
## bedeportediario.rec2-Gatorade -0,8013  2,5845
## bedeportediario.rec3-other    -3,5365  0,8284

Modelo_ero_19.logit <- svyglm(Erosinbord ~ jugodiario.nodiario, design = diseniopost1, 
    family = quasibinomial())
summary(Modelo_ero_19.logit)
## 
## Call:
## svyglm(formula = Erosinbord ~ jugodiario.nodiario, design = diseniopost1, 
##     family = quasibinomial())
## 
## Survey design:
## postStratify(disenio1, ~categor.rec + Sexo, tabla.pob)
## 
## Coefficients:
##                             Estimate Std. Error t value Pr(>|t|)    
## (Intercept)                   -3,195      0,347   -9,20  3,3e-11 ***
## jugodiario.nodiario2-Diario    0,180      0,504    0,36     0,72    
## ---
## Signif. codes:  0 '***' 0,001 '**' 0,01 '*' 0,05 '.' 0,1 ' ' 1 
## 
## (Dispersion parameter for quasibinomial family taken to be 0,9706)
## 
## Number of Fisher Scoring iterations: 5
confint(Modelo_ero_19.logit)
##                               2,5 % 97,5 %
## (Intercept)                 -3,8755 -2,515
## jugodiario.nodiario2-Diario -0,8085  1,168



Modelo_ero_20.logit <- svyglm(Erosinbord ~ JugFrutas.rec, design = diseniopost1, 
    family = quasibinomial())
summary(Modelo_ero_20.logit)
## 
## Call:
## svyglm(formula = Erosinbord ~ JugFrutas.rec, design = diseniopost1, 
##     family = quasibinomial())
## 
## Survey design:
## postStratify(disenio1, ~categor.rec + Sexo, tabla.pob)
## 
## Coefficients:
##                                      Estimate Std. Error t value Pr(>|t|)
## (Intercept)                            -3,717      0,955   -3,89   0,0004
## JugFrutas.rec2-Todos los dias           0,489      0,957    0,51   0,6120
## JugFrutas.rec3-Mas de 3 veces al dia    1,169      1,030    1,14   0,2637
##                                         
## (Intercept)                          ***
## JugFrutas.rec2-Todos los dias           
## JugFrutas.rec3-Mas de 3 veces al dia    
## ---
## Signif. codes:  0 '***' 0,001 '**' 0,01 '*' 0,05 '.' 0,1 ' ' 1 
## 
## (Dispersion parameter for quasibinomial family taken to be 0,9706)
## 
## Number of Fisher Scoring iterations: 6
confint(Modelo_ero_20.logit)
##                                        2,5 % 97,5 %
## (Intercept)                          -5,5887 -1,845
## JugFrutas.rec2-Todos los dias        -1,3857  2,365
## JugFrutas.rec3-Mas de 3 veces al dia -0,8498  3,188




Modelo_ero_21.logit <- svyglm(Erosinbord ~ bedeportediario.rec, design = diseniopost1, 
    family = quasibinomial())
summary(Modelo_ero_21.logit)
## 
## Call:
## svyglm(formula = Erosinbord ~ bedeportediario.rec, design = diseniopost1, 
##     family = quasibinomial())
## 
## Survey design:
## postStratify(disenio1, ~categor.rec + Sexo, tabla.pob)
## 
## Coefficients:
##                               Estimate Std. Error t value Pr(>|t|)    
## (Intercept)                     -3,121      0,334   -9,34  1,2e-10 ***
## bedeportediario.rec2-Gatorade    0,892      0,864    1,03     0,31    
## bedeportediario.rec3-other      -1,354      1,114   -1,22     0,23    
## ---
## Signif. codes:  0 '***' 0,001 '**' 0,01 '*' 0,05 '.' 0,1 ' ' 1 
## 
## (Dispersion parameter for quasibinomial family taken to be 0,8143)
## 
## Number of Fisher Scoring iterations: 6
confint(Modelo_ero_21.logit)
##                                 2,5 %  97,5 %
## (Intercept)                   -3,7755 -2,4656
## bedeportediario.rec2-Gatorade -0,8013  2,5845
## bedeportediario.rec3-other    -3,5365  0,8284

Modelo_ero_22.logit <- svyglm(Erosinbord ~ Bebidas_energizantes.rec, design = diseniopost1, 
    family = quasibinomial())
summary(Modelo_ero_22.logit)
## 
## Call:
## svyglm(formula = Erosinbord ~ Bebidas_energizantes.rec, design = diseniopost1, 
##     family = quasibinomial())
## 
## Survey design:
## postStratify(disenio1, ~categor.rec + Sexo, tabla.pob)
## 
## Coefficients:
##                                               Estimate Std. Error t value
## (Intercept)                                     -3,342      0,302  -11,08
## Bebidas_energizantes.rec2-Todos los dias         0,277      0,286    0,97
## Bebidas_energizantes.recMas de 3 veces al dia    2,052      0,670    3,06
##                                               Pr(>|t|)    
## (Intercept)                                    2,6e-13 ***
## Bebidas_energizantes.rec2-Todos los dias        0,3396    
## Bebidas_energizantes.recMas de 3 veces al dia   0,0041 ** 
## ---
## Signif. codes:  0 '***' 0,001 '**' 0,01 '*' 0,05 '.' 0,1 ' ' 1 
## 
## (Dispersion parameter for quasibinomial family taken to be 0,9451)
## 
## Number of Fisher Scoring iterations: 6
confint(Modelo_ero_22.logit)
##                                                 2,5 %  97,5 %
## (Intercept)                                   -3,9339 -2,7509
## Bebidas_energizantes.rec2-Todos los dias      -0,2843  0,8386
## Bebidas_energizantes.recMas de 3 veces al dia  0,7381  3,3659


Modelo_ero_23.logit <- svyglm(Erosinbord ~ RefrCola.rec, design = diseniopost1, 
    family = quasibinomial())
summary(Modelo_ero_23.logit)
## 
## Call:
## svyglm(formula = Erosinbord ~ RefrCola.rec, design = diseniopost1, 
##     family = quasibinomial())
## 
## Survey design:
## postStratify(disenio1, ~categor.rec + Sexo, tabla.pob)
## 
## Coefficients:
##                                     Estimate Std. Error t value Pr(>|t|)
## (Intercept)                           -4,109      0,753   -5,45  3,4e-06
## RefrCola.rec2-Todos los dias           0,908      0,740    1,23     0,23
## RefrCola.rec3-Mas de 3 veces al dia    1,451      0,883    1,64     0,11
##                                        
## (Intercept)                         ***
## RefrCola.rec2-Todos los dias           
## RefrCola.rec3-Mas de 3 veces al dia    
## ---
## Signif. codes:  0 '***' 0,001 '**' 0,01 '*' 0,05 '.' 0,1 ' ' 1 
## 
## (Dispersion parameter for quasibinomial family taken to be 0,9703)
## 
## Number of Fisher Scoring iterations: 6
confint(Modelo_ero_23.logit)
##                                       2,5 % 97,5 %
## (Intercept)                         -5,5849 -2,632
## RefrCola.rec2-Todos los dias        -0,5413  2,358
## RefrCola.rec3-Mas de 3 veces al dia -0,2790  3,182