Étude Cas-témoin réalisée chez 449 femmes: Analyse et interprétation

library(rmdformats)
library(readxl)
library(tidyverse)

library(gtsummary)
library(gt)
library(epitools)
library(ggstats)
library(mice)

library(labelled)
library(questionr)

INFARCTUS_M <- read_excel("D:/ENATSE 2/Logiciel R/Analyse_infarctus_26/Infarct (1).xls")

INFARCTUS_M[INFARCTUS_M =="."]<-NA

View(INFARCTUS_M)

str(INFARCTUS_M)

INFARCTUS_M$INFARCT <- as.character(INFARCTUS_M$INFARCT)
INFARCTUS_M$INFARCT[INFARCTUS_M$INFARCT == "0"] <- "Témoins"
INFARCTUS_M$INFARCT[INFARCTUS_M$INFARCT == "1"] <- "Cas"

INFARCTUS_M$CO <- as.character(INFARCTUS_M$CO)
INFARCTUS_M$CO[INFARCTUS_M$CO == "0"] <- "Jamais"
INFARCTUS_M$CO[INFARCTUS_M$CO == "1"] <- "Oui"

INFARCTUS_M$TABAC <- as.character(INFARCTUS_M$TABAC)
INFARCTUS_M$TABAC[INFARCTUS_M$TABAC == "0"] <- "Non"
INFARCTUS_M$TABAC[INFARCTUS_M$TABAC == "1"] <- "Fumeuse actuelle"
INFARCTUS_M$TABAC[INFARCTUS_M$TABAC == "2"] <- "Ancienne fumeuse"

INFARCTUS_M$HTA <- as.character(INFARCTUS_M$HTA)
INFARCTUS_M$HTA[INFARCTUS_M$HTA == "0"] <- "Non"
INFARCTUS_M$HTA[INFARCTUS_M$HTA == "1"] <- "Oui"

INFARCTUS_M$ATCD <- INFARCTUS_M$ATCD |>
  fct_recode(
    "Non" = "0",
    "Oui" = "1"
  )

INFARCTUS_M$POIDS <- as.numeric(INFARCTUS_M$POIDS)

INFARCTUS_M$IMC <- as.numeric(INFARCTUS_M$IMC)

INFARCTUS_M$INFARCT <- INFARCTUS_M$INFARCT |>
  fct_relevel(
    "Témoins", "Cas"
  )

INFARCTUS_M$TABAC <- INFARCTUS_M$TABAC |>
  fct_relevel(
    "Non", "Ancienne fumeuse", "Fumeuse actuelle"
  )

INFARCTUS_M = INFARCTUS_M %>% mutate(across(where(is.character),as.factor))

str(INFARCTUS_M)

summary(INFARCTUS_M)

INFARCT_imp=mice(INFARCTUS_M,method = c("","","","","","norm","", "norm","logreg",""),m=5)  

INFARCTUS_M=complete(INFARCT_imp)

## Normalité de l'âge
shapiro.test(INFARCTUS_M$AGE)

qqnorm(INFARCTUS_M$AGE)
qqline(INFARCTUS_M$AGE)

## Normalité du poids des femmes
shapiro.test(INFARCTUS_M$POIDS)

qqnorm(INFARCTUS_M$POIDS)
qqline(INFARCTUS_M$POIDS)

## Normalité de la taille des femmes
shapiro.test(INFARCTUS_M$TAILLE)

qqnorm(INFARCTUS_M$TAILLE)
qqline(INFARCTUS_M$TAILLE)

## Normalité de l'IMC
shapiro.test(INFARCTUS_M$IMC)

qqnorm(INFARCTUS_M$IMC)
qqline(INFARCTUS_M$IMC)

R=theme_gtsummary_language(language ="fr",decimal.mark = ",",big.mark = " ")

TBL <- tbl_summary(INFARCTUS_M,include = -NUMERO,
                   statistic = list(all_categorical() ~ "{n} ({p})"),                 
                   digits=list(
                     all_categorical()~ c(0,1)
                     
                   )) %>%
  modify_header(label="**Variables**")%>%
  modify_caption("**Tableau 1:** Description des variables en étude ") %>% 
  bold_labels()%>%
  as_flex_table()
  
TBL

TBL1 <- tbl_summary(INFARCTUS_M,include = c(INFARCT,CO,TABAC,ATCD,
                                           HTA),
                   statistic = list(all_categorical() ~ "{n} ({p})"),                 
                   digits=list(
                     all_categorical()~ c(0,1)
                     
                   )) %>%
  modify_header(label="**Variables**")%>%
  bold_labels()%>%
  modify_caption("**Tableau 2:** Répartition des variables qualitatives")

tbl_summary(INFARCTUS_M,include = -NUMERO,
statistic = list(all_categorical() ~ "{n} ({p})"
                 ),
digits=list(
  all_categorical()~ c(0,1),
  all_continuous() ~ 1
),by=INFARCT)%>%
  add_overall(last=T,col_label= "**Total** N= {N}")%>% 
  add_p()%>%
  bold_p()%>%
  bold_labels() %>% 
  modify_header(
    list(
      label ~ "**Variables**",
      all_stat_cols() ~ "**_{level}_** (n={n}, {style_percent(p)}%)",
      stat_0 ~ "**TOTAL** (N={N})",p.value~"**Test de comparaison** (p-valeur)"
    )
  ) %>%
  modify_footnote(everything() ~ NA) %>%
  modify_spanning_header(all_stat_cols() ~ "**Infarctus du myocarde**")%>%
  modify_caption("Tableau 3: Facteurs associés à l'infarctus de myocarde chez les femmes (analyse bivariée)" 
  
)%>%
  as_flex_table()

## Le modèle 

mod = glm(
  INFARCT ~ CO+AGE+TABAC+HTA+POIDS+ATCD+TAILLE,
  family = binomial,data = INFARCTUS_M
)

mod

tbl_regression(mod, 
               exponentiate = T,
               add_estimate_to_reference_rows = TRUE
               )%>%
  bold_labels() %>%
  bold_p()%>%
  italicize_levels()%>%
  modify_header(label="**Variables**",p.value= "**p**")%>%
  modify_footnote(everything() ~ NA)%>%
  remove_abbreviation(abbreviation = NULL)%>%
  modify_caption("Tableau 4: Facteurs associés à l'infarctus de myocarde chez les femmes (analyse multivariée)" )%>%
  as_flex_table()

ggcoef_model(mod,exponentiate = T)

ggcoef_table(mod,exponentiate = T)

# Création de la variable classe d'âge
INFARCTUS_M <- mutate(INFARCTUS_M,
                       CLASSE_AGE= 
                         case_when(AGE >=  15  &   AGE <= 49   ~  "15-49",
                                   AGE >=  50  &   AGE <= 100  ~  "50-100"
                          )
                      )

### Transformation de la nature de la variable CLASSE_AGE en facteur et labellisation

INFARCTUS_M$CLASSE_AGE = as.factor(INFARCTUS_M$CLASSE_AGE)

var_label(INFARCTUS_M) <- list(
  NUMERO       =          "Numéro",
  INFARCT      =          "Infarctus du myocarde",
  CO           =          "Prise de contraceptifs oraux",
  TABAC        =          "Consommation de tabac",
  AGE          =          "Age (en année)",
  POIDS        =          "Poids (Kg)",
  TAILLE       =          "Taille (Cm)",
  ATCD         =          "Antécédents familiaux de maladie cardio-vasculaire",
  HTA          =          "Hypertension artérielle",
  IMC          =          "Indice de masse corporelle",
  CLASSE_AGE   =          "Classe d'âge")

tbl_summary(INFARCTUS_M,include = -c(NUMERO,AGE),
            statistic = list(all_categorical() ~ "{n} ({p})"
            ),
            digits=list(
              all_categorical()~ c(0,1),
              all_continuous() ~ 1
            ),by=INFARCT)%>%
  add_overall(last=T,col_label= "**Total** N= {N}")%>% 
  add_p()%>%
  bold_p()%>%
  bold_labels() %>% 
  modify_header(
    list(
      label ~ "**Variables**",
      all_stat_cols() ~ "**_{level}_** (n={n}, {style_percent(p)}%)",
      stat_0 ~ "**TOTAL** (N={N})",p.value~"**Test de comparaison** (p-valeur)"
    )
  ) %>%
  modify_footnote(everything() ~ NA) %>%
  modify_spanning_header(all_stat_cols() ~ "**Infarctus du myocarde**")%>%
  modify_caption("Tableau 5: Facteurs associés à l'infarctus de myocarde chez les femmes (analyse bivariée)" 
                 
  )%>%
  as_flex_table()

tbl_uvregression(
  data = INFARCTUS_M,
  include = -c(NUMERO,AGE),
  y = INFARCT,
  method = glm,
  method.args = list(family = binomial),
  exponentiate = TRUE,
  add_estimate_to_reference_rows = TRUE
) %>%
  add_global_p() %>% 
  bold_p() %>% 
  modify_header(label="**Variables**",p.value= "**p**") %>% 
  as_flex_table()

## CLASSE_AGE - POIDS

ggplot(INFARCTUS_M) +
 aes(x = CLASSE_AGE, y = POIDS, fill = CLASSE_AGE) +
 geom_boxplot() +
 stat_summary(fun = mean, geom = "point", shape = 20, size = 4, color = "red")+
 scale_fill_hue(direction = 1) +
 labs(x = "Classe d'âge ", y = "Poids (Kg)", title = "Distibution du poids des femmes selon la classe d'âge") +
 theme_classic() +
 theme(plot.title = element_text(size = 16L, face = "bold", hjust = 0.5), axis.title.y = element_text(size = 13L, 
 face = "bold.italic", hjust = 1), axis.title.x = element_text(size = 13L, face = "bold.italic", hjust = 1), 
 axis.text.y = element_text(size = 14L), axis.text.x = element_text(size = 14L))

## CLASSE_AGE - TAILLE

ggplot(INFARCTUS_M) +
  aes(x = CLASSE_AGE, y = TAILLE, fill = CLASSE_AGE) +
  geom_boxplot() +
  stat_summary(fun = mean, geom = "point", shape = 20, size = 4, color = "red")+
  scale_fill_hue(direction = 1) +
  labs(x = "Classe d'âge ", y = "Taille (Cm)", title = "Distibution de la taille des femmes selon la classe d'âge") +
  theme_classic() +
  theme(plot.title = element_text(size = 16L, face = "bold", hjust = 0.5), axis.title.y = element_text(size = 13L, 
                                                                                                       face = "bold.italic", hjust = 1), axis.title.x = element_text(size = 13L, face = "bold.italic", hjust = 1), 
        axis.text.y = element_text(size = 14L), axis.text.x = element_text(size = 14L))

mod1 = glm(
  INFARCT ~ CO+TABAC+HTA+POIDS+ATCD+TAILLE+CLASSE_AGE,
  family = binomial,data = INFARCTUS_M
)

mod1

tbl_regression(mod1, 
               exponentiate = T,
               add_estimate_to_reference_rows = TRUE
)%>%
  bold_labels() %>%
  bold_p()%>%
  italicize_levels()%>%
  modify_header(label="**Variables**",p.value= "**p**")%>%
  modify_footnote(everything() ~ NA)%>%
  remove_abbreviation(abbreviation = NULL)%>%
  modify_caption("**Tableau 6: Facteurs associés à l'infarctus de myocarde chez les femmes (analyse multivariée)**" )%>%
  as_flex_table()

### Graphiques

ggcoef_model(mod1,exponentiate = T)

ggcoef_table(mod1,exponentiate = T)