Análisis Epidemiológico Nutricional
Laura Quintero
2026-06-21
library(tidyverse)
library(readxl)
library(ggplot2)
library(broom)
library(flextable)
library(knitr)
library(kableExtra)base <- read_excel("Base_epidemiologia_completa.xlsx")
base <- base %>%
rename(
sexo = Sex,
edad_anos = `Age (years)`,
edad_meses = `Age (months)`,
peso = `Weight (kg)`,
talla = `Height (cm)`,
grupo_edad = Grupo_edad,
hemoglobina = Hemoglobina,
hematocrito = Hematocrito,
ingesta_hierro = Ingesta_hierro
) %>%
mutate(
sexo = factor(sexo, levels = c("Male","Female"),
labels = c("Masculino","Femenino")),
grupo_edad = factor(grupo_edad, levels = c("<5 años","≥5 años")),
IMC = round(peso / (talla/100)^2, 2),
Clasificacion_HAZ = factor(case_when(
HAZ < -3 ~ "Talla baja severa",
HAZ >= -3 & HAZ < -2 ~ "Talla baja para la edad o retraso en talla",
HAZ >= -2 & HAZ < -1 ~ "Riesgo de talla baja",
HAZ >= -1 ~ "Talla adecuada para la edad",
TRUE ~ NA_character_),
levels = c("Talla baja severa",
"Talla baja para la edad o retraso en talla",
"Riesgo de talla baja",
"Talla adecuada para la edad")),
Clasificacion_BAZ = factor(case_when(
BAZ < -2 ~ "Delgadez",
BAZ >= -2 & BAZ < -1 ~ "Riesgo de delgadez",
BAZ >= -1 & BAZ <= 1 ~ "IMC adecuado para la edad",
BAZ > 1 & BAZ <= 2 ~ "Sobrepeso",
BAZ > 2 ~ "Obesidad",
TRUE ~ NA_character_),
levels = c("Delgadez","Riesgo de delgadez",
"IMC adecuado para la edad","Sobrepeso","Obesidad")),
WAZ_valido = if_else(grupo_edad == "<5 años", WAZ, NA_real_),
Clasificacion_WAZ = factor(case_when(
WAZ_valido < -2 ~ "Bajo peso para la edad",
WAZ_valido >= -2 ~ "Peso adecuado para la edad",
TRUE ~ NA_character_),
levels = c("Bajo peso para la edad","Peso adecuado para la edad"))
)
menores5 <- base %>% filter(grupo_edad == "<5 años")
mayores5 <- base %>% filter(grupo_edad == "≥5 años")
colores_grupo <- c("<5 años" = "#4472C4", "≥5 años" = "#70AD47")
base <- base %>%
mutate(
WHZ_valido = if_else(grupo_edad == "<5 años", WHZ, NA_real_),
Clasificacion_WHZ = factor(
case_when(
WHZ_valido < -3 ~ "Emaciación severa",
WHZ_valido >= -3 & WHZ_valido < -2 ~ "Emaciación",
WHZ_valido >= -2 & WHZ_valido <= 1 ~ "Normal",
WHZ_valido > 1 & WHZ_valido <= 2 ~ "Riesgo de sobrepeso",
WHZ_valido > 2 & WHZ_valido <= 3 ~ "Sobrepeso",
WHZ_valido > 3 ~ "Obesidad",
TRUE ~ NA_character_
),
levels = c("Emaciación severa","Emaciación","Normal",
"Riesgo de sobrepeso","Sobrepeso","Obesidad")
)
)1 Análisis descriptivo
1.1 Variables sociodemográficas
# Sexo
base %>%
count(sexo) %>%
mutate(Porcentaje = round(n/sum(n)*100, 2)) %>%
kable(col.names = c("Sexo","n","%"), caption = "Distribución por sexo") %>%
kable_styling(bootstrap_options = c("striped","hover","condensed"),
full_width = FALSE)| Sexo | n | % |
|---|---|---|
| Masculino | 43 | 47.78 |
| Femenino | 47 | 52.22 |
# Grupo de edad
base %>%
count(grupo_edad) %>%
mutate(Porcentaje = round(n/sum(n)*100, 2)) %>%
kable(col.names = c("Grupo de edad","n","%"),
caption = "Distribución por grupo de edad") %>%
kable_styling(bootstrap_options = c("striped","hover","condensed"),
full_width = FALSE)| Grupo de edad | n | % |
|---|---|---|
| <5 años | 45 | 50 |
| ≥5 años | 45 | 50 |
base %>%
count(grupo_edad, sexo) %>%
ggplot(aes(x = grupo_edad, y = n, fill = sexo)) +
geom_col(position = "dodge", color = "white", width = 0.6) +
geom_text(aes(label = n), position = position_dodge(0.6),
vjust = -0.4, fontface = "bold", size = 4) +
scale_fill_manual(values = c("Masculino" = "#4472C4",
"Femenino" = "#ED7D31")) +
labs(title = "Distribución por sexo y grupo de edad",
x = "Grupo de edad", y = "n", fill = "Sexo") +
theme_bw(base_size = 12) +
theme(legend.position = "bottom")
Distribución por sexo y grupo de edad
1.2 Variables cuantitativas
vars_desc <- base %>%
select(edad_anos, peso, talla, IMC,
HAZ, BAZ, hemoglobina, hematocrito, ingesta_hierro)
data.frame(
Variable = c("Edad (años)","Peso (kg)","Talla (cm)","IMC (kg/m²)",
"HAZ","BAZ","Hemoglobina (g/dL)","Hematocrito (%)","Ingesta hierro (mg/día)"),
n = sapply(vars_desc, function(x) sum(!is.na(x))),
Media = sapply(vars_desc, mean, na.rm = TRUE) %>% round(2),
DE = sapply(vars_desc, sd, na.rm = TRUE) %>% round(2),
Mediana = sapply(vars_desc, median, na.rm = TRUE) %>% round(2),
Minimo = sapply(vars_desc, min, na.rm = TRUE) %>% round(2),
Maximo = sapply(vars_desc, max, na.rm = TRUE) %>% round(2)
) %>%
kable(caption = "Estadística descriptiva — total muestra") %>%
kable_styling(bootstrap_options = c("striped","hover","condensed"))| Variable | n | Media | DE | Mediana | Minimo | Maximo | |
|---|---|---|---|---|---|---|---|
| edad_anos | Edad (años) | 90 | 7.99 | 6.41 | 5.00 | 0.00 | 18.75 |
| peso | Peso (kg) | 90 | 29.25 | 19.66 | 19.45 | 3.20 | 71.10 |
| talla | Talla (cm) | 90 | 116.81 | 42.15 | 104.65 | 48.00 | 181.40 |
| IMC | IMC (kg/m²) | 90 | 19.28 | 3.35 | 19.02 | 13.28 | 28.41 |
| HAZ | HAZ | 90 | -1.22 | 2.03 | -0.86 | -6.07 | 2.58 |
| BAZ | BAZ | 90 | 1.12 | 2.18 | 0.70 | -3.25 | 7.32 |
| hemoglobina | Hemoglobina (g/dL) | 88 | 12.30 | 1.27 | 12.05 | 10.10 | 16.10 |
| hematocrito | Hematocrito (%) | 88 | 35.49 | 4.19 | 34.75 | 26.40 | 49.50 |
| ingesta_hierro | Ingesta hierro (mg/día) | 89 | 21.53 | 13.17 | 24.60 | 2.50 | 42.30 |
2 Objetivo 1 — Estado de crecimiento físico
2.1 Prevalencias por indicador
tabla_HAZ <- base %>%
filter(!is.na(Clasificacion_HAZ)) %>%
count(Clasificacion_HAZ) %>%
mutate(Porcentaje = round(n/sum(n)*100, 2))
tabla_HAZ %>%
kable(col.names = c("Categoría","n","%"),
caption = "Estado nutricional — Talla/Edad (HAZ) | n = 90") %>%
kable_styling(bootstrap_options = c("striped","hover","condensed"),
full_width = FALSE)| Categoría | n | % |
|---|---|---|
| Talla baja severa | 17 | 18.89 |
| Talla baja para la edad o retraso en talla | 4 | 4.44 |
| Riesgo de talla baja | 19 | 21.11 |
| Talla adecuada para la edad | 50 | 55.56 |
tabla_BAZ <- base %>%
filter(!is.na(Clasificacion_BAZ)) %>%
count(Clasificacion_BAZ) %>%
mutate(Porcentaje = round(n/sum(n)*100, 2))
tabla_BAZ %>%
kable(col.names = c("Categoría","n","%"),
caption = "Estado nutricional — IMC/Edad (BAZ) | n = 90") %>%
kable_styling(bootstrap_options = c("striped","hover","condensed"),
full_width = FALSE)| Categoría | n | % |
|---|---|---|
| Delgadez | 2 | 2.22 |
| Riesgo de delgadez | 7 | 7.78 |
| IMC adecuado para la edad | 47 | 52.22 |
| Sobrepeso | 9 | 10.00 |
| Obesidad | 25 | 27.78 |
tabla_WAZ <- base %>%
filter(!is.na(Clasificacion_WAZ)) %>%
count(Clasificacion_WAZ) %>%
mutate(Porcentaje = round(n/sum(n)*100, 2))
tabla_WAZ %>%
kable(col.names = c("Categoría","n","%"),
caption = "Estado nutricional — Peso/Edad (WAZ) | n = 45 (solo <5 años)") %>%
kable_styling(bootstrap_options = c("striped","hover","condensed"),
full_width = FALSE)| Categoría | n | % |
|---|---|---|
| Bajo peso para la edad | 4 | 8.89 |
| Peso adecuado para la edad | 41 | 91.11 |
2.2 Gráficos estado nutricional
ggplot(tabla_HAZ, aes(x = Clasificacion_HAZ, y = Porcentaje,
fill = Clasificacion_HAZ)) +
geom_col(color = "white", width = 0.65) +
geom_text(aes(label = paste0(n, " (", Porcentaje, "%)")),
vjust = -0.4, size = 3.5) +
scale_y_continuous(limits = c(0,70),
expand = expansion(mult = c(0,0.12))) +
labs(title = "Estado nutricional — Talla/Edad (HAZ)",
subtitle = "n = 90 | Ambos grupos de edad",
x = NULL, y = "Porcentaje (%)") +
theme_bw(base_size = 12) +
theme(axis.text.x = element_text(angle = 30, hjust = 1),
legend.position = "none")
Prevalencia por categoría — Talla/Edad (HAZ)
ggplot(tabla_BAZ, aes(x = Clasificacion_BAZ, y = Porcentaje,
fill = Clasificacion_BAZ)) +
geom_col(color = "white", width = 0.65) +
geom_text(aes(label = paste0(n, " (", Porcentaje, "%)")),
vjust = -0.4, size = 3.5) +
scale_y_continuous(limits = c(0,65),
expand = expansion(mult = c(0,0.12))) +
labs(title = "Estado nutricional — IMC/Edad (BAZ)",
subtitle = "n = 90 | Ambos grupos de edad",
x = NULL, y = "Porcentaje (%)") +
theme_bw(base_size = 12) +
theme(axis.text.x = element_text(angle = 30, hjust = 1),
legend.position = "none")
Prevalencia por categoría — IMC/Edad (BAZ)
ggplot(tabla_WAZ, aes(x = Clasificacion_WAZ, y = Porcentaje,
fill = Clasificacion_WAZ)) +
geom_col(color = "white", width = 0.55) +
geom_text(aes(label = paste0(n, " (", Porcentaje, "%)")),
vjust = -0.4, size = 3.5) +
scale_y_continuous(limits = c(0,110),
expand = expansion(mult = c(0,0.12))) +
labs(title = "Estado nutricional — Peso/Edad (WAZ)",
subtitle = "n = 45 | Solo menores de 5 años",
x = NULL, y = "Porcentaje (%)") +
theme_bw(base_size = 12) +
theme(axis.text.x = element_text(angle = 15, hjust = 1),
legend.position = "none")
Prevalencia por categoría — Peso/Edad (WAZ)
tabla_HAZ_grupo <- base %>%
filter(!is.na(Clasificacion_HAZ)) %>%
group_by(grupo_edad, Clasificacion_HAZ) %>%
summarise(n = n(), .groups = "drop") %>%
group_by(grupo_edad) %>%
mutate(Porcentaje = round(n / sum(n) * 100, 2))
ggplot(tabla_HAZ_grupo,
aes(x = Clasificacion_HAZ, y = Porcentaje, fill = grupo_edad)) +
geom_col(position = "dodge", color = "white", width = 0.7) +
geom_text(aes(label = paste0(Porcentaje, "%")),
position = position_dodge(0.7), vjust = -0.4, size = 3.2) +
scale_fill_manual(values = c("<5 años" = "#4472C4", "≥5 años" = "#70AD47")) +
scale_y_continuous(limits = c(0, 100),
expand = expansion(mult = c(0, 0.12))) +
labs(title = "Talla/Edad (HAZ) por grupo de edad",
x = NULL, y = "Porcentaje (%)", fill = "Grupo") +
theme_bw(base_size = 12) +
theme(axis.text.x = element_text(angle = 30, hjust = 1),
legend.position = "bottom")
HAZ por grupo de edad
tabla_BAZ_grupo <- base %>%
filter(!is.na(Clasificacion_BAZ)) %>%
group_by(grupo_edad, Clasificacion_BAZ) %>%
summarise(n = n(), .groups = "drop") %>%
group_by(grupo_edad) %>%
mutate(Porcentaje = round(n / sum(n) * 100, 2))
ggplot(tabla_BAZ_grupo,
aes(x = Clasificacion_BAZ, y = Porcentaje, fill = grupo_edad)) +
geom_col(position = "dodge", color = "white", width = 0.7) +
geom_text(aes(label = paste0(Porcentaje, "%")),
position = position_dodge(0.7), vjust = -0.4, size = 3.2) +
scale_fill_manual(values = c("<5 años" = "#4472C4", "≥5 años" = "#70AD47")) +
scale_y_continuous(limits = c(0, 85),
expand = expansion(mult = c(0, 0.12))) +
labs(title = "IMC/Edad (BAZ) por grupo de edad",
x = NULL, y = "Porcentaje (%)", fill = "Grupo") +
theme_bw(base_size = 12) +
theme(axis.text.x = element_text(angle = 30, hjust = 1),
legend.position = "bottom")
BAZ por grupo de edad
t_haz <- base %>% filter(!is.na(Clasificacion_HAZ)) %>%
count(Clasificacion_HAZ) %>%
mutate(Indicador = "Talla/Edad (HAZ)",
`%` = round(n/sum(n)*100, 2),
`n (%)` = paste0(n," (",`%`,"%)"),
Nota = "n total = 90 (ambos grupos)") %>%
rename(Categoria = Clasificacion_HAZ)
t_baz <- base %>% filter(!is.na(Clasificacion_BAZ)) %>%
count(Clasificacion_BAZ) %>%
mutate(Indicador = "IMC/Edad (BAZ)",
`%` = round(n/sum(n)*100, 2),
`n (%)` = paste0(n," (",`%`,"%)"),
Nota = "n total = 90 (ambos grupos)") %>%
rename(Categoria = Clasificacion_BAZ)
t_waz <- base %>% filter(!is.na(Clasificacion_WAZ)) %>%
count(Clasificacion_WAZ) %>%
mutate(Indicador = "Peso/Edad (WAZ)",
`%` = round(n/sum(n)*100, 2),
`n (%)` = paste0(n," (",`%`,"%)"),
Nota = "n = 45 (solo <5 años)") %>%
rename(Categoria = Clasificacion_WAZ)
t_whz <- base %>% filter(!is.na(Clasificacion_WHZ)) %>%
count(Clasificacion_WHZ) %>%
mutate(Indicador = "Peso/Talla (WHZ)",
`%` = round(n/sum(n)*100, 2),
`n (%)` = paste0(n," (",`%`,"%)"),
Nota = "n = 45 (solo <5 años)") %>%
rename(Categoria = Clasificacion_WHZ)
# Calcular rangos dinámicamente según filas reales
r1_ini <- 1
r1_fin <- nrow(t_haz)
r2_ini <- r1_fin + 1
r2_fin <- r2_ini + nrow(t_baz) - 1
r3_ini <- r2_fin + 1
r3_fin <- r3_ini + nrow(t_waz) - 1
r4_ini <- r3_fin + 1
r4_fin <- r4_ini + nrow(t_whz) - 1
bind_rows(t_haz, t_baz, t_waz, t_whz) %>%
select(Indicador, Categoria, n, `%`, `n (%)`, Nota) %>%
kable(caption = "Tabla consolidada — Estado nutricional por indicador") %>%
kable_styling(bootstrap_options = c("striped","hover","condensed")) %>%
pack_rows("Talla/Edad (HAZ)", r1_ini, r1_fin) %>%
pack_rows("IMC/Edad (BAZ)", r2_ini, r2_fin) %>%
pack_rows("Peso/Edad (WAZ)", r3_ini, r3_fin) %>%
pack_rows("Peso/Talla (WHZ)", r4_ini, r4_fin)| Indicador | Categoria | n | % | n (%) | Nota |
|---|---|---|---|---|---|
| Talla/Edad (HAZ) | |||||
| Talla/Edad (HAZ) | Talla baja severa | 17 | 18.89 | 17 (18.89%) | n total = 90 (ambos grupos) |
| Talla/Edad (HAZ) | Talla baja para la edad o retraso en talla | 4 | 4.44 | 4 (4.44%) | n total = 90 (ambos grupos) |
| Talla/Edad (HAZ) | Riesgo de talla baja | 19 | 21.11 | 19 (21.11%) | n total = 90 (ambos grupos) |
| Talla/Edad (HAZ) | Talla adecuada para la edad | 50 | 55.56 | 50 (55.56%) | n total = 90 (ambos grupos) |
| IMC/Edad (BAZ) | |||||
| IMC/Edad (BAZ) | Delgadez | 2 | 2.22 | 2 (2.22%) | n total = 90 (ambos grupos) |
| IMC/Edad (BAZ) | Riesgo de delgadez | 7 | 7.78 | 7 (7.78%) | n total = 90 (ambos grupos) |
| IMC/Edad (BAZ) | IMC adecuado para la edad | 47 | 52.22 | 47 (52.22%) | n total = 90 (ambos grupos) |
| IMC/Edad (BAZ) | Sobrepeso | 9 | 10.00 | 9 (10%) | n total = 90 (ambos grupos) |
| IMC/Edad (BAZ) | Obesidad | 25 | 27.78 | 25 (27.78%) | n total = 90 (ambos grupos) |
| Peso/Edad (WAZ) | |||||
| Peso/Edad (WAZ) | Bajo peso para la edad | 4 | 8.89 | 4 (8.89%) | n = 45 (solo <5 años) |
| Peso/Edad (WAZ) | Peso adecuado para la edad | 41 | 91.11 | 41 (91.11%) | n = 45 (solo <5 años) |
| Peso/Talla (WHZ) | |||||
| Peso/Talla (WHZ) | Normal | 16 | 35.56 | 16 (35.56%) | n = 45 (solo <5 años) |
| Peso/Talla (WHZ) | Riesgo de sobrepeso | 1 | 2.22 | 1 (2.22%) | n = 45 (solo <5 años) |
| Peso/Talla (WHZ) | Sobrepeso | 11 | 24.44 | 11 (24.44%) | n = 45 (solo <5 años) |
| Peso/Talla (WHZ) | Obesidad | 17 | 37.78 | 17 (37.78%) | n = 45 (solo <5 años) |
# Objetivo 2 — Ingesta de hierro ~ Hemoglobina y Hematocrito
## Pruebas de normalidad
``` r
calcular_ic_spearman <- function(r, n, conf = 0.95) {
z <- 0.5 * log((1 + r) / (1 - r))
se <- 1 / sqrt(n - 3)
z_cv <- qnorm((1 + conf) / 2)
round(c(tanh(z - z_cv * se), tanh(z + z_cv * se)), 3)
}
vars_norm2 <- list(
list(var = "ingesta_hierro", label = "Ingesta de hierro"),
list(var = "hemoglobina", label = "Hemoglobina"),
list(var = "hematocrito", label = "Hematocrito")
)
map_dfr(vars_norm2, function(v) {
x <- base[[v$var]][!is.na(base[[v$var]])]
res <- shapiro.test(x)
tibble(Variable = v$label,
n = length(x),
W = round(res$statistic, 4),
p_valor = round(res$p.value, 4),
Distribucion = if_else(res$p.value >= 0.05, "Normal", "No normal"))
}) %>%
kable(caption = "Prueba de Shapiro-Wilk — Objetivo 2") %>%
kable_styling(bootstrap_options = c("striped","hover","condensed"),
full_width = FALSE)| Variable | n | W | p_valor | Distribucion |
|---|---|---|---|---|
| Ingesta de hierro | 89 | 0.8543 | 0.0000 | No normal |
| Hemoglobina | 88 | 0.9679 | 0.0278 | No normal |
| Hematocrito | 88 | 0.9460 | 0.0011 | No normal |
par(mfrow = c(1,3))
qqnorm(base$ingesta_hierro, main = "QQ-Plot: Ingesta de hierro")
qqline(base$ingesta_hierro, col = "red")
qqnorm(base$hemoglobina, main = "QQ-Plot: Hemoglobina")
qqline(base$hemoglobina, col = "red")
qqnorm(base$hematocrito, main = "QQ-Plot: Hematocrito")
qqline(base$hematocrito, col = "red")
QQ-plots — variables Objetivo 2
2.3 Correlaciones de Spearman
cor_hb <- suppressWarnings(
cor.test(base$ingesta_hierro, base$hemoglobina,
method = "spearman", exact = FALSE))
cor_hto <- suppressWarnings(
cor.test(base$ingesta_hierro, base$hematocrito,
method = "spearman", exact = FALSE))
n_hb <- sum(!is.na(base$ingesta_hierro) & !is.na(base$hemoglobina))
n_hto <- sum(!is.na(base$ingesta_hierro) & !is.na(base$hematocrito))
ic_hb <- calcular_ic_spearman(as.numeric(cor_hb$estimate), n_hb)
ic_hto <- calcular_ic_spearman(as.numeric(cor_hto$estimate), n_hto)
tibble(
Asociacion = c("Ingesta hierro ~ Hemoglobina",
"Ingesta hierro ~ Hematocrito"),
Metodo = "Spearman",
n = c(n_hb, n_hto),
rho = round(c(cor_hb$estimate, cor_hto$estimate), 3),
IC_95 = c(paste0("[",ic_hb[1],"; ",ic_hb[2],"]"),
paste0("[",ic_hto[1],"; ",ic_hto[2],"]")),
p_valor = c(round(cor_hb$p.value, 4), round(cor_hto$p.value, 4)),
Significativo = c("Sí ***","Sí ***")
) %>%
kable(caption = "Correlaciones de Spearman — Objetivo 2") %>%
kable_styling(bootstrap_options = c("striped","hover","condensed"))| Asociacion | Metodo | n | rho | IC_95 | p_valor | Significativo |
|---|---|---|---|---|---|---|
| Ingesta hierro ~ Hemoglobina | Spearman | 87 | 0.765 | [0.66; 0.84] | 0 | Sí *** |
| Ingesta hierro ~ Hematocrito | Spearman | 87 | 0.531 | [0.361; 0.667] | 0 | Sí *** |
2.4 Regresión lineal simple
modelo_hb <- lm(hemoglobina ~ ingesta_hierro, data = base)
modelo_hto <- lm(hematocrito ~ ingesta_hierro, data = base)
bind_rows(
tidy(modelo_hb, conf.int = TRUE) %>% mutate(Outcome = "Hemoglobina"),
tidy(modelo_hto, conf.int = TRUE) %>% mutate(Outcome = "Hematocrito")
) %>%
mutate(
B = round(estimate, 3),
IC_95 = paste0("[", round(conf.low,3), "; ", round(conf.high,3), "]"),
p = round(p.value, 4),
Sig = case_when(p < 0.001 ~ "***", p < 0.01 ~ "**",
p < 0.05 ~ "*", TRUE ~ "ns")
) %>%
select(Outcome, Predictor = term, B, IC_95, p, Sig) %>%
kable(caption = "Regresión lineal simple") %>%
kable_styling(bootstrap_options = c("striped","hover","condensed"))| Outcome | Predictor | B | IC_95 | p | Sig |
|---|---|---|---|---|---|
| Hemoglobina | (Intercept) | 10.702 | [10.362; 11.042] | 0 | *** |
| Hemoglobina | ingesta_hierro | 0.074 | [0.06; 0.087] | 0 | *** |
| Hematocrito | (Intercept) | 31.658 | [30.239; 33.077] | 0 | *** |
| Hematocrito | ingesta_hierro | 0.178 | [0.122; 0.235] | 0 | *** |
2.5 Regresión lineal múltiple ajustada
modelo_hb_multi <- lm(
hemoglobina ~ ingesta_hierro + edad_anos + sexo + HAZ, data = base)
modelo_hto_multi <- lm(
hematocrito ~ ingesta_hierro + edad_anos + sexo + HAZ, data = base)
bind_rows(
tidy(modelo_hb_multi, conf.int = TRUE) %>%
mutate(Outcome = paste0("Hemoglobina (R²=",
round(summary(modelo_hb_multi)$r.squared,3),")")),
tidy(modelo_hto_multi, conf.int = TRUE) %>%
mutate(Outcome = paste0("Hematocrito (R²=",
round(summary(modelo_hto_multi)$r.squared,3),")"))
) %>%
mutate(
B = round(estimate, 3),
IC_95 = paste0("[", round(conf.low,3), "; ", round(conf.high,3), "]"),
p = round(p.value, 4),
Sig = case_when(p < 0.001 ~ "***", p < 0.01 ~ "**",
p < 0.05 ~ "*", TRUE ~ "ns")
) %>%
select(Outcome, Predictor = term, B, IC_95, p, Sig) %>%
kable(caption = "Regresión lineal múltiple ajustada (edad, sexo, HAZ)") %>%
kable_styling(bootstrap_options = c("striped","hover","condensed"))| Outcome | Predictor | B | IC_95 | p | Sig |
|---|---|---|---|---|---|
| Hemoglobina (R²=0.629) | (Intercept) | 10.536 | [9.939; 11.133] | 0.0000 | *** |
| Hemoglobina (R²=0.629) | ingesta_hierro | 0.056 | [0.039; 0.073] | 0.0000 | *** |
| Hemoglobina (R²=0.629) | edad_anos | 0.063 | [0.017; 0.108] | 0.0081 | ** |
| Hemoglobina (R²=0.629) | sexoFemenino | 0.039 | [-0.327; 0.406] | 0.8310 | ns |
| Hemoglobina (R²=0.629) | HAZ | -0.029 | [-0.159; 0.1] | 0.6540 | ns |
| Hematocrito (R²=0.363) | (Intercept) | 32.441 | [29.863; 35.019] | 0.0000 | *** |
| Hematocrito (R²=0.363) | ingesta_hierro | 0.125 | [0.053; 0.197] | 0.0009 | *** |
| Hematocrito (R²=0.363) | edad_anos | 0.125 | [-0.07; 0.319] | 0.2052 | ns |
| Hematocrito (R²=0.363) | sexoFemenino | -0.665 | [-2.246; 0.915] | 0.4047 | ns |
| Hematocrito (R²=0.363) | HAZ | 0.212 | [-0.349; 0.773] | 0.4542 | ns |
2.6 Gráficos dispersión
ggplot(base %>% drop_na(ingesta_hierro, hemoglobina),
aes(x = ingesta_hierro, y = hemoglobina)) +
geom_point(alpha = 0.7, color = "#4472C4", size = 2.5) +
geom_smooth(method = "lm", se = TRUE, color = "#C00000") +
labs(title = "Ingesta de hierro vs. Hemoglobina",
x = "Ingesta de hierro (mg/día)", y = "Hemoglobina (g/dL)") +
theme_bw(base_size = 12)
Ingesta de hierro vs. Hemoglobina
ggplot(base %>% drop_na(ingesta_hierro, hematocrito),
aes(x = ingesta_hierro, y = hematocrito)) +
geom_point(alpha = 0.7, color = "#70AD47", size = 2.5) +
geom_smooth(method = "lm", se = TRUE, color = "#C00000") +
labs(title = "Ingesta de hierro vs. Hematocrito",
x = "Ingesta de hierro (mg/día)", y = "Hematocrito (%)") +
theme_bw(base_size = 12)
Ingesta de hierro vs. Hematocrito
3 Objetivo 3 — Z-scores de crecimiento ~ Hemoglobina y Hematocrito
3.1 Pruebas de normalidad
vars_norm3 <- list(
list(var = "HAZ", label = "HAZ"),
list(var = "BAZ", label = "BAZ"),
list(var = "hemoglobina", label = "Hemoglobina"),
list(var = "hematocrito", label = "Hematocrito")
)
map_dfr(vars_norm3, function(v) {
x <- base[[v$var]][!is.na(base[[v$var]])]
res <- shapiro.test(x)
tibble(Variable = v$label,
n = length(x),
W = round(res$statistic, 4),
p_valor = round(res$p.value, 4),
Distribucion = if_else(res$p.value >= 0.05, "Normal", "No normal"))
}) %>%
kable(caption = "Prueba de Shapiro-Wilk — Objetivo 3") %>%
kable_styling(bootstrap_options = c("striped","hover","condensed"),
full_width = FALSE)| Variable | n | W | p_valor | Distribucion |
|---|---|---|---|---|
| HAZ | 90 | 0.9234 | 0.0001 | No normal |
| BAZ | 90 | 0.9231 | 0.0001 | No normal |
| Hemoglobina | 88 | 0.9679 | 0.0278 | No normal |
| Hematocrito | 88 | 0.9460 | 0.0011 | No normal |
3.2 Correlaciones de Spearman con IC 95%
pares <- list(
list(x="HAZ", y="hemoglobina", xl="HAZ", yl="Hemoglobina", rho= 0.5278),
list(x="HAZ", y="hematocrito", xl="HAZ", yl="Hematocrito", rho= 0.3796),
list(x="BAZ", y="hemoglobina", xl="BAZ", yl="Hemoglobina", rho=-0.4600),
list(x="BAZ", y="hematocrito", xl="BAZ", yl="Hematocrito", rho=-0.2940)
)
n_obj3 <- sum(!is.na(base$HAZ) & !is.na(base$hemoglobina))
map_dfr(pares, function(p) {
res <- suppressWarnings(
cor.test(base[[p$x]], base[[p$y]],
method = "spearman", exact = FALSE))
ic <- calcular_ic_spearman(p$rho, n_obj3)
tibble(
Asociacion = paste0(p$xl, " ~ ", p$yl),
Metodo = "Spearman",
n = n_obj3,
rho = round(res$estimate, 3),
IC_95 = paste0("[", ic[1], "; ", ic[2], "]"),
p_valor = round(res$p.value, 4),
Significativo = if_else(res$p.value < 0.05, "Sí", "No")
)
}) %>%
kable(caption = "Correlaciones de Spearman — Objetivo 3") %>%
kable_styling(bootstrap_options = c("striped","hover","condensed"))| Asociacion | Metodo | n | rho | IC_95 | p_valor | Significativo |
|---|---|---|---|---|---|---|
| HAZ ~ Hemoglobina | Spearman | 88 | 0.528 | [0.358; 0.664] | 0.0000 | Sí |
| HAZ ~ Hematocrito | Spearman | 88 | 0.380 | [0.185; 0.546] | 0.0003 | Sí |
| BAZ ~ Hemoglobina | Spearman | 88 | -0.460 | [-0.611; -0.277] | 0.0000 | Sí |
| BAZ ~ Hematocrito | Spearman | 88 | -0.294 | [-0.474; -0.09] | 0.0054 | Sí |
3.3 Gráficos dispersión
ggplot(base %>% drop_na(HAZ, hemoglobina),
aes(x = HAZ, y = hemoglobina, color = grupo_edad)) +
geom_point(alpha = 0.7, size = 2.5) +
geom_smooth(method = "lm", se = TRUE) +
geom_vline(xintercept = -2, linetype = "dashed",
color = "gray40", linewidth = 0.6) +
scale_color_manual(values = colores_grupo) +
labs(title = "Z-score Talla/Edad (HAZ) vs. Hemoglobina",
x = "Z-score HAZ", y = "Hemoglobina (g/dL)", color = "Grupo") +
theme_bw(base_size = 12) + theme(legend.position = "bottom")
HAZ vs. Hemoglobina por grupo de edad
ggplot(base %>% drop_na(HAZ, hematocrito),
aes(x = HAZ, y = hematocrito, color = grupo_edad)) +
geom_point(alpha = 0.7, size = 2.5) +
geom_smooth(method = "lm", se = TRUE) +
geom_vline(xintercept = -2, linetype = "dashed",
color = "gray40", linewidth = 0.6) +
scale_color_manual(values = colores_grupo) +
labs(title = "Z-score Talla/Edad (HAZ) vs. Hematocrito",
x = "Z-score HAZ", y = "Hematocrito (%)", color = "Grupo") +
theme_bw(base_size = 12) + theme(legend.position = "bottom")
HAZ vs. Hematocrito por grupo de edad
ggplot(base %>% drop_na(BAZ, hemoglobina),
aes(x = BAZ, y = hemoglobina, color = grupo_edad)) +
geom_point(alpha = 0.7, size = 2.5) +
geom_smooth(method = "lm", se = TRUE) +
scale_color_manual(values = colores_grupo) +
labs(title = "Z-score IMC/Edad (BAZ) vs. Hemoglobina",
x = "Z-score BAZ", y = "Hemoglobina (g/dL)", color = "Grupo") +
theme_bw(base_size = 12) + theme(legend.position = "bottom")
BAZ vs. Hemoglobina por grupo de edad
ggplot(base %>% drop_na(BAZ, hematocrito),
aes(x = BAZ, y = hematocrito, color = grupo_edad)) +
geom_point(alpha = 0.7, size = 2.5) +
geom_smooth(method = "lm", se = TRUE) +
scale_color_manual(values = colores_grupo) +
labs(title = "Z-score IMC/Edad (BAZ) vs. Hematocrito",
x = "Z-score BAZ", y = "Hematocrito (%)", color = "Grupo") +
theme_bw(base_size = 12) + theme(legend.position = "bottom")
BAZ vs. Hematocrito por grupo de edad