その要件だと以下のような感じでいけると思います

library(tidyverse)

df <- tibble(
time = rep(1:4, each = 5),
id = rep(1:5, 4),
a = rnorm(20),
b = rnorm(20),
c = rnorm(20),
d = rnorm(20),
e = rnorm(20)
)

df |>
pivot_wider(id_cols = id, names_from = time, values_from = c(a, e)) |>
ggplot(aes(a_1, e_4)) +　geom_point()