Skip to main content

Table 1 Specification of the AFT models to predict ages at final menstrual period (T i )

From: Bayesian estimation of associations between identified longitudinal hormone subgroups and age at final menstrual period

Model

AFT models

M 0

log(T i  ‐ 40) = α 0 + x T i θ + ε i

M 1

log(T i  ‐ 40) = α 0 + α 1 D i  + α 2 ω i  + x T i θ + ε i

M 2

log(T i  ‐ 40) = α 0 + α 1 D i  + α 2 ω i  + α 3 μ i (40)(1 ‐ D i ) + α 4 μ i (45)D i  + x T i θ + ε i

M 3

log(T i  ‐ 40) = α 0 + α 1 D i  + α 2 ω i  + α 3 ν i (40)(1 ‐ D i ) + α 4 ν i (45)D i  + x T i θ + ε i

M 4

log(T i  ‐ 40) = α 0 + α 1 D i  + α 2 ω i  + α 3 μ i (40)(1 ‐ D i ) + α 4 μ i (45)D i  + α 5 ν i (40)(1 ‐ D i ) + α 6 ν i (45)D i  + x T i θ + ε i

  1. Note:
  2. x T i = (adjusted log (BMI) at age 40, adjusted AMH at age 40, smoking, race)
  3. ε i N(0, σ 2) is the residual for log-normal AFT model
  4. D i is the early FSH rise class indicator and ω i is the within-subject variability in FSH; μ i (τ) and ν i (τ) are the mean FSH level and its rate of change at age τ, respectively