Skip to main content

Table 1 Model fit statistics for quadratic LCGMM comparing solutions for 1 to 7 latent classes

From: Childhood adiposity trajectories are associated with late adolescent blood pressure: birth to twenty cohort

Sex Latent classes BICa LMR-LRTb (P-value) BLRTc (P-value) Convergence Entropy value Log Likelihood Number of parameters
Boys 1 23605.4 N/A N/A Yes N/A −11738.3 19
2 22846.6 0.03 0.00 Yes 0.98 −11345.4 23
3 d 22644.4 0.13 0.67 Yes 0.99 −11230.7 27
4 22544.7 0.53 1.00 Yes 0.98 −11167.3 31
5 22476.5 0.77 0.67 Yes 0.93 −11119.7 35
6 22438.4 0.61 1.00 Yes 0.94 −11087.1 39
7     NO    43
Girls 1 27949.8 N/A N/A Yes N/A −13909.8 19
2 27226.2 0.00 0.00 Yes 0.94 −13534.3 23
3 27155.0 0.10 0.00 Yes 0.76 −13485.0 27
4 d 27083.9 0.23 0.00 Yes 0.80 −13435.7 31
5 27026.0 0.45 0.6 Yes 0.82 −13393.1 35
6 26990.9 0.63 1.00 Yes 0.78 −13361.8 39
7 26961.2 0.27 1.00 Yes 0.80 −13333.3 43
  1. LCGMM latent class growth mixture modeling
  2. aBayesian Information Criterion
  3. bLo, Mendell and Rubin Likelihood Ratio Test
  4. cBootstrapped Likelihood Ratio Test, N/A – not applicable
  5. dThe optimal class solutions according to the model fit criteria are shown in bold