Skip to main content
. Author manuscript; available in PMC: 2017 Feb 2.
Published in final edited form as: Neuropsychol Rev. 2015 Mar 3;25(1):97–112. doi: 10.1007/s11065-015-9280-2

Table 2.

Multilevel fixed effect estimates of MZ/DZ ratios by cognitive domain.

Domain N Obs Fixed Effect Estimate (ln) se LCL UCL Expected rMZ/rDZ “Model”
Verbal 21 Intercept 1.379 0.112 1.114 1.644 2.97 ADE*
Age 0.000 0.009 -0.020 0.021 --
Age2 0.000 0.000 -0.001 0.000 --
Spatial 14 Intercept 1.148 0.102 0.885 1.410 2.15 ADE
Age -0.005 0.009 -0.030 0.018 -- --
Age2 0.000 0.000 -0.001 0.001 -- --
Speed 20 Intercept 1.116 0.092 0.892 1.340 2.05 A?E
Age 0.007 0.011 -0.017 0.031 -- --
Age2 0.001 0.000 0.000 0.002 -- --
General Memory 23 Intercept 1.017 0.185 0.565 1.469 1.76 ACE
Age 0.004 0.011 -0.019 0.028 -- --
Age2 0.000 0.000 0.000 0.001 -- --
Working Memory 15 Intercept 1.453 0.190 1.004 1.902 3.28 ADE
Age -0.012 0.012 -0.044 0.020 -- --
Age2 -0.001 0.001 -0.003 0.001 -- --
Executive Functioning 8 Intercept 1.283 0.125 -0.307 2.872 2.61 ADE
Age 0.006 0.004 -0.005 0.017 -- --

Note. Each ratio was transformed accordingly: ln (rMZ/rDZ +1). N Obs = number of data points contributing to each analysis. Analyses were weighted based on pair sample size. The fixed effect estimates and standard errors are presented from multi-level modeling in SAS Proc Mixed (Cary, NC), where LCL=95% Lower Confidence Limit and UCL = 95% Upper confidence limit. The intercept reflects the expected ratio in natural log units, adjusted for age- and age-squared terms centered at 65 years, with the exception of Executive Functioning that entered only the linear age term. If the intercept estimates exceed 1.099 that would indicate an ADE model may be warranted when back-calculated into raw ratio units (i.e., exceed 2.0). Back-calculated expected rMZ/rDZ ratios are provided. ADE* = support of ADE model based on 95% CI of the intercept that excludes values 1.099 and below.