Skip to main content
. Author manuscript; available in PMC: 2012 Sep 1.
Published in final edited form as: Psychol Bull. 2011 Sep;137(5):753–784. doi: 10.1037/a0023262

Table 2.

Analyses with White Matter Hyperintensities (WMH) as the Brain Structure Variable

C B A–C A–C.B A–B A–B.C B–C B–C.A
Cook et al. (2002), N = 43, Ages 60 to 93
Trail Making B Peri-ventricular WMH .57 .53 .47 .42 .33 .09
Deep WMH .57 .53 .22 .07 .30 .22
Shipley Abstraction Peri-ventricular WMH −.57 −.53 .47 .42 −.33 −.09
Deep WMH −.57 −.55 .22 .15 −.21 −.11
Gunning-Dixon & Raz (2003), N = 139, Ages 50 to 81
Working Memory (Verbal) Frontal WMH −.27 −.22 .40 .37 −.22 −.13
Temporal WMH −.27 −.23 .48 .46 −.19 −.07
WCST (Perseverations) Frontal WMH .27 .16 .40 .33 .34 .26
Temporal WMH .27 .25 .48 .47 .17 .05
Gootjes et al. (2007), N = 36, Ages 50 to 81
Dichotic listening Deep WMH −.59 −.57 .67 .66 −.41 −.03
Charlton et al. (2010), N = 104, Ages 50 to 93
Immediate memory WMH −.57 −.55 .48 .46 −.30 −.04
Rabbitt et al. (2007), N = 65, Ages 64 to 85
Culture Fair IQ WMH −.41 −.40 .38 .37 −.18 −.03
AH4-1 IQ WMH −.41 −.36 .38 .32 −.27 −.14
AH4-2 IQ WMH −.48 −.47 .38 .37 −.20 −.02
WAIS Vocabulary WMH .21 .12 .38 .34 .28 .22
Mill Hill A Vocabulary WMH −.16 −.13 .38 .37 −.13 −.08
Mill Hill B Vocabulary WMH .16 .09 .38 .36 .21 .16
Forward Digit Span WMH −.21 −.15 .38 .35 −.21 −.14
Backward Digit Span WMH .19 .07 .38 .33 .34 .30
Word Span WMH .17 .09 .38 .35 .24 .19
Memory for People WMH .29 .20 .38 .32 .31 .23
Memory for Objects in a Circle WMH −.52 −.46 .38 .28 −.34 −.18
Visual Search WMH .31 .19 .38 .29 .38 .30
Letter-Digit Substitution WMH .33 .19 .38 .26 .44 .36
Phonemic fluency WMH .22 .16 .38 .35 .22 .15
Category fluency WMH .35 .24 .38 .28 .38 .29
Brixton WMH .37 .34 .38 .35 .22 .09
Rey-Osterrieth Figure WMH −.56 −.51 .38 .28 −.33 −.15

Note: All of the B–C combinations within a given study presumably involved the same research participants. It is likely that there was at least some overlap in the samples reported in different articles by the same research team. However, the degree of overlap was not always stated, and because results were not reported separately for the new and old participants, the data were not suitable for meta-analyses.

Entries in bold represent combinations that were most consistent with the mediation model in that the reduction in A–C after control of B was larger than both the reduction of A–B after control of C and the reduction of B–C after control of A.

NA indicates that the estimates could not be computed because the relevant correlations were not available.