Age contributes more to interindividual variability in cognitive performance than other individual difference characteristics. Although the impact of age on both executive functions (EFs) and intelligence has been frequently investigated, there is little research on the relationship between the two constructs across different ages. The aim of our study was to explore the relationships between EFs and intelligence by the comparison of younger and older healthy adults. Additionally, the aim was to investigate age-related predictability of intelligence with EFs.
Data from five subtests of the Executive Functions Module of the German Neuropsychological Assessment Battery (NAB) and ten core subtests of the German Wechsler Adult Intelligence Scale – Fourth Edition (WAIS-IV) were used to examine the relationship between EFs and intelligence in a comparison of two age groups: individuals aged 18–59 years (N = 65) and individuals aged 60–88 years (N = 61).
The results demonstrate moderate to substantial correlations between the NAB Executive Functions Module and the WAIS-IV. In most cases, there were no significant differences between younger and older participants in respect to the correlationship between the two measures. Particularly, the NAB Categories and Word Generation subtests showed a consistent correlation pattern for both age groups. The two complex and multifaceted tasks correlated substantially with most of the WAIS-IV indices and were most frequently included in the WAIS-IV best-fitting prediction models; whereas NAB Letter Fluency, which is a specific language-related task, was least frequently included in the WAIS-IV best-fitting models. Nevertheless, significant differences between the two age groups were identified, too; and in all cases, the older age group had the higher correlation coefficient. Particularly, the NAB Judgment subtest correlated more strongly with the WAIS-IV Verbal Comprehension Index (VCI), z = 2.29, p = .011, q = .42; Perceptual Reasoning Index, z =2.15, p =.016, q = .39; Working Memory Index, z = 2.01, p =.022, q = .37; Full Scale IQ (FSIQ), z = 2.38, p =.008, q = .43; and General Ability Index, z = 2.53, p =.006, q = .46. Additionally, there was a stronger correlation between the Executive Functions Index (EFI) and VCI, z = 2.22, p =.013, q = .40 and a considerable difference in the total variance explained by the VCI prediction models between older (75%) and younger participants (28%). Moreover, the NAB subtests along with age better predicted the WAIS-IV FSIQ in older than in younger participants, with the total variance explained, 72% in 60- to 88-year-olds, and 45% in 18- to 59-year-olds.
The results indicate that intelligence performance can be better predicted by complex or multi-domain EF tasks rather than by well-structured, domain specific EF tasks. Furthermore, the relationship between EFs and intelligence may become stronger as age advances, which is consistent with the dedifferentation-hypothesis (Balinsky, 1941; Baltes & Lindenberger, 1997). In particular, EFs may be stronger related to Gc at older ages, which can partially be explained by the Investment theory (Cattell, 1963, 1987). Additionally, the potential diagnostic utility of the Judgment subtest at older ages should be considered.
Ageing , Measurement and Psychometrics , Group differences
