No abstract

Background: Fluid intelligence declines with advancing age, starting in early adulthood. Within-subject declines in fluid intelligence are highly correlated with contemporaneous declines in the ability to live and function independently. To support healthy aging, the mechanisms underlying these declines need to be better understood. Methods: In this pre-registered analysis, we applied latent growth curve modelling to investigate the neural determinants of longitudinal changes in fluid intelligence across three time points in 185,317 individuals (N=9,719 two waves, N=870 three waves) from the UK Biobank (age range: 39-73 years). Results: We found a weak but significant effect of cross-sectional age on the mean fluid intelligence score, such that older individuals scored slightly lower. However, the mean longitudinal slope was positive, rather than negative, suggesting improvement across testing occasions. Despite the considerable sample size, the slope variance was non-significant, suggesting no reliable individual differences in change over time. This null-result is likely due to the nature of the cognitive test used. In a subset of individuals, we found that white matter microstructure (N=8839, as indexed by fractional anisotropy) and grey-matter volume (N=9931) in pre-defined regions-of-interest accounted for complementary and unique variance in mean fluid intelligence scores. The strongest effects were such that higher grey matter volume in the frontal pole and greater white matter microstructure in the posterior thalamic radiations were associated with higher fluid intelligence scores. Conclusions: In a large preregistered analysis, we demonstrate a weak but significant negative association between age and fluid intelligence. However, we did not observe plausible longitudinal patterns, instead observing a weak increase across testing occasions, and no significant individual differences in rates of change, likely due to the suboptimal task design. Finally, we find support for our preregistered expectation that white- and grey matter make separate contributions to individual differences in fluid intelligence beyond age.

Fluid intelligence declines with advancing age, starting in early adulthood. Within-subject declines in fluid intelligence are highly correlated with contemporaneous declines in the ability to live and function independently. To support healthy aging, the mechanisms underlying these declines need to be better understood. In this pre-registered analysis, we applied latent growth curve modelling to investigate the neural determinants of longitudinal changes in fluid intelligence across three time points in 185,317 individuals (N=9,719 two waves, N=870 three waves) from the UK Biobank (age range: 39-73 years).We found a weak but significant effect of cross-sectional age on the mean fluid intelligence score, such that older individuals scored slightly lower. However, the mean longitudinal slope was positive, rather than negative, suggesting improvement across testing occasions. Despite the considerable sample size, the slope variance was non-significant, suggesting no reliable individual differences in change over time. This null-result is likely due to the nature of the cognitive test used. In a subset of individuals, we found that white matter microstructure (N=8839, as indexed by fractional anisotropy) and grey-matter volume (N=9931) in pre-defined regions-of-interest accounted for complementary and unique variance in mean fluid intelligence scores. The strongest effects were such that higher grey matter volume in the frontal pole and greater white matter microstructure in the posterior thalamic radiations were associated with higher fluid intelligence scores. In a large preregistered analysis, we demonstrate a weak but significant negative association between age and fluid intelligence. However, we did not observe plausible longitudinal patterns, instead observing a weak increase across testing occasions, and no significant individual differences in rates of change, likely due to the suboptimal task design. Finally, we find support for our preregistered expectation that white- and grey matter make separate contributions to individual differences in fluid intelligence beyond age.

Despite the reliability of intelligence measures in predicting important life outcomes such as educational achievement and mortality, the exact configuration and neural correlates of cognitive abilities remain poorly understood, especially in childhood and adolescence. Therefore, we sought to elucidate the factorial structure and neural substrates of child and adolescent intelligence using two cross-sectional, developmental samples (CALM: N=551 (N=165 imaging), age range: 5-18 years, NKI-Rockland: N=337 (N=65 imaging), age range: 6-18 years). In a preregistered analysis, we used structural equation modelling (SEM) to examine the neurocognitive architecture of individual differences in childhood and adolescent cognitive ability. In both samples, we found that cognitive ability in lower and typical-ability cohorts is best understood as two separable constructs, crystallized and fluid intelligence, which became more distinct across development, in line with the age differentiation hypothesis. Further analyses revealed that white matter microstructure, most prominently the superior longitudinal fasciculus, was strongly associated with crystallized (gc) and fluid (gf)abilities. Finally, we used SEM trees to demonstrate evidence for developmental reorganization of gc and gf and their white matter substrates such that the relationships among these factors dropped between 7-8 years before increasing around age 10. Together, our results suggest that shortly before puberty marks a pivotal phase of change in the neurocognitive architecture of intelligence.

The thickness and surface area of cortex are genetically distinct aspects of brain structure, and may be affected differently by age. However, their potential to differentially predict age and cognitive abilities has been largely overlooked, likely because they are typically aggregated into the commonly used measure of volume. In a large sample of healthy adults (N=647, aged 18-88), we investigated the brain-age and brain-cognition relationships of thickness, surface area, and volume, plus five additional morphological shape metrics. Cortical thickness was the metric most strongly associated with age cross-sectionally, as well as exhibiting the steepest longitudinal change over time (subsample N=261, aged 25-84). In contrast, surface area was the best single predictor of age-residualized cognitive abilities (fluid intelligence), and changes in surface area were most strongly associated with cognitive change over time. These findings were replicated in an independent dataset (N=1345, aged 18-93). Our results suggest that cortical thickness and surface area make complementary contributions the age-brain-cognition triangle, and highlight the importance of considering these volumetric components separately.

Previous evidence suggests that modifiable lifestyle factors, such as engagement in leisure activities, might slow the age-related decline of cognitive functions. Less is known, however, about which aspects of lifestyle might be particularly beneficial to healthy cognitive ageing, and whether they are associated with distinct cognitive domains (e.g. fluid and crystallized abilities) differentially. We investigated these questions in the cross-sectional Cambridge Centre for Ageing and Neuroscience (Cam-CAN) data (N = 708, age 18-88), using datadriven exploratory structural equation modelling, confirmatory factor analyses, and age-residualized measures of cognitive differences across the lifespan. Specifically, we assessed the relative associations of the following five lifestyle factors on age-related differences of fluid and crystallized age-adjusted abilities: education/SES, physical health, mental health, social engagement, and intellectual engagement. We found that higher education, better physical and mental health, more social engagement and a greater degree of intellectual engagement were each individually correlated with better fluid and crystallized cognitive age-adjusted abilities. A joint path model of all lifestyle factors on crystallized and fluid abilities, which allowed a simultaneous assessment of the lifestyle domains, showed that physical health, social and intellectual engagement and education/SES explained unique, complementary variance, but mental health did not make significant contributions above and beyond the other four lifestyle factors and age. The total variance explained for fluid abilities was 14% and 16% for crystallized abilities. Our results are compatible with the hypothesis that intellectually and physically challenging as well as socially engaging activities are associated with better crystallized and fluid performance across the lifespan.

No abstract

Mediation - or models where an intervening variable is thought to propagate an affect from a cause to an outcome of interest - is a ubiquitous statistical tool in the behavioral and medical sciences. Despite a long history of thoughtful critiques of its use in applied research, mediation remains inferentially attractive and easy to implement in widely-available software. Here we highlight a challenge of mediation that has not yet received appropriate consideration - namely the potential for improper causal inferences when the mediator is a derived, or composite, score of a set of items or measures (e.g., sums, means, differences, etc.). We show that composites have the potential to confound or mask different causal processes that occur at the individual item level, leading to spurious mediation effects. As composite measures are near-universal in many fields from the psychological to biomedical sciences, this limitation of inferences potentially impacts a very broad set of research questions. In this paper we demonstrate this issue in a diversity of mediators, using both empirical examples of grey matter volume and depression questionnaires and a wide array of simulated examples. Additionally, we provide several potential analytical approaches with may help diagnose spurious mediation with composite measures and provide greater insight into the causal structure of our data. As part of this set of approaches, we advocate a transition from traditional, regression-based approaches to mediation towards a structural equation modeling (SEM) framework which provides many advantages. We also note additional conditions under which interrupted mediation can occur, providing directions for future developments in this area.