Processing speed and executive functions in cognitive aging: How to disentangle their mutual relationship?

https://doi.org/10.1016/j.bandc.2012.02.001Get rights and content

Abstract

The processing-speed theory and the prefrontal-executive theory are competing theories of cognitive aging. Here we used a theoretically and methodologically-driven framework to investigate the relationships among measures classically used to assess these two theoretical constructs. Twenty-eight young adults (18–32 years) and 39 healthy older adults (65–80 years) performed a battery of nine neuropsychological and experimental tasks assessing three executive function (EF) components: Inhibition, Updating, and Shifting. Rate of information processing was evaluated via three different experimental and psychometric tests. Partial correlations analyses suggested that 2-Choice Reaction Time (CRT) performance is a more pure measure of processing speed than Digit Symbol Substitution Test (DSST) performance in the elderly. Hierarchical regression analyses showed that, although measures of processing speed and EF components share mutual variance, each measure was independently affected by chronological age. The unique adverse effect of age was more important for processing speed than for EF. The processing-speed theory and the prefrontal-executive theory of cognitive aging were shown not to be mutually exclusive but share mutual variance. This implies the need to control for their mutual relationship before examining their unique potential role in the explanation of age-related cognitive declines. Caution has still to be taken concerning the tasks used to evaluate these theoretical constructs.

Highlights

► Processing speed and executive functions as mediators of cognitive aging. ► Twelve experimental tasks to assess these two constructs in young and older adults. ► The two constructs share mutual variance but are independently affected by age. ► The unique adverse effect of age was more important for processing speed. ► More work is still needed to refine the tasks used to assess these two constructs.

Introduction

There is considerable interest in the area of cognitive and brain sciences in understanding and explaining the age-related declines in cognitive functioning. Numerous theories and models, from neurophysiological perspectives to more behavioral and cognitive perspectives, have been proposed (see Cabeza et al., 2005, Reuter-Lorenz and Park, 2010, Schaie and Willis, 2011, for reviews). To date, the most influential and empirically tested models in neuropsychology and cognitive literature are the processing speed theory (Salthouse, 1996) and the prefrontal-executive theory (West, 1996). Speed of information processing and executive functioning have been thus proposed as theoretical constructs and many researchers have studied them as candidates responsible for mediating (or even explaining), either in whole or in part, the age-related declines in cognitive functioning. As we will review, the conclusions are remarkably mixed: around half of the studies support the slowdown in processing speed and the other half support the decline in executive functions (EF) to be the actual mediator. However, the relationships between these two theoretical models and the tasks used to test them are understudied and poorly understood. It is necessary to determine if speed of information processing and EF are clearly independent, and to elucidate each constructs’ explanatory contribution to the understanding of general age-related cognitive decline. The principal objective of the present study is to propose a theory-based, methodological framework concerning the operationalization and the measurement of EF and speed of information processing, with the aim of examining their potential mutual relationships.

Over the last two decades, the so-called global and local theories of cognitive aging have received considerable attention. On the one side, one of these theories, referred to as the processing-speed theory, contends that age-related cognitive declines can be accounted for by a single or global mechanism: the generalized slowing of cognitive processing. This generalized slowing is thought to be due in a large part to a diffuse or global deterioration of white matter integrity throughout the brain. On the other side, the prefrontal-executive theory states that local structural and functional changes in frontal cortex areas lead to specific declines in executive abilities, which in turn lead to more general cognitive deficits.

Following years of empirical and theoretical work, Salthouse (1996) proposed the processing-speed theory of cognitive aging which assumes that the major factor contributing to the negative age-related effects in fluid cognition measures is a reduction in the speed with which fundamental cognitive operations can be executed (see also Salthouse, 1991, Salthouse, 1994, Salthouse, 2000). Accordingly, this generalized slowing has a detrimental effect not only on the quantitative but also on the qualitative dimension of performance for a variety of cognitive skills. For this author, cognitive performance is degraded because relevant basic cognitive operations are executed too slowly and hence, slowed processing reduces the amount of simultaneously available information needed for higher level processing (see Salthouse, 1996). Recent neuroimaging studies have proposed that a disruption of white matter integrity in the whole brain may be at the core of this generalized slowing (see Gunning-Dixon et al., 2009, Penke et al., 2010, Salami et al., 2011; but see also Kennedy & Raz, 2009 for clear dissociations between specific cognitive functions and specific white matter regions). Indeed, age-related differences in white matter integrity have been found to influence different measures of cognitive functioning, particularly processing speed (see Madden, Bennett, & Song, 2009 for a review). Since Salthouse seminal work, numerous studies have supported the processing-speed theory and have shown that controlling for processing speed, strongly reduces or even eliminates age differences in such domains as memory (Bryan and Luszcz, 1996, Clarys et al., 2002), general intelligence and reasoning (Hertzog and Bleckley, 2001, Zimprich and Martin, 2002), or spatial abilities (Finkel, McArdle, Reynolds, & Pedersen, 2007; see Verhaeghen & Salthouse, 1997 for a review). Taken together, these results offer a solid foundation for the global approach and for the slowing of processing speed as the key responsible of cognitive aging (see Deary, Johnson, & Starr, 2010 and Rabbitt et al., 2007 for recent development and interpretation).

At the same time, another theoretical model of cognitive aging, the prefrontal-executive theory, came from different lines of research and was formalized by West (1996), following the work of Dempster (1992). The rationale for this model is that EF, which are thought to involve higher-order functions of control and coordination of more basic or fundamental cognitive operations (see below) and the anterior brain structures (frontal lobes), are especially sensitive to the effects of normal aging (see Phillips & Henry, 2008). Neuropsychological and neuroimaging studies have shown that an important substrate for executive processes are the frontal lobes, which are the cerebral structure the most vulnerable to the advancing in age (see Raz and Rodrigue, 2006, Raz et al., 2007, West, 1996). Moreover, Duncan and colleagues (2000) have shown that “general intelligence derives from a specific frontal system important in the control of diverse forms of behavior” (p. 457). Given the important age-related declines in these functions as well as their underpinning cerebral structures, the hypothesis of prefrontal-executive decline appears to be also a good candidate for explaining the age effects on cognition. Numerous studies have then documented important age-related decrements in EF (Fisk and Sharp, 2004, Jurado and Rosselli, 2007; see Phillips & Henry, 2008 for a review) and their mediating role in the decline due to normal aging in memory (Clarys et al., 2009, Parkin, 1997), strategy and meta-cognition (Bouazzaoui et al., 2010, Taconnat et al., 2006) or activities of daily living (Vaughan & Giovanello, 2010).

Theory-dedicated lines of research have provided empirical evidence for the processing-speed theory and to the prefrontal-executive theory. A few studies have examined, in the same pool of data, the validity of both speed of information processing and EF as mediators of normal aging on various cognitive domains. However, the results of these studies are remarkably mixed. Clarys and collaborators (Clarys et al., 2002, Clarys et al., 2007) and Parkin and Java (2000) found that processing speed emerges as a more fundamental mediator of age-related differences than executive functioning in a number of episodic memory measures. Furthermore, some authors argued against the reality of EF as a distinct construct relevant to the aging literature (Parkin, 1998, Salthouse, 2005, Salthouse et al., 2003). For instance, Salthouse and coworkers have proposed that EF may in fact represent a combination of more well-established cognitive abilities, such as reasoning and perceptual speed. Others showed that controlling processing speed eliminates age differences in central executive functioning (Fisk and Sharp, 2004, Fisk and Warr, 1996). However, almost as many studies reported that executive functioning is a more fundamental factor mediating age-related differences in episodic memory than processing speed (Baudouin et al., 2009, Bugaiska et al., 2007) or showed that age-related executive decline persists beyond the slowing in processing speed (Keys & White, 2000). Additionally, some studies have shown that processing speed and prefrontal-executive functions are independent contributors to the negative effects of aging on fluid intelligence (Bugg et al., 2006, Charlton et al., 2008, Schretlen et al., 2000).

It is very difficult to draw conclusions from the available research for at least three reasons. First, the great heterogeneity of tasks used to evaluate either processing speed or EF makes the comparison of the results very difficult. Second, there is a lack of clear consensus on the theoretical framework concerning the construct of executive functions and the way to evaluate them as well as processing speed. Third, and most important, the shared commonality between the two constructs has not been adequately studied.

A limitation of past studies is that they used very different tasks or lack a clear theory-based rationale for task selection. Concerning the EF construct, some studies used only one task to infer executive performance; e.g., typically a multi-component neuropsychological test such as the Wisconsin Card Sorting Test (WCST, Bugaiska et al., 2007, Schretlen et al., 2000) or the Tower of Hanoi test (Bugg et al., 2006). Others used the same task to evaluate different executive processes; for instance the Stroop task both for inhibition (Salthouse et al., 2003, Vaughan and Giovanello, 2010) and flexibility (Charlton et al., 2008). These examples point out the need for a clear theoretically-driven framework to explicate the impact of EF on cognitive declines. Despite an extensive neuropsychological literature, the construct of EF has been difficult to operationalize. Executive functions refer to a set of cognitive processes whose principal function is to facilitate behavioral adaptation to new or complex situations, in particular when routines are inappropriate. These are thought to encompass formulation of goal and implementing of strategy, planning, action sequencing and monitoring, mental flexibility, inhibition and updating of working memory (see Bryan and Luszcz, 2000, Jurado and Rosselli, 2007, Rabbitt, 1997 for reviews). It appears thus that EF could be characterized by different fundamental executive processes that traditional neuropsychological tests are unable to clearly separate. Here we adopt the theoretical framework developed by Miyake and coworkers (2000) that proposes that EF is fractionated, with different tasks tapping executive processes that are at least partially independent, yet correlated enough to be a unique construct. Miyake et al. propose a three factor structure. These authors have shown in a college-students population, that EF can be subdivided into three separate components: Inhibition of prepotent responses, Updating of working memory, and mental set Shifting (Miyake et al., 2000). Few studies have examined whether or not this organization of EF is age-independent. Two studies replicated Miyake and collaborators’ three-factor model (Fisk and Sharp, 2004, Vaughan and Giovanello, 2010), whereas another identified only two factors including Shifting and Updating but not Inhibition (Hull, Martin, Beier, Lane, & Hamilton, 2008). Again, the number and the nature of the tasks used to assess the different EF components seem to be of critical importance to understand this discrepancy (see Hull et al., 2008 for a discussion).

A limit concerning the EF construct is this problem of task impurity, inherent to the definition of EF in itself. This refers to the fact that it is very difficult to measure directly an EF because it is operating on other cognitive processes (Miyake et al., 2000, Rabbitt, 1997). Accordingly, to address these limitations in the present study, we used nine tasks for assessing the three postulated EF components: Inhibition, Shifting and Updating. By computing a composite score for each EF component, from three different experimental tasks tapping this specific component, we sought a more robust and reliable measure of performance reflecting more selectively the postulated component than each individual experimental task.

The construct of processing speed is much more clearly defined that the one of EF. However, many of the tasks used in past studies have not been very well controlled and often reflect more than pure processing speed. For example, numerous authors (Bryan and Luszcz, 1996, Charlton et al., 2008, Clarys et al., 2007, Parkin and Java, 2000, Salthouse et al., 2003) have used the Digit Symbol Substitution Task (DSST) from the Wechsler adult intelligence scale (WAIS, Wechsler, 1981, Wechsler, 1997) as a measure of processing speed. However, this task is known to involve many other processes than just processing speed; e.g., working memory, visuo-manual coordination, and intelligence (see Parkin and Java, 2000, Piccinin and Rabbitt, 1999, Salthouse, 1992). Moreover, Baudouin et al. (2009) recently showed that the DSST actually involves processing speed and executive processes, particularly in the elderly. These results may explain why DSST performance appears often as the best mediator of age-related cognitive declines. But, as a consequence, it is not possible to isolate the roles of processing-speed and EF in this mediation. Salthouse, 1991, Salthouse, 1994, Salthouse, 1996, Salthouse, 2000 has presented extensive theoretical, methodological, and empirical work on processing speed. He isolated motor speed (with minimal cognitive requirements and reflecting primarily the speed of sensory and motor processes) and perceptual speed (involving various types of cognitive operations, such as comparison, substitution, or transformation, in addition to sensory and motor processes). He argued that perceptual speed is more important than motor speed as a mediator of cognitive aging because it is the speed of central operations that primarily contributes to the relations between age and measures of cognitive functioning (Salthouse, 1994, Salthouse, 1996). However, it is difficult to clearly control what the tasks used to evaluate these constructs (often paper and pencil tasks) really measure in terms of information processing, because they include both motor and cognitive processes, and they do not allow to fully control for the various types of cognitive operations involved. As an illustration of this problem, Salthouse (1994) admitted that motor speed and perceptual speed constructs are not really distinguishable and strongly correlated (i.e., r = .93).

One limit concerning the power of the processing-speed theory is then how processing speed is defined and experimentally measured. Again, to overcome this limit, we think it is important to use a well-defined theoretical framework and well-controlled experimental tasks. The amount of information given by a stimulus and the rate to process it, measured by the Reaction Time (RT), was formalized mathematically by the so-called Hick-Hyman’s law (Hick, 1952, Hyman, 1953).1 Within this framework, we can precisely control the quantity of information one has to process during an experimental RT task. In the present study, we used three different measures of processing speed; a simple auditory RT task (SRT), assumed to measure the speed of only basic sensory and motor processes (the stimulus and the required response are known and prepared in advance); a stimulus-response (S-R) compatible visual 2-choice RT (CRT) task, assumed to measure the speed to process 1 bit of information (see footnote 1); and the DSST from the WAIS III (Wechsler, 1997), assumed to involve both executive and speed processes. Note that, among these three tasks, only the CRT and the SRT allow to clearly quantify the amount of information to be processed (in the case of a 2-choice RT task, 1 bit of information). Concerning the SRT task, one has merely to detect the onset of the stimulus (because the nature of the stimulus, as well as the required response, are known in advance), and thus less than 1 bit of information is processed. Finally, the DSST does not clearly allow to quantify the amount of information to be processed because its performance involves various sensory, motor and cognitive processes, which are difficult to isolate in this paper and pencil task.

The objective of this study was to examine the relationships between measures used to assess two theoretical constructs, processing speed and EF. More specifically, using a theoretically-driven framework, we examined the mutual variance shared by these different measures of processing speed and prefrontal-executive functioning.

Section snippets

Participants

Twenty-eight young adults (11 men, 17 women) from 18 to 32 years of age (M = 22.7 ± 3.3 years) and 39 older adults (17 men, 22 women) from 65 to 80 years (M = 71.2 ± 4.4 years), free of any motor, cardiovascular, or neurological disease, volunteered for this study. Chi-square tests revealed that the proportion of men and women did not differ between the two age groups (p > .05). They all had normal or corrected visual acuity and were in good health according to their personal physician. Only the older

Age-related differences in executive tasks and processing speed measures

The results of the MANCOVAs revealed that Age group had a significant negative effect on the inhibition scores (Wilks’ lambda = .672; F(3, 62) = 10. 1, p < .0001; ηp2=.328), on the updating scores (Wilks’ lambda = .515; F(3, 62) = 19.4, p < .0001; ηp2=485), and on the shifting scores (Wilks’ lambda = 427; F(3, 62) = 26.8, p < .0001; ηp2=.573). All the subsequent univariate ANCOVAs conducted on each of the executive tasks revealed a significant effect of Age group (all ps < .005).

The results of the MANCOVA concerning

Discussion

The principal purpose of the current study was to examine relationships between different measures that have been used to support two competing theoretical constructs that account for cognitive aging: processing speed and executive functions. Several measures were used to evaluate, in samples of young and older adults, three EF components identified by Miyake and collaborators’ factorial structure (2000). The contribution of processing speed obtained from three different measures was also

Acknowledgments

This work was supported by grants from Région Poitou-Charentes and the University of Poitiers. The authors greatly thank Prof. Phillip D. Tomporowski for his valuables comments on an earlier version of this manuscript.

References (77)

  • A. Miyake et al.

    The unity and diversity of executive functions and their contributions to complex frontal lobe tasks: A latent variable analysis

    Cognitive Psychology

    (2000)
  • N. Raz et al.

    Differential aging of the brain: Patterns, cognitive correlates and modifiers

    Neuroscience and Biobehavioral Reviews

    (2006)
  • T.A. Salthouse

    Aging and measures of processing speed

    Biological Psychology

    (2000)
  • C. Albinet et al.

    Aging and concurrent task performance. Cognitive demand and motor control

    Educational Gerontology

    (2006)
  • C.T. Albinet et al.

    Increased heart rate variability and executive performance after aerobic training in the elderly

    European Journal of Applied Physiology

    (2010)
  • A. Allport

    Selection for action: Some behavioral and neurophysiological considerations of attention and action

  • A. Baddeley

    Random Generation and the Executive Control of Working Memory

    Quarterly Journal of Experimental Psychology: Section A

    (1998)
  • P.B. Baltes et al.

    Emergence of a powerful connection between sensory and cognitive functions across the adult life span: A new window to the study of cognitive aging?

    Psychology and Aging

    (1997)
  • E.A. Berg

    A simple objective technique for measuring flexibility in thinking

    The Journal of General Psychology

    (1948)
  • J. Beringer

    ERTS: A flexible software tool for developing and running psychological reaction time experiments on IBM PCs

    Behavior Research Methods

    (1994)
  • J. Bryan et al.

    Speed of information processing as a mediator between age and free-recall performance

    Psychology and Aging

    (1996)
  • J. Bryan et al.

    Measurement of executive function: Considerations for detecting adult age differences

    Journal of Clinical and Experimental Neuropsychology

    (2000)
  • R. Cabeza et al.

    Cognitive neuroscience of aging: Linking cognitive and cerebral aging

    (2005)
  • D. Clarys et al.

    Ageing, remembering, and executive function

    Memory

    (2009)
  • D. Clarys et al.

    Aging and episodic memory: Contributions of executive function and processing speed

    L’Annee Psychologique

    (2007)
  • F. Collette

    Exploration des fonctions exécutives par imagerie cérébrale

  • F. Collette et al.

    Exploring the unity and diversity of the neural substrates of executive functioning

    Human Brain Mapping

    (2005)
  • I.J. Deary et al.

    Are processing speed tasks biomarkers of cognitive aging?

    Psychology and Aging

    (2010)
  • Deltour, J. J. (1993). Echelle de vocabulaire de Mill Hill de J.C. Raven. Adaptation française et normes européennes du...
  • J. Duncan et al.

    A neural basis for general intelligence

    Science

    (2000)
  • D. Finkel et al.

    Age changes in processing speed as a leading indicator of cognitive aging

    Psychology and Aging

    (2007)
  • J.E. Fisk et al.

    Age-related impairment in executive functioning: Updating, inhibition, shifting, and access

    Journal of Clinical and Experimental Neuropsychology

    (2004)
  • J.E. Fisk et al.

    Age and working memory: The role of perceptual speed, the central executive, and the phonological loop

    Psychology and Aging

    (1996)
  • F.M. Gunning-Dixon et al.

    Aging of cerebral white matter: A review of MRI findings

    International Journal of Geriatric Psychiatry

    (2009)
  • D. Head et al.

    Neuroanatomical and cognitive mediators of age-related differences in episodic memory

    Neuropsychology

    (2008)
  • W.E. Hick

    On the rate of gain of information

    Quarterly Journal of Experimental Psychology

    (1952)
  • R. Hull et al.

    Executive function in older adults: A structural equation modeling approach

    Neuropsychology

    (2008)
  • R. Hyman

    Stimulus information as a determinant of reaction time

    Journal of Experimental Psychology

    (1953)
  • Cited by (140)

    • Post-stroke Depressive Symptoms and Cognitive Performances: A Network Analysis

      2023, Archives of Physical Medicine and Rehabilitation
    • Action-speech and gesture-speech integration in younger and older adults: An event-related potential study

      2022, Journal of Neurolinguistics
      Citation Excerpt :

      As mentioned earlier, online (transient) information processing (Salthouse, 1994), updating (Zuber et al., 2019) and revising (Wlotko et al., 2010) is challenged in older adults. In addition, processing speed deteriorates as a result of cognitive aging (Albinet et al., 2012; Kerchner et al., 2012). Because the object was present one 1 s the critical stimulus, participants had plenty of time to already process part of the information related to the critical stimulus.

    View all citing articles on Scopus
    View full text