Development of action representation during adolescence

https://doi.org/10.1016/j.neuropsychologia.2006.07.010Get rights and content

Abstract

During adolescence the body undergoes many physical changes. These changes necessitate an updating of internal models of action. Here, we tested the hypothesis that internal models undergo refinement between adolescence and adulthood. We investigated the chronometry of executed and imagined hand actions, which relies on internal models, in 40 adolescents (24 males; mean age 13.1 years) and 33 adults (15 males; mean age 27.5 years). In two different motor imagery tasks, the time it took each participant to execute a hand movement was compared with the time it took them to imagine making that movement. For all participants, movement execution time significantly correlated with movement imagery time. However, there was a significant increase in the execution–imagery time correlation between adolescence and adulthood. Cognitive-motor efficiency per se did not change as indexed by both similar execution and imagery times on both tasks for the adolescents and adults. That it was only the correlation between imagined and executed actions that changed with age suggests that the developmental change was specific to generating accurate motor images and not a result of general cognitive improvement with age. The results support the notion that aspects of internal models are refined during adolescence. We suggest that this refinement may be facilitated by the development of parietal cortex during adolescence.

Introduction

During adolescence, the body is subject to height, weight, organ and musculoskeletal development (Coleman & Hendry, 1989). Representations of the body and its kinematics might therefore be expected to change with age. This kind of “action representation” is a component of an internal forward model, which is a neural system that simulates the dynamic behaviour of the body in relation to the environment (Wolpert, Ghahramani, & Jordan, 1995). It has been proposed that these internal models make predictions about actions, limb kinematics and parameters of the external world and enable successful planning and execution of movement (Wolpert, 1997). Optimal motor control is thought to depend on an internal forward model making predictions of the consequences of the movement, based on an efference copy of the motor command (von Holst, 1954, Wolpert et al., 1995, Wolpert, 1997). Discrepancies between predicted states and desired states result in error signals that can be used to fine-tune actions and to provide the motor instructions required by the muscles to achieve the desired effect. Internal models are constantly updated based on the actions and experiences of the person in the world (Miall & Wolpert, 1996; Wolpert et al., 1995). As body kinematics change during adolescence, the representation and prediction of actions made by internal models require updating.

It has been posited that by studying conscious motor imagery, it is possible to access the unconscious process of action representation (Jeannerod, 1997). Motor imagery, which is thought to involve the activation of internal models of action, can be considered a first-person process of the participant feeling herself executing an action. A motor image is therefore a conscious equivalent to a prediction for that action (Jeannerod, 1994, Jeannerod, 1997). Recently, Voss et al. have shown that internal model prediction occurs even in the absence of movement (Voss, Ingram, Haggard, & Wolpert, 2006). Impairment in internal models occurs in patients with parietal cortex lesions (Sirigu et al., 1996; Wolpert, Goodbody, & Husain, 1998) as well as in people with schizophrenia (Maruff, Wilson, & Currie, 2003). Together with neuroimaging studies (Gerardin et al., 2000; Lacourse, Orr, Cramer, & Cohen, 2005; Stephan et al., 1995), the results suggest that motor imagery is associated with activity in parietal cortex. This is supported by the proposal that internal models are stored in parietal cortex (Blakemore & Sirigu, 2003; Wolpert et al., 1998).

Psychophysical experiments have shown that there are parallels between the parameters affecting actual movements and imagined movements. In particular, the time course of imagined and executed actions is highly correlated in normal adults (e.g. Decety & Jeannerod, 1995; Sirigu et al., 1995, Sirigu et al., 1996). The temporal invariance between executed and imagined movements suggests that the same motor representation (internal model) governs an action whether it is executed or imagined, and time constraints operate in the same way in both modalities of action. A close association between execution and imagery time for a given action, therefore, is an indicator of the ability to represent that action. In patients with parietal lesions, however, this timing relationship is lost (Sirigu et al., 1996, Maruff et al., 2003, Wolpert et al., 1998).

Studies using grip force modulation paradigms have shown that age four to age six is a critical time for the development of internal models (Blank et al., 1999, Blank et al., 2000; Paré & Dugas, 1999). To our knowledge, there have been no studies to date on the development of internal models or action imagery beyond childhood and none has studied the development during adolescence. Perhaps because research on the development of the adolescent brain is relatively new, mechanisms of cognitive function during adolescence remain poorly understood. Recent MRI data has demonstrated that that the parietal cortex, which is associated with internal models of action, undergoes a particularly protracted course of development compared with sensory regions of the brain such as the visual cortex (Giedd et al., 1999, Gogtay et al., 2004). In light of previous histological studies, these MRI data are thought to reflect synaptic pruning, the elimination of unused synapses, and a simultaneous process of myelination of developing fibre tracts occur during this period, stabilising in early adulthood (Huttenlocher, 1979; Yakovlev & Lecours, 1967).

Given the association of internal models with parietal cortex, is it logical to expect that parietal development might affect internal models. To investigate the hypothesis that internal models undergo a period of development during adolescence, we tested the ability of 40 young adolescents and a control group of 33 adults to use motor imagery. As an index of the development of the action representation system, we measured the correspondence between the time course of every participant's executed (E) and imagined (I) actions in each age group, on two different motor imagery tasks. To compare this correspondence between age groups, we compared the execution–imagery (E–I) correlations of each individual in the two age groups. To determine whether developmental change was specific to the development of internal models and not a result of general cognitive-motor improvement, the change in E–I correlation with age was compared to the change in general cognitive-motor efficiency with age.

Section snippets

Participants

Seventy-three participants were recruited and divided into two age categories. Forty adolescents (24 males; mean age 13.1 years, S.D. = 1.4) and 33 adults (15 males; mean age 27.5 years, S.D. = 7.9) took part in the study. Adolescent participants were from state comprehensive primary and secondary schools in the London area and adults were students and staff at University College London. Participants were all right handed and none had a history of psychiatric, neurological, developmental or

FPIQ

All participants scored highly on the FPIQ, and there were no differences between groups in the mean total scores or in any of the subscales (mean score (adolescents) = 24; mean score (adults) = 26 out of a total of 28; see Table 1). A participant's imagery and action data were excluded from the analysis if his or her mean FPIQ score fell three or more standard deviations below the mean score from his or her age group. One adolescent's data fell into this category. This analysis indicated that all

Discussion

The results of the current study demonstrate a tight correlation between action execution (E) and imagery (I) times in both adolescents and adults for two different imagery tasks: the fingers task and the triple 8 task. However, for both tasks, there was a significant increase in the degree of correspondence between E and I between adolescence and adulthood. To our knowledge, this is the first study to demonstrate that although adolescents are able to form motor images, this ability improves

References (42)

  • E.R. Sowell et al.

    Localizing age-related changes in brain structure between childhood and adolescence using statistical parametric mapping

    Neuroimage

    (1999)
  • J.A. Stevens

    Interference effects demonstrate distinct roles for visual and motor imagery during the mental representation of human action

    Cognition

    (2005)
  • A.W. Toga et al.

    Mapping brain maturation

    Trends in Cognitive Neurosciences

    (2006)
  • E. von Holst

    Relations between the central nervous system and the peripheral organs

    British Journal of Animal Behaviour

    (1954)
  • P. Viviani et al.

    The relationship between linear extent and velocity in drawing movements

    Neuroscience

    (1983)
  • P.H. Wilson et al.

    Abnormalities of motor and praxis imagery in children with DCD

    Human Movement Science

    (2001)
  • D.M. Wolpert

    Computational approaches to motor control

    Trends in Cognitive Sciences

    (1997)
  • A. Binet

    La psychologie du raisonnement

    (1885)
  • S.-J. Blakemore et al.

    Action prediction in the cerebellum and in the parietal lobe

    Experimental Brain Research

    (2003)
  • C.J. Coleman et al.

    The nature of adolescence

    (1989)
  • P.M. Fitts

    The informed capacity of the human motor system in controlling the amplitude of movements

    Journal of Experimental Psychology

    (1954)
  • Cited by (0)

    View full text