Abstract
Aging involves changes in psychomotor performance. Few studies are focused on psychomotor skills among older people due, in part, to the inexistence of valid instruments in the field. The purpose of this article is to analyze the factor structure model of the Portuguese version of Exámen Géronto-Psychomoteur. The confirmatory factor analysis was completed in a sample of 497 older persons, aged between 60 and 99 years, with and without dementia (74.4% female; M = 78.0; standard deviation = 8.6). A baseline one-factor model was compared against 2 three-factor models (first and second order) that were developed based on the previous exploratory factor analysis. Fit indices for the one-factor model were slightly higher when compared with other models; however, the second-order model seems to be more representative of human behavior. The results of this study provide evidence to support a three-factor model: cognition, motor function, and physical aspects.
The life expectancy of the world population and more specifically the Portuguese population is increasing, and health professionals are faced with new symptomatology in the field of geriatrics (Michel, Roux, Albaret, & Soppelsa, 2009). According to the indicators for 2008 of the National Institute of Statistics of Portugal (Instituto Nacional de Estatística, 2009), life expectancy at birth in Portugal increased by 77.41 years in 2002 to 78.70 years in 2008. In addition, according to these indicators, there are currently about 2 million people aged 65 years or older, and it is expected that by 2050 there will be over 3.5 million.
Oliveira et al. (2008) highlighted the aging profile of the Portuguese population, where approximately 21.4% of the elderly population has a functional dependence, 43.8% have a risk of falling, and 68.4% have an unfavorable level physical activity.
The healthy aging is composed by different biomarkers that involve capacities such as physical capability, cognitive, physiological, and musculoskeletal functions (Lara et al., 2015). However, the prevalence of chronic disease increases with age and life expectancy of today’s society, with a possible increase in morbidity and psychosocial disorders (Sawatzky, Liu-Ambrose, Miller, & Marra, 2007). Otherwise, variables such as level of education and income seem to be protective of health decline (Blake, 2014). The geriatric assessment must be complete (by both direct observation and application of scales and questionnaires), covering functional status, mental health and social functioning, and dimensions that are interconnected and influence each other (Paixão & Reichenheim, 2005). According to P.-L. Lee (2014), it seems that there are connections between the practice of physical activity and the cognitive function, being that, in its own study, the persons who had more frequency of physical activity at leisure time and had more cognitive activities showed better executive functions.
The new paradigm of supports provision, for adequate persons-centered plans, favors the specific assessment of older people, especially with dementia (Guitard, Basse, & Albaret, 2005; Martin & Albaret, 2013). Early detection of cognitive deficits and functional decline is critical because of the increased risk of developing dementia (Bartfay, Bartfay, & Gorey, 2014). The studies seem to indicate the need for evaluation in several areas—visuospatial capacity, executive functions, attention, and execution speed—so that an early diagnosis of cognitive deficits can be established, and as the evaluation of the memory alone is not enough (Johnson, Storandt, Morris, & Galvin, 2009). Psychomotor evaluation should focus on different processes and mechanisms that affects movement behavior and should be framed within a valid theoretical model evolving psychomotor and cognitive abilities in an ecological–dynamic approach of psychomotor development and performance (Martin & Albaret, 2013). This evaluation precedes psychomotor therapy that integrates interventions based on body action and mediation, aiming the development of cognitive, emotional, and behavioral functions (Probst, Knapen, Poot, & Vancampfort, 2010). Based on an integral view of the human being, in the psychiatric context, psychomotor therapy is a complement and support to pharmacological treatment (Probst et al., 2010). The benefits of psychomotor therapy in the promotion of healthy aging or as a rehabilitation tool in pathological aging have been reported, but it is fundamental to present scientific evidence of the effectiveness of this intervention, especially regarding a primary intervention focused on dementia. Studies have reported that early treatment of dementia is associated with improved quality of life, better performance of daily living activities, and greater patient safety (Bartfay et al., 2014), but it is important to understand the role of evaluation and psychomotor intervention in the screening and therapeutic intervention in neurological and psychiatric disorders such as dementia.
The assessment related with psychomotor skills, in the field of older people, has been carried out by applying a set of existing tests that assess different areas (Cerioli, 2009; Hugonot-Diener, Barbeau, Michel, Thomas-Antérion, & Robert, 2010; e.g., Mini Mental State Examination for the cognitive state; Tinetti Scale for static and dynamic balance; Rey Complex Figure Test for visual memory and constructive practice, etc.). Adaptations of psychomotor batteries for children are also organized, but they are not validated for the aged population (with and without pathology).
In 2011 appears a French instrument, the Examen Géronto-Psychomoteur (EGP; Michel, Soppelsa, & Albaret, 2011), for assessing psychomotor skills of people older than 60 years. The EGP, conceptually, concerns motor skills, perception and cognition, and verbal level and nonverbal communication, considering the connection with memory (Michel et al., 2011). This is the first instrument developed and adapted for older people within psychomotor therapy field in Portugal. Besides the need of such validated instrument, this study also focuses on the analysis of hierarchical structure of psychomotor skills for older people with dementia. Results will be relevant for more adequate interventions with this subgroup and will enable a comprehensive understanding of how domains are related to each other.
The psychometric properties of Portuguese version of EGP (P-EGP) were tested previously, and three main factors emerged from exploratory factor analysis (EFA): cognitive prevalence, motor prevalence, and physical constraints (Morais, Santos, & Lebre, 2016). Nevertheless, there is no validation study of a multidimensional structure model for the assessment of psychomotor skills of older adults. Furthermore, in Portugal, it is still tradition the use of “mere translations” of others instruments from other countries or adapted versions from children psychomotor batteries. In this study, the confirmatory factor analysis (CFA) was used to add a level of statistical precision in the validation of P-EGP (especially for the construct validity) and to eventually confirm the subdomains founded in EFA (Atkinson et al., 2011). Furthermore, there are no such data available on the original version.
Therefore, our goal in this article is to present the model (or models) behind the assessment of psychomotor competences in older people, obtained through CFA. Based on previous EFA (Morais et al., 2016), we hypothesized a model with three main components: cognitive, motor, and physical aspects. The validation of such instrument, measuring psychomotor skills of older people in Portugal, will also enable its use in clinical and cross-cultural studies. Furthermore, the purpose of this article is also to add the theoretical and empirical development achieved in this field, and at the same time, it tries to bring evidence and knowledge to national and international research in psychomotor therapy area.
Method
Sample
This nonnormative and convenience sample comprised evaluations of 497 participants, with and without dementia, aged between 60 and 99 years (M = 78.0; standard deviation = 8.6). The exclusion criteria were the existence of a motor, sensory, or learning disability as well as a nonstable psychiatric diagnosis. The participants were volunteers or receiving some kind of support of the institutions where the psychomotor therapists were working. In what concerns age and gender distribution, 43% of the participants are aged between 75 and 84 years, and 74.4% were women. More than half of the participants do not exercise regularly (68.6%) and have few nonpharmacological therapies (65.6%). The median cognitive score was 21.3 (measured with Mini-Mental State Examination). In addition, 55.7% of the subjects have a diagnosis of a chronic illness condition different from dementia (e.g., diabetes mellitus, osteoporosis).
Almost half of the sample (48.9%) received care in 20 residential or nursing homes, which provides supports and services for the older adults located on the north, center and south of Portugal metropolitan area of Lisbon and Madeira Island.
Instrument
The EGP is an instrument that assesses psychomotor skills of people older than 60 years and aims to establish an individual psychomotor profile. This instrument may contribute, along with additional medical diagnosis, to guide individualized intervention plans and evaluate the effectiveness of intervention projects, particularly in the area of dementia care (Michel et al., 2011). EGP consists of 17 items that assess the following dimensions: static and dynamic balance, joint mobilizations, praxis, fine motor skills of upper and lower limbs, knowledge of body parts, vigilance, perception, verbal and perceptive memory, spatial and temporal domain, and verbal and nonverbal communication. Each item is scored on a 6-point scale, and most of the items are composed by a set of subitems. The order of application can be modified, with the exception of three items that assess memory (Items 10, 11, and 12) that must be applied in a specific sequence and time interval. A pause is recommended at Item 10.4 if the subject is feeling tired, in pain, or uncomfortable (Michel, Soppelsa, & Albaret, 2010; Michel et al., 2011). EGP also involves clinical observation with reporting qualitative aspects such as posture, problems with balance and walking, functional aspects, the handshake quality or greeting mode, tonic-emotional reactions, the quality of the movement, laterality, coordination/dissociation of upper and lower limbs, tremors, and involuntary movements observed during the application (Michel et al., 2010, 2011).
Procedures
The research protocol was approved by the ethical board of a research center. All participants provided signed informed consent for participation in this study. Contacts with support providers for older people were contacted to collaborate in the research.
The P-EGP was applied by psychomotor therapists specifically trained on how to administer and score each item in accordance with ethical principles, importance of psychomotor evaluation, and the scale administration guidelines. A previous training was conducted by the researchers to all applicants. In this training, special attention was given to the quotation of the scale items in order to reduce the occurrence of human error. Moreover, most evaluations were conducted by the lead author of this article. In addition, sociodemographic data and medical history such as age, education level, morbidity, support needs, therapies, and physical activity routine were collected. Data collection took place between January 2013 and September 2014 in therapeutic rooms or offices (with the elimination of stimuli when necessary). Each application had an average duration of 50 to 70 minutes.
Statistical analyses were undertaken using the Statistical Package for Social Sciences (IBM Corp. Released, 2012), version 20.0 and AMOS 22.0.
Data Analysis
The appropriateness of the model was assessed through a variety of absolute and incremental goodness-of-fit indexes (GFIs; see “Results” section).
Based on the hypothetical underlying constructs for the psychomotor assessment in older people, three models were developed to represent the best fit for the overall data. In Model 1, one-factor model was used as a baseline comparison against the other models. Model 2 was a first-order CFA model testing aimed to verify the model found in the EFA and assumed that all three factors are correlated with each other. Once the first-order CFA model testing was confirmed, Model 3 was further tested: A second-order CFA model to see whether the three factors could be explained by a broader general factor, which we defined as “psychomotor performance.”
Model quality of fitness was evaluated using the chi-square (χ2/df ), comparative fit index (CFI), GFI, adjusted GFI (AGFI), normed fit index (NFI), and root mean square error of approximation (RMSEA; Arbuckle, 2007).
The model was considered to have acceptable fit if χ2/df was lower than 5; CFI, GFI, and AGFI were higher than 0.8; NFI was higher than 0.9; and the RMSEA was lower than 0.08 (Arbuckle, 2007; Hooper, Coughlan, & Mullen, 2008; Hu & Bentler, 1999). Model adjustment was performed, step-by-step, via Modification Indices analysis (higher than 11; p < .001) and based on theory (Arbuckle, 2007). To complete the CFA estimating the construct reliability, composite reliability and the average variance extracted were computed. The cutoff points for these indices are 0.70 and 0.50, respectively (Herrera-López, Gómez-Ortiz, Ortega-Ruiz, Jolliffe, & Romera, 2017).
Translation and Adaptation of EGP and Psychometric Properties
The description of procedures related with the translation and adaptation of EGP for the Portuguese context are already described in a previous paper (Morais et al., 2016). In addition, in this previous study, the results related with preliminary psychometric properties analysis were found to be moderate to excellent indexes for content validity (Morais et al., 2016). The instrument demonstrated excellent internal consistency (all domain reliabilities exceeding .90; Morais et al., 2016). The total score of P-EGP (α = .92) was superior to the original study (α = .83; Michel et al., 2011). A strong stability was also found as the Pearson’s values were all above .60, and the coefficient for total score of P-EGP was .97 (Morais et al., 2016). Finally, three main components from EFA were obtained: Physical Constraints, Motor Prevalence, and Cognitive Prevalence. The two domains that constitute the factor “Physical Constraints” had weakly correlations with many other EGP domains and appeared to be a factor with specific characteristics.
Results
The goal of this article is to present the model (or models) behind the assessment of psychomotor competences in older people using P-EGP.
Confirmatory Factor Analysis
We used a CFA, using maximum likelihood to estimate model parameters, to determine the fit of the three models. For Models 2 and 3, all items presented factor loadings above .40, ranging from .57 (item static balance I) to .90 (item time domain).
Model 3 is presented in Figure 1. According to fit indices (Table 1), all models have acceptable fit.
Model 3 (second-order model). The measurement model variables (e.g., static balance I; fine motor upper limbs) are presented in squares; the latent variables (e.g., Psychomotor Performance) are presented in oval. The numbers on arrows between variables are coefficients. SB I = static balance I; SB II = static balance II; DB I = dynamic balance I; DB II = dynamic balance II; JMUL = joint mobilization upper limbs; JMLL = joint mobilization lower limbs; FMUL = fine motor upper limbs; FMLL = fine motor lower limbs; PR = praxis; KBP = knowledge of body parts; V = vigilance; PM = perceptive memory; SD = space domain; VM = verbal memory; P = perception; TD = time domain; C = communication; EGP = Examen Géronto-Psychomoteur.
Fit Indices From Confirmatory Factor Analysis Models.
Note. df = degree of freedom; CFI = comparative fit index; GFI = goodness-of-fit index; AGFI = adjusted GFI; NFI = normed fit index; RMSEA = root mean square error of approximation.
Apparently, Model 1 seems to have the better adjustment, as it has the lower χ2/df and RMSEA and the higher CFI, GFI, AGFI, and NFI. Nevertheless, Models 2 and 3 have stronger regression weights for each domain (>0.50) and lower standard errors of the coefficients (Table 2). The domains with lower scores for Model 1 (static balance I, dynamic balances I and II, and joint mobilization of upper and lower limbs) have much better results when attributed to a single factor. In addition, Model 3 seems more “clean,” as it presents just two covariances (e1–e3 and e10–e11) to make this model fit. Model 3 also has good values of composite reliability (0.77) and average variance extracted (0.54).
Standardized Regression Weights and Factor Correlations by Model and Latent Construct.
Note. EGP = Examen Géronto-Psychomoteur.
aStandardized factor coefficients and the standard errors of the coefficients. Entries marked with (–) were constrained at a raw factor coefficient of 1.0 and thus yielded no standard error estimates.
Discussion
This analysis of validity of the P-EGP, as far as we are concerned, is the first study on this matter and extends the knowledge in the literature. Having a valid instrument to evaluate psychomotor skills in older people with dementia will guide more adequate elderly person-centered plan. P-EGP has the adequate psychometric properties to assess the psychomotor competences of older people (Morais et al., 2016), contributing to improved care in this population, namely, supporting psychomotor therapists in dementia early signs detection, and in the guidance for the development of individualized plan and person-centered planning strategies in geriatrics area (Michel et al., 2011). However, and due to the inexistence of such procedures, there was a need to reinforce the construct validity of this instrument using a confirmatory method. This study intended to present the model (or models) behind the assessment of psychomotor competences in older people. Statistic scores showed empirical support of construct validity, and findings allow credibility to improve psychomotor competence on Portuguese older adults with dementia.
Based on previous results (Morais et al., 2016), we hypothesized that a three-factor model would be more adjusted for the psychomotor assessment in geriatrics. We tested three models—one factor, three-factor first order, and three-factor second order—and all models have acceptable adjustment fit indexes. Surprisingly, Model 1 (one factor) has slightly better scores for these criteria, which may be due to the perspective of the psychomotor assessment as a process that includes a comprehensive and holistic view of the person, in which cognitive and motor domains are closely interrelated (Michel et al., 2011). In addition, pathologies like dementia are possible to observe concomitantly with a decline in cognitive abilities, changes in emotional control, and social behavior (Bergh & Selbæk, 2012), which affect the functional capacity (Helvik, Engedal, Benth, & Selbæk, 2014). The high prevalence of dementia in our sample could have created a distorting effect on the fit indices of the models. Although in theoretical terms psychomotor symptoms or skills involve the development of perceptual and cognitive processes (Beheydt et al., 2015; Reuben et al., 2013), recent research continues to study cognitive and motor aspects separately, especially in pathologies like dementia (Boyle, Cohen, Paul, Moser, & Gordon, 2002). In addition, we found that frequently psychomotor skills are assessed by the reaction time to a task (Beheydt et al., 2015) and not the process itself or the time of the movement to achieve a goal. However, these studies often conclude that psychomotor competences evolve diverse cognitive aspects, and it is very difficult to separate these areas (Beheydt et al., 2015). These authors state that an interdisciplinary approach of psychomotor retardation in older adults depressed seems to be in order. Varalta et al. (2015) founded significant association between balance skills and executive functions in patients with Parkinson’s disease. In the point of view of the therapy with older people, many studies refer the benefits of physical activity on maintaining cognitive function in late life or preventing a cognitive decline (P.-L. Lee, 2014; Y. Lee et al., 2015).
What is the purpose of separating cognitive and motor aspects in the study of psychomotor competences? Can these aspects be divided, in the psychomotor performance? If we look carefully to Models 2 and 3, we can see that each EGP domain have better regression weights when attributed to a factor (cognitive, motor, or physical). In a neurobiological perspective, we know that cognitive performance have different neural substract than motor performance (e.g., the executive functions are supported by prefrontal and frontal areas of the brain; Poranen-Clark et al., 2015), although in reality most of our actions involve both areas simultaneously (Probst et al., 2010), including perceptual processes. Systematically, science has been showing us that combined cognitive and motor assessment or treatment has better predictive values for several pathologies and prevents/treats better many situations such as fall risk (Schoene et al., 2015), dementia (Ho, Cheung, Chan, Cheung, & Lam, 2015), with a greater impact on daily functioning (Schoene et al., 2015).
Lara et al. (2015) proposed a panel of biomarkers of healthy aging and, although they divide them into physical capability, cognitive, physiological and musculoskeletal, and endocrine and immune functions, they assume that combinations of these biomarkers may have a better predictive value for biological age and for the rate of aging among adults. For example, in the creation of National Institutes of Health Toolbox for Assessment of Neurological and Behavioral Function (Reuben et al., 2013), the authors separated the assessment of motor function from other domains. However, it is mentioned that some of motor functions (e.g., dexterity) are associated with cognitive functions (e.g., visual memory), and most of them are correlated with performance in daily living in older adults.
So, according to the most recent studies, Models 2 and 3 seem to be more representative of human function, as they preserve a different “identity” for cognitive, motor, and physical behavior but maintain interrelations between the factors. In addition, in Model 3, we can say that there is a Psychomotor Performance “construct” that joins three areas—Cognition, Motor Function, and Physical Aspects.
Model 3 seems to respect professional and scientific purposes, as it allows tracing cognitive, motor, and physical profiles for the patient, but also represents the holistic point of view of the human being and the person as a unity of body and mind through the concept of psychomotor performance. Based on previous studies (Morais et al., 2015), we found that most of P-EGP domains are correlated not only with mental state but also with performance in basic and instrumental activities of daily living. This is so important because the main goal of psychomotor therapy in geriatrics is to maintain the functional autonomy of the person, and so, more research for a deeper analysis of the relations between psychomotor competences, cognitive status, and functional performance, in older adults with and without dementia is suggested.
In this study, the psychomotor competences of all participants (with and without dementia) were analyzed as a group, assuming that the P-EGP model will behave in the same manner for both populations. However, in a future study, we will present a multigroup structural analysis in order to investigate whether the three models from the CFA are invariant across the presence or nonpresence of dementia.
Conclusion
The results of this study showed that fit indices for the one-factor model were slightly superior when compared with the other models; however, the second-order model seems to be more representative of human behavior. The results of this study provide evidence to support a three-factor model for the assessment of psychomotor competences, and Portuguese version will retain the original structure. As a convenience sample was used, further research will need to confirm results with older adults in different stages of dementia. P-EGP assumes an important role for evaluation and planning purposes and may be used in clinical settings and situations as well in future research for measuring the effectiveness of psychomotor interventions (Guitard et al., 2005).
This is one of the few studies that use CFA in psychomotor assessment instruments and in the elderly population, with the validation of a scale that has already shown high correlations with other countries (e.g., Uruguay). The results founded encourage further research in this area and enable comparative analyses between different countries. At the same time, they allow monitoring of psychomotor interventions.
Footnotes
Acknowledgments
The authors would like to thank Ana Antunes for her valuable help on inserting data and all the colleagues who applied Portuguese version of Examen Géronto-Psychomoteur. Nothing would have been possible without their cooperation and excellent work. The authors would also like to thank the board of directors and the technical staff from the elderly care services where the data collection took place for welcoming the researchers of this study.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.