Mobility and cognition in patients with cognitive impairment and Alzheimer’s disease

Abstract

Background: The number of people with cognitive impairment and dementia is increasing worldwide. This development will put considerable strain on health services, and it will be increasingly important to help people with dementia to live safely at home for as long as possible. One of the main threats to independent living is impairments in abilities related to mobility, such as balance, walking, and spatial navigation. These abilities are scarcely studied in clinical practice. Increased knowledge of the character of such mobility impairments in patients with mild cognitive impairment and dementia of various severity may help in the development of interventions aimed at postponing the loss of mobility skills. Aims: The overall aims of this thesis were (1) to explore how mobility performance differs between groups with different levels of cognitive impairment (papers II and III) and how mobility changes over one year in people with early onset Alzheimer’s disease (AD) (paper I), and (2) to explore the relationship between mobility and different domains of cognitive function (papers II and III). Methods: This thesis includes three papers which all apply a cross-sectional design. We also applied a one-year longitudinal follow-up in paper I. The samples in the three studies mainly consist of patients recruited from memory clinics in Malmø, Sweden (paper I) and in Oslo, Norway (papers II and III). Paper I included 72 patients with early onset dementia in the cross-sectional part, and 25 of the 42 patients with early onset AD also attended a one-year follow-up. Paper II included 170 patients in three groups: 33 with subjective cognitive impairment (SCI) or mild cognitive impairment (MCI), 99 with mild AD, and 38 with moderate AD. In paper III, we used a subsample of 128 patients from the sample included in paper II: 19 with SCI, 20 with MCI, and 89 with mild AD. Outcomes: In all three papers, we used performance-based measures of mobility. In paper I, we used timed measures of several mobility tasks. In paper II, we used the Balance Evaluation Systems Test (BESTest) to assess the various aspects of balance. In paper III, we used the Floor Maze Test to assess spatial navigation during walking. The cognitive domains were in papers II and III assessed using tests from the test battery in the Norwegian Dementia Register. Statistical analyses: Between-group differences were analyzed using different versions of analysis of variance (papers I and II), Chi-square tests (papers I and III) and the Mann- Whitney U-test (paper III). The relationship between the mobility measures and the cognitive domains were analyzed using multiple regression analysis (papers II and III). Changes over time were analyzed using the paired samples t-test (paper I). Results: Paper I: Patients with early onset AD had inferior mobility performance compared to patients with other forms of early onset dementia. The performance of the patients with early onset AD deteriorated from baseline to the one-year follow-up. Papers II and III: We found between-group differences (based on severity of cognitive impairment) in all aspects of balance measured by the BESTest and also in spatial navigation measured by the Floor Maze Test. The worst performance was in the group with the most pronounced cognitive impairment. Executive function was the only cognitive domain independently associated with all the mobility outcomes in the multivariate models. Conclusions: We found differences between each of the groups in all aspects of balance (paper II) and also in spatial navigation (paper III). In our longitudinal study of early onset AD we also saw a small decline in mobility over one year (paper I). Although our findings from these cross-sectional indicate a decline in mobility through the stages from SCI to MCI, mild AD, and moderate AD, these findings need to be confirmed in longitudinal studies. With regard to the second aim, executive function was associated with all aspects of balance, and also with spatial navigation. The explained variances were generally high in the models of the aspects of balance. However, for spatial navigation, the models provided only minor explained variances. Future studies are needed to validate the Floor Maze Test against reallife navigation.