Investigating the Mini‐BESTest's construct validity in elderly with Parkinson's disease
The Mini‐BESTest has recently emerged as a promising clinical balance test. Reflecting the construct of dynamic balance,3 the Mini‐BESTest consists of items relevant to the specific balance impairments among PwPD.1 However, despite having been presented as a unidimensional balance test3 (suggesting that only the total score should be considered), the fact that the latest version of the Mini‐BESTest4 clearly defines four subcomponents (anticipatory postural adjustments, reactive postural control, sensory orientation, and dynamic gait) may suggest the test to be multidimensional. Accordingly, recent studies have investigated the scores of these components separately.5 Although such investigations appear appealing, it is difficult to evaluate the relevance of their outcomes as instructions regarding how to approach these subcomponents are lacking.
Most studies on the psychometric properties of the Mini‐BESTest have examined criterion validity,8 “the degree to which the score of an instrument is an adequate measure of a gold standard,”11 by treating other common balance tests as the gold standard. However, none of these tests12 have been developed to measure the construct of dynamic balance.3 In fact, no gold standard seems to exist for the measurement of dynamic balance. When a gold standard is lacking for the construct to be investigated, the more appropriate approach is to study construct validity.14 This can be performed by formulating hypotheses regarding: (i) expected differences between subgroups of individuals (known‐groups validity) or (ii) expected associations between instruments assessing theoretically related (convergent validity) or unrelated (divergent validity) constructs.11
Previous research on known‐groups validity has shown that the Mini‐BESTest was able to discriminate between PwPD with mild and severe motor severity.9 Others have found that the Mini‐BESTest's total score was significantly lower (i.e., worse) among recurrent fallers compared to non‐recurrent fallers in people with mild‐to‐severe PD.7 However, to our knowledge, no study has investigated the Mini‐BESTest's ability to discriminate between people with and without PD; people with mild and moderate severity of PD; nor between people with mild‐to‐moderate PD with and without a history of recurrent falls. In our view, the ability for any clinical tool to separate between people with different capabilities of the construct to be measured is not only an important aspect of validity, but is also vital for its clinical utility.
This study aimed to investigate the Mini‐BESTest's construct validity. In particular, we hypothesized that the Mini‐BESTest scores would be significantly lower (i.e., worse) among: (i) PwPD compared to healthy controls, (ii) PwPD with moderate compared to mild motor severity, and (iii) PwPD with a history of recurrent falls compared to non‐recurrent falls.
In addition, for convergent validity, we expected a moderate‐to‐strong correlation between the total score of the Mini‐BESTest and the timed up and go (TUG), a test developed to measure the construct of physical mobility. For divergent validity, we expected a poor correlation between the scores of the Mini‐BESTest and a questionnaire measuring activities of daily living in PwPD (part II of the Unified Parkinson's Disease Rating Scale; UPDRS, part II). Unlike the TUG (and the Mini‐BESTest), the UPDRS, part II, is based on the participants' own perceptions of their abilities regarding a variety of situations related to activities of daily living.