Concurrent validity of the Swedish version of the life-space assessment questionnaire

In later life, the risk of experiencing limited mobility is increased [1], with limited activity in daily life and restricted participation in social life as a consequence [2, 3]. This risk is evident in Sweden and other Western countries despite general good accessibility in public environments and in housing [4], in cities as well as more rural areas. In order to guide and evaluate mobility-related interventions, it is essential to gain more knowledge about older people’s mobility, and factors related to mobility limitations. For these purposes valid instruments to measure mobility are vital. Most measurements of mobility used for clinical and research purposes focus on performance at a specific point of time or other aspects such as mobility-related tiredness [4]; mobility limited to walking [5]; mobility as an aspect of personal or instrumental Activities of Daily Living (ADL) [6]; or environmental aspects influencing personal mobility [7]. Unlike these instruments, the Life-Space Assessment (LSA), developed in the USA, is an instrument focusing on mobility with respect to reaching different areas defined as life-spaces extending from the room where the person sleeps, to mobility outside one’s hometown [8, 9]. In addition, this instrument goes beyond measuring the mobility a person is capable of a certain day by considering attained life-spaces during the previous month, The LSA also takes into consideration the frequency and independence of a person’s mobility [10, 11]. The LSA has been used to measure mobility in different populations, for example community-dwelling older adults [12], people with eye diseases [13], wheelchair users [1], power mobility device users [14], older people after hospitalization [15], and palliative care patients [16]. The questionnaire has previously been translated, e.g. into French-Canadian [14], Finnish [17, 18], Japanese [19], Spanish [20, 21], and Portuguese versions [20].

Versions translated to native languages are needed to ensure standardised measurements across settings. These versions need to be tested to determine the reliability and validity of the measurements. A newly translated Swedish version of The LSA has been tested for test-retest reliability, with the ratings consistent between the two test occasions for all four LSA scoring methods (total LSA, independent, assistive and maximal life-space scores) [22].

The validity of the original version [23] as well as the Portuguese and Spanish versions [20] has been demonstrated, but the validity of the Swedish version of the LSA is yet to be tested. The purpose of the present study was to examine the concurrent validity of the Swedish version of LSA (LSA-S), by comparing LSA-scores to other measures of mobility.

In the present study the concurrent validity of the Swedish version of the LSA (LSA-S) was examined by comparing the LSA-scores to other measures of mobility. Most correlations between the different LSA measures and the other mobility-related variables were large (above 0.50). Thus, in terms of concurrent validity, thus LSA was shown to be a valid measure of mobility in a Swedish setting among community-dwelling older persons.

Measures reflecting aspects of mobility that presumably could be related to the LSA were chosen. In fact, as far as we know, there is no golden standard described in the literature that could be used to assess concurrent validity related to assessment of life-space mobility. This could be considered a limitation in the present study. However, previous studies on the validity of the LSA have, similarly to our study, used measures of physical performance, including SPPB [20], and ADL [20], or have focused on associations between physical activity and life-space mobility [18]. In addition, our study also considered participation in community activities since that may reflect life-space mobility. Similar to our study, these previous studies found evidence for validity of the LSA by observing significant correlations between the LSA and other mobility-related variables. In fact, most correlations between the different the LSA measures and the other mobility-related variables were also large (above 0.50) in previous studies [20]. Moreover, similar levels of positive correlations were found between physical activity and life-space mobility in a recent study [18].

As previously shown [10], correlations between the different LSA scores were large overall with respect to LSA total, independent and assistive life-space (0.78–0.84), but lower between maximal life-space and the other three LSA scores (0.45–0.58). This may be due to the ceiling effect of maximal life-space found in this study, with most of the study participants reaching the highest score possible. Maximal life-space includes compensation by both persons and equipment, whereas the other LSA scores only consider the persons’ own ability, or in the case of assistive life-space, with the assistance of equipment but without help of another person. With the help of both another person and equipment the person’s own ability is highly compensated, potentially creating this ceiling effect of the maximal life-space score. Thus maximal life-space includes somewhat different aspects of mobility, for example social support [10], than LSA total and independent as well as assistive life-space, which may explain the low correlations. Low correlations were also found between maximal life-space and the other mobility-related variables. In addition to the narrow range of values on the maximal life-space in this study sample, these low correlations may also be due to the fact that several of the measures (unlike maximal life-space), focused on independent performance. Overall, independent life-space probably best reflects the chosen mobility-related variables compared to the other LSA scores. Not surprisingly, this was especially true for “SPPB total score”, “transfers” and “transportation”. In fact, the two last (“transfers” and “transportation”), were based on independent performance.

The mobility-related variables that were analysed in relation to LSA in this study included different aspects of mobility, ranging from more functional aspects, as in SPPB, through more activity-related issues, as in “stair climbing” and the other ADL activities, to also include aspects of participation as in the “community activities” variable. If the environment is supportive, as in Swedish and other Western context, and other alternatives are available, stair climbing is not a prerequisite for community mobility, and this may offer one explanation for the lower correlations with the LSA for that variable. Generally low correlations were found related to “transfers”, potentially due to the fact that values were observed in only the two highest categories of this variable in our sample. It is worth mentioning that low values for this variable may indicate substantial mobility problems, but such problems were not evident in this sample of community-dwelling older adults. Generally, SPPB and “transportation” achieved the highest correlations with the LSA scores. Both SPPB and “transportation” are likely to reflect functional and activity-related factors that support or indicate life-space mobility, while some of the other variables (“travel for fun” and “community activities”) not only indicate life-space mobility but also elements of participation such as motivation, engagement and opportunities.

Comparing the LSA total scores with other mobility-related variables indicated consistency between measures, i.e., higher means of LSA total score also meant an increase in ability as measured by other mobility-related variables. Similar to the previous study on the LSA test-retest reliability [22] we recommend using the total score since that seems to give the most complete picture of mobility compared to the three separate life-space levels.

Finally, it may be considered a limitation that not all variables, i.e., “transfers”, “transportation” “food shopping”, “travel for pleasure” and “community activities”, were part of a standardised test. However, these variables have been used in similar surveys by the research group before, and found to work well in similar contexts. “Travel for pleasure” and “community activities” were based on self-reported, time-based frequency of participation (daily to never). The items of “transfers”, “transportation” and “food shopping” were self-rated by the participants using the ordinal scale (1 = not able/2 = large problem/3 = some problem/4 = no problem). The participants were not given any further definition of these values, so their rating were subjective. This may be considered a limitation, but could be motivated by the argument that mobility is more than moving from A to B and thus not only affected by objective conditions, but also influenced by previous and current subjective feelings and individual experiences [25]. It should also be noted that the variable of “transportation” includes walking as well as transport mobility including various vehicles.

Another limitation relates to the rather high levels of especially assistive and maximum life-space indicating a ceiling effect of these measures, or similar to many other gerontological studies, that we have identified a rather healthy sample.

LSA was shown to be a valid measure of mobility when using the LSA total, independent LS or assistive LSA.