Life-space mobility assessment in older people in Finland; measurement properties in winter and spring

Life-space mobility refers to the size of the spatial area (bedroom, home, outside home, neighborhood, town, and beyond town) that an individual purposely moves through in daily life, and to the frequency of movement within a specific time and the need for assistance[1]. Thus, it reflects the actual performance of mobility activities in daily life. Importantly, life-space mobility has been associated with quality of life in older people[2].

Life-space mobility can be studied using the University of Alabama at Birmingham Study of Aging Life-Space Assessment[1]. Previous research has established the validity and reliability of the scale among relatively healthy older people in the USA[1, 3, 4], South American[5] populations and in powered wheel chair users in Canada[6]. It has been argued that the measurement properties of a scale are context-specific[7–9]. Thus, sample characteristics (e.g. age and health status) and for example contextual factors, such as climate and physical environment, may affect life-space mobility and consequently also the accuracy and reliability of the assessment. The present study was conducted in Finland, a Nordic country that has a climate including long, cold winters and warm summers, and rapid changes in temperature[10]. Furthermore, Finland has extensive rural areas and towns that may cover large surface areas[11].

In older people the most common physical activity is walking[12, 13]. Life-space mobility is a measure that captures the outdoor activity of an individual, irrespective of the mode of transportation[1, 14], Even when using motorized transportation, walking is a prerequisite for moving in larger life-space areas. A previous study demonstrated a relationship between physical activity and life-space mobility[4], and another study used life-space mobility as an indicator of physical activity[15]. Previous studies have shown that season and weather circumstances affect the level of physical activity and walking, even in countries with a temperate climate (for review see[16, 17]). Extremely low or high outdoor temperatures[18–23], shorter daylight time[19, 24], fewer hours of sunshine[19, 20], rainfall[20–22, 24], higher wind speed[20, 22], and snow and icy conditions[22, 25] have been identified as barriers for physical activity and walking behavior. Most previous studies have focused on children, adolescents and working-aged adults (for review see[16, 17]). Studies among older people have shown similar associations and some researchers have suggested that the effects of weather may be more pronounced among older people[19, 24–26]. Still, whether life-space mobility is affected by season or weather has not been studied before.

The aim of the present study was to investigate selected measurement properties of the Life-Space Assessment in community-dwelling older people in Finland. We studied the distribution, one-year responsiveness and two-week reproducibility of the scale items and composite scores. Considering the potential impact of season and outdoor temperature on outdoor behavior patterns, we also conducted stratified analyses for season (winter and spring).

This study assesses the measurement properties of the Life-Space Assessment in a Nordic country and is the first study specifically accounting for the effects of season. The test-retest reproducibility of the Life-Space Assessment was fair, however, it seemed to be somewhat compromised in winter conditions, with average temperatures below zero and snow on the ground. In addition, the scoring of items at higher life-space levels was more variable, which may reflect the daily reality of community-dwelling older people. Further, self-reported decline in health or mobility status and age over 85 were associated with larger decreases in life-space mobility score over the one-year follow-up period, thus indicating that the Life-Space Assessment was responsive to changes.

In accordance with previous studies[1, 5], life-space mobility score had a normal distribution, while the indicators of maximal, assisted and independent life-space and the life-space dependency score had ceiling effects. Also we found that the life-space mobility score and assisted life-space were responsive to change[1] as indicated by a larger change in participants reporting declines in health or mobility status over one year. The overall change in life-space mobility score related to declines in health or mobility was relatively small (1.5-2.5 points) in these community-dwelling older people compared to other studies. In clinical samples changes of 22-23 points have been reported due to surgery-related hospital admissions[30] or gynecological oncology[31], changes of nine points for non-surgical hospital admissions[30], and changes of twelve points for urogynecological surgery[31], respectively. On theoretical basis, Baker et al.[1] suggested that a change of ≥10 points can be considered clinically meaningful. In our population-based sample, about half of the participants reported such a decrease or increase between baseline and the one-year follow-up. This percentage is similar to the 6-month change reported by Baker et al.[1].

While individual change in life-space mobility and other life-space measures over the year varied, the median change was zero. It is known that community-dwelling older people experience different trajectories of declines and recovery in function[32]. A previous study of our group showed that life-space mobility scores in each 5-year age-group of advancing age were on average 10 points lower[33], the annual change in the present study did not reach the level needed to be considered clinically meaningful[1] in any age-group. However, the decrease in life-space mobility score accelerated in those aged 85 years and above. Further, individual decreases and increases of more than 10 points occurred in all age-groups. Consequently, recovery from restrictions in the life-space seems possible even in those above 85 years.

In general, the test-retest reproducibility reported in the present study was comparable to that reported by Curcio et al.[5] and Auger et al.[6], but lower than that reported by Baker et al.[1]. Also, the reproducibility of the life-space mobility scores was somewhat higher than the reproducibility of the other indicators of life-space[1, 5, 6]. However, in the two-week period between the test and retest, about one-third of the participants experienced a clinically meaningful change of ≥10 points[1] in positive or negative direction, which could not be attributed to a self-reported change in their situation. Similarly as for other questionnaires that assess behavior, such as self-report questionnaires on physical activity, some variation will be due to the normal variation in daily life[34]. More specifically concerning life-space mobility, previous studies have suggested that trips over longer distances occur occasionally. In older people over the age of 65 long-distance trips may occur even less than once a month[35] and may thus not always be captured within the time interval of the Life-Space Assessment.

One possible explanation of the fact that the agreement between the test and retest was slightly poorer in the winter sample for the majority of measures could be that extreme temperatures and climatologic circumstances (such as slippery roads) in the winter may affect outdoor mobility of older people[20, 22, 24, 25]. Although in our study, the 4-weeks average temperatures for the test and retest were not different, in the winter the variation in life-space mobility may increase due to snowfall or icy conditions[22, 25]. Unfortunately, such data were not available in our study.

In accordance with our expectations that were based on previous literature on physical activity and walking[19, 20, 22, 24, 26, 36], life-space mobility scores were somewhat higher in spring. In spring, we observed some ambiguity in defining the “home” of some individuals for which life-space was assessed. Many Finnish people own second homes or summer cottages often in another town, and they may live in their city home in winter and all summer in their cottage in a rural region[37, 38]. In the present study, we defined as the “home” of the participant their primary home in case of frequent visits for few days or a week at the time, and their secondary home in case of long-term stay (e.g. whole summer). The rating of life-space was experienced as difficult particularly in periods of transition between the different “homes”, which occur most often in the spring (and fall)[37, 38]. Furthermore the use of a second home most likely affects the frequency of movement. This may be reflected in the finding that the variability in frequency at the “town” and “beyond town” level was larger in the spring sample than in the winter. In the test-retest sample two participants moved from the winter home to the summer cottage between the test and the retest, reflecting the daily reality of the LISPE study participants. Naturally, this may also affect longitudinal changes for example between baseline and follow-up data. Administering follow-up interviews on a date as close as possible to the date of a baseline interview, may not guarantee comparable circumstances since climatologic circumstances may differ regardless. Therefore we suggest that climatologic circumstances are an important factor to consider in future longitudinal studies on outdoor mobility behavior of older individuals, and that data collection should include variables able to capture such information.

Anecdotally, in the present study the interviewers and participants identified certain other difficulties when completing the Life-Space Assessment that may explain part of the variation in scoring. For example, defining the neighborhood and town area appeared sometimes difficult. In Finland, in the past decades many towns have been merged to cover extensive areas. Without structurally providing a clear-cut definition of the distance related to the town area, participants may have applied different interpretations of what area to relate to when responding. Moreover, participants living close to town borders found the use of the term “town” confusing. Further, the definition of the neighborhood was also dependent on interpretation, particularly when living in a rural area. While a distance was occasionally provided in case of ambiguity, as suggested by Baker et al.[1] and Auger et al.[6], it is unknown how other participants have interpreted the life-space areas. Consequently, inter-individual differences in interpretation may have contributed to the variation demonstrated in the results, particularly at higher life-space levels. In the original publication by Baker et al.[1], the distance perceived as neighborhood and town remained constant among study participants. The consensus distance that defined the town area[1] reported in the Life-Space Assessment manual (16 km; 10 miles) is much shorter than the distance potentially travelled within study areas such as Jyväskylä municipality in Finland. Further study is needed to determine whether the distances related to the respective life-space levels are valid in different national contexts, for example in the Nordic countries. Although this requires further study, we speculate that clear-cut definitions of the distances routinely stated in conjunction with each life-space level should be considered in future studies.

Limitations of our study include that our population-based sample consisted of relatively healthy older people, which reduced the variation in life-space mobility at the lower life-space levels. In addition, when interpreting the results of the reproducibility study, it should be taken into account that we used a relatively small convenience sample. In the responsiveness analyses, selective loss to follow-up due to declining health – a common issue in aging research – may have led to underestimation of the decrease in life-space mobility. Finally, inter-individual differences in the interpretation of the life-space areas and intra-individual transitions between different homes may have slightly contributed to the variation in results.

Testing the Life-Space Assessment in a Nordic country demonstrated that the life-space mobility score was responsive to change. The test-retest reproducibility was fair but subject to seasonal variation, that is, the reproducibility was somewhat compromised in winter conditions. Further study is warranted to obtain insight in the factors contributing to seasonal effects in life-space mobility. The present study also showed that the mobility of older people at life-space levels of “town” and “beyond town” may be variable. We speculate that this likely reflects the real life situation rather than results from a reproducibility problem, but should however, be confirmed in further studies. Finally, researchers should be aware of the challenges when using the Life-Space Assessment in different contexts. Structurally defining distances for the different life-space levels may be helpful to improve reproducibility, but this requires further study.