Our objective was to study the impact of an extensive range of socio-demographic, health-related and psychosocial variables on levels of fear of falling and avoidance of activity in older persons who avoid activity due to fear of falling. The results clearly show that due to the interrelationship between socio-demographic, health-related and psychosocial variables in older people, only a few of them correlated independently with either severe fear of falling and avoidance of activity due to this fear. Although we identified four socio-demographic, five health-related and five psychosocial factors significant univariate correlates of fear of falling, the independent correlation of only female sex, limitations in activity of daily living and falls history remained when covariates were taken into account. With respect to avoidance of activity due to fear of falling, we identified two socio-demographic, five health-related and five psychosocial significant univariate correlates; this pattern is consistent with fear of falling except for low education and living alone status being non-significant. When all significant univariate correlates were taken into account, only higher age and limitations in activity of daily living correlated independently with severe avoidance of activity. The correspondence between correlates for fear of falling and avoidance of activity may partly be explained by the interrelation of both outcomes. In effect, we were trying to determine what influences avoidance of activity in addition to fear of falling.
In a recent review Scheffer and colleagues  identified falls history, female sex and higher age as the main risk factors of fear of falling. In addition, decline in physical and mental functioning and quality of life as well as risk of falling were identified as the main consequences of fear of falling. These conclusions are partly similar to our findings in Table 2 when we compared older persons with and without severe fear of falling. However, in our final multivariate model where we took all significant univariate associations into account, only the impact of age, limitations in activity of daily living and falls history remained significant in the model. This latter finding indicates that none of the psychosocial factors we studied can differentiate between severe and mild levels of fear of falling.
Our study has several limitations. First, all measures were self-reported and performance-based measures were not included in the current study. Future studies should include objective measures of balance, gait, muscle strength, vision, etc. to establish a further understanding of the complex construct of fear of falling. Second, our study was cross-sectional so we may not take causal inferences from our data (e.g., between avoidance of activity and limitations in activity of daily living). Future research may identify which correlates can be considered as predictors and which as consequences of severe fear of falling and related avoidance of activity. Third, we included community-living older persons which can be considered as a heterogeneous sample. Future studies could focus on whether different patterns of correlates exist between more frail and less frail older persons. Finally, we included only persons who reported at least mild fear of falling and avoidance of activity. Also having people without fear of falling in our study would have enabled us to say whether, for example, women are more likely to develop any level of fear of falling or only severe fear of falling. However, our interest was not in the development of fear of falling, but rather in those factors that differentiate between mild and severe levels of fear and related avoidance behavior. From clinical point of view particularly older people with sever fear of falling and avoidance of activity due to fear of falling are relevant for interventions. Therefore it is important to identify the factors associated with severe levels of fear of falling and related avoidance behavior. Such knowledge may help to specify the contents, approach and length of interventions for these persons. So from clinical point of view it is relevant and important to distinguish between mild and severe levels of fear of falling.
The strength of our study is that in contrast to previous studies we focused on severe levels of fear of falling and avoidance of activity. We identified old age, female sex, limitations in activity of daily living, impaired vision, poor perceived health, chronic morbidity, falls, low general self-efficacy, low mastery, loneliness, feelings of anxiety and symptoms of depression as significant correlates of severe fear of falling and avoidance of activity. Although the knowledge about the unique associations of specific variables with levels of fear of falling is of interest for theoretical and scientific reasons, the awareness of the above univariate associations may help to specify the concepts for developing interventions and programmes to reduce severe levels fear of falling in old age, particularly in their early stages of development. A further strength of our study was the attempt to separate fear of falling and the resultant avoidance of activity as outcomes; particularly the latter is under researched in previous studies. This is important, as while fear of falling in itself may be an unwelcome anxiety for an older person, it is the related avoidance of activity that is arguably more significant in terms of its effect on social engagement, functional decline, quality of life, and further falls, particularly when this avoidance is severe. Our work suggests that avoidance of activity in older people with severe levels of fear of falling may be particularly high in those of advanced age and with limitations in activities of daily living, people for whom continued independence may be especially fragile and who will need rapid intervention to maintain that independence.