Next Article in Journal
A Multilevel Analysis of Support for Immigrants’ Social Rights in Latin America
Previous Article in Journal
Reducing Youth Incarceration: From Trauma-Informed Confinement to Community-Based Services
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Social Sciences
Volume 15
Issue 6
10.3390/socsci15060379
socsci-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Predictors of Avoidance Behavior in Fear of Falling Among Older Adults: A Latent Profile Analysis

by
Tatyana K. Konovalchik
1,2 and
Olga Yu. Strizhitskaya
3,*
1
Department of Medical Psychology, Faculty of Psychology, Saint Petersburg State University, Saint Petersburg 199034, Russia
2
Department of Clinical Psychology, Faculty of Applied Psychology, Saint Petersburg State Autonomous Educational Institution of Higher Education “Saint Petersburg State Institute of Psychology and Social Work”, Saint Petersburg 199178, Russia
3
Department of Developmental and Differential Psychology, Faculty of Psychology, Saint Petersburg State University, Saint Petersburg 199034, Russia
*
Author to whom correspondence should be addressed.
Soc. Sci. 2026, 15(6), 379; https://doi.org/10.3390/socsci15060379
Submission received: 28 February 2026 / Revised: 6 June 2026 / Accepted: 8 June 2026 / Published: 10 June 2026
(This article belongs to the Topic Environmental and Social Transformations, Population Ageing, and Sustainable Health Equity)
Versions Notes

Abstract

Objectives: Fear of falling (FoF) is a common psychological phenomenon in later life and is often accompanied by avoidance behavior and activity restriction. Although FoF is associated with anxiety, depressive symptoms, reduced self-efficacy, and fear of loss of autonomy, older adults with FoF may differ substantially in the configuration of these characteristics. The present study aimed to identify data-derived profiles of older adults based on FoF, avoidance behavior, self-efficacy, and fear of loss of autonomy, and to examine profile-specific psychological predictors of FoF and avoidance behavior. Methods: The main analytical sample included 217 older adults aged 60–97 years (M = 76.45, SD = 10.14) with Mini-Mental State Examination scores of 20 or higher. Latent profile analysis was conducted using FoF, avoidance behavior, self-efficacy, and fear of loss of autonomy. Anxiety components, depressive symptoms, coping strategies, pain catastrophizing, and loneliness-related indicators were examined in class-specific regression models. The stability of the class solution was tested across different MMSE cut-off scores. Between-class comparisons were conducted for functional, fall-related, socio-demographic, and psychological indicators. Results: A three-class solution was selected and interpreted as adaptive, vulnerable, and maladaptive profiles. The profile structure remained relatively consistent across MMSE cut-off scores, including in the broader sample with MMSE ≥ 15. The classes did not differ significantly in postural balance or number of falls, suggesting that the profiles could not be fully explained by objective fall-risk indicators. Significant between-class differences were found for age, daily pain level, and state social defense. Class-specific regression models suggested that psychological variables associated with FoF and avoidance behavior differed across profiles. Pain appraisal and emotion-related coping were more relevant in the adaptive profile, phobic anxiety and anxious appraisal of future events in the vulnerable profile, and anxiety-related, depressive, interpersonal, and coping-related factors in the maladaptive profile. All reported associations remained significant after false discovery rate correction. Conclusions: FoF and avoidance behavior are related but not identical phenomena and vary across data-derived psychological profiles. A profile-oriented approach may provide a more differentiated understanding of activity restriction in older adults and help identify profile-specific targets for psychological support.

1. Introduction

Fear of falling is a serious and widespread problem among older adults and may occur in individuals with different levels of objective balance, physical functioning, and fall experience (Tinetti et al. 1990; Tinetti and Powell 1993; Lord et al. 2007). It is considered a factor associated with restrictions in daily activity, reduced autonomy, and lower quality of life (Xiong et al. 2024; Sattler et al. 2025; Even-Zohar et al. 2025). Fear of falling is also regarded as a recognized psychological predictor of fall risk alongside many other factors. At the same time, evidence suggests that fear of falling is only moderately associated with actual falls, whereas its behavioral consequences, primarily activity restriction or avoidance, are more closely related to functional status and level of independence in older adults than to the occurrence of falls itself (Delbaere et al. 2004; Zijlstra et al. 2007). Given the need to improve the quality of care for older adults with fear of falling, this phenomenon requires further in-depth investigation.
Contemporary research describes fear of falling not only as an independent phenomenon, but also as a complex multilevel construct closely associated with emotional, cognitive, motivational, and behavioral characteristics of older adults. A central problem in this field is that older adults with fear of falling are often treated as a relatively homogeneous group, although the same level of fear may be accompanied by different behavioral, cognitive, emotional, and motivational patterns. Some individuals may maintain daily activity despite fear, whereas others may substantially restrict activity even when objective fall-risk indicators are not markedly impaired. Therefore, understanding fear of falling requires attention not only to its overall severity, but also to heterogeneity in the psychological configurations in which fear and avoidance occur. Fear of falling is most commonly examined in relation to anxiety, depression, self-efficacy, fear of loss of autonomy and dependence, coping strategies, and avoidance behavior (Payette et al. 2016; Savvakis et al. 2024; Shimizu et al. 2024).
When fear of falling and avoidance behavior are considered within the framework of Protection Motivation Theory, the formation of a protective behavioral response depends on how older adults subjectively evaluate several factors. Within this framework, behavior is regulated not by fear itself, but by cognitive processes involved in the appraisal of threat and coping ability. An important role is played by subjective threat appraisal, which includes perceptions of the likelihood of falling and the seriousness of its possible consequences. These perceptions are shaped both by environmental features associated with fall risk and by internal concerns and expectations related to the possible loss of physical independence, habitual lifestyle, and social autonomy. Equally important is the appraisal of one’s ability to cope with the threat. This includes beliefs about the effectiveness of protective behavior, confidence in one’s own ability to carry it out, and the subjective costs associated with such actions. According to PMT, maladaptive protective responses may develop when high perceived threat is combined with low perceived coping ability (Rogers 1983; Preissner et al. 2023). In such situations, avoidance may extend beyond adaptive caution and lead to activity restriction even in relatively safe conditions (Denkinger et al. 2015; Auais et al. 2017; Türkmen Keskin et al. 2025; Wu et al. 2023). Avoidance behavior may include a wide range of strategies, from refraining from potentially challenging motor tasks to substantial restriction of life-space mobility and increased dependence on assistance from others (Baek et al. 2024). Once avoidance becomes established, older adults may lose the opportunity to confirm their ability to cope with the situation at the behavioral level, which in turn may intensify fear and contribute to the formation of a vicious circle (Tinetti and Powell 1993). Thus, fear of falling may be viewed as an affective component associated with the choice of behavioral strategies, whereas avoidance represents one of the possible behavioral manifestations of this response.
The diversity of psychological aspects associated with fear of falling complicates the identification of more precise targets for psychological support in older adults. In this regard, it appears important to identify groups differing in their affective, cognitive, and behavioral characteristics, as well as to examine psychological factors associated with the maintenance of fear of falling and avoidance behavior (Yang et al. 2025; Tabacchi et al. 2025; Xiong et al. 2024). Therefore, the present study attempted to identify latent participant profiles that would allow a more meaningful description of the heterogeneous structure of the sample (Kongsted and Nielsen 2017).
Within the framework of Protection Motivation Theory, we assumed that the profiles could be based on indicators reflecting its main components. Accordingly, indicators of subjective intensity of fear of falling, severity of avoidance behavior, fear of loss of autonomy, and self-efficacy were selected as variables for latent profile analysis.
Subjective intensity of fear of falling was selected as the affective core of threat experience and perceived vulnerability. Fear of loss of autonomy reflects the subjective threat associated with the consequences of falling and is closely related to expectations of dependence, loss of control over daily life, habitual social roles, and personal agency, which gives fear of falling a pronounced psychosocial meaning (Payette et al. 2016; Savvakis et al. 2024; Shimizu et al. 2024). Empirical evidence suggests that subjective evaluations of autonomy and satisfaction with independence are associated with the severity of fear of falling. Lower perceived autonomy is associated with higher levels of fear of falling independently of somatic factors (Damar and Barutcu 2022), whereas expectations of losing independent functioning often constitute an important psychological basis of fear of falling (Imani and Godde 2021). Self-efficacy was conceptualized as a component of coping appraisal. It reflects subjective confidence in one’s ability to safely perform daily activities and manage fall-related risk (Bandura 1977). In fear of falling research, self-efficacy is most often operationalized through measures of falls efficacy, reflecting confidence in one’s ability to prevent falls or cope with their consequences during everyday activities (Almajid et al. 2025; Jang et al. 2025; Meyer et al. 2025). Lower self-efficacy has been associated with more pronounced fear of falling, reduced subjective autonomy, and increased avoidance behavior even at comparable levels of physical functioning (Lee and Tak 2023; Lee and Kim 2024; Imani and Godde 2021). Conversely, higher levels of fall-related self-efficacy are associated with reduced fear of falling and improved daily activity, including within exercise-based intervention programs (Li et al. 2005; Preissner et al. 2023). Avoidance behavior was conceptualized as a behavioral manifestation of protective responding that may range from adaptive caution to maladaptive activity restriction.
Thus, based on these indicators, we assumed that older adults may differ not only in the absolute severity of fear of falling or avoidance behavior, but also in the configuration of threat experience, autonomy concerns, perceived coping ability, and protective behavioral response. For this reason, a profile-based approach was used, as it allows the identification of data-derived subgroups characterized by different combinations of these indicators rather than by a single symptom level. This approach made it possible to examine whether fear of falling and avoidance behavior are embedded in different psychological configurations within the sample.
As psychological characteristics associated with fear of falling and avoidance, we selected aspects most frequently examined in contemporary research. At the same time, we aimed to cover not only affective difficulties, primarily anxiety and depression, which are traditionally associated with fear of falling, but also the psychosocial aspect represented by attitudes toward loneliness, as well as the coping-related component represented by emotional, cognitive, and behavioral coping strategies (Konovalchik 2024).
The aim of the study was to examine psychological variables associated with fear of falling and avoidance behavior in older adults while taking into account participant profiles derived from indicators of subjective severity of fear of falling, avoidance behavior, self-efficacy, and fear of loss of autonomy.
Protection Motivation Theory was used as a conceptual framework for selecting the profile indicators and organizing the interpretation of the findings. However, given the cross-sectional design of the study, the present analysis cannot test causal mechanisms of threat appraisal, coping appraisal, or protective motivation. Instead, it allows us to examine whether theoretically relevant indicators form distinct data-derived profiles and whether fear of falling and avoidance behavior show different patterns of association with psychological variables within these profiles.
Research questions
  • Can qualitatively distinct latent profiles be identified based on indicators of fear of falling, avoidance behavior, self-efficacy, and fear of loss of autonomy?
  • What profile-specific patterns of association are observed between fear of falling, avoidance behavior, and psychological factors not used in profile construction, including anxiety, depressive symptoms, attitudes toward loneliness, and coping strategies?

2. Materials and Methods

2.1. Data Collection and Sample

Participants were recruited from the nursing home with medical care “Our Care” (Saint Petersburg, Russia), the residential care facility “Zarya” (Saint Petersburg, Russia), and the Saint Petersburg State Budgetary Residential Social Care Institution “Nursing Home for Older Adults and People with Disabilities No. 1” (Saint Petersburg, Russia) during routine cognitive and clinical assessments. Initial eligibility criteria included age from 60 to 97 years, a Mini-Mental State Examination (MMSE) score of 15 or higher, absence of psychotic conditions, and normal or corrected-to-normal vision and hearing.
At admission, participants were assessed by a geriatrician using the Berg Balance Scale, the Barthel Index, and the MMSE. These assessments were conducted as part of the initial clinical examination and were not administered simultaneously with the main study protocol. The Barthel Index was completed by a specialist and was based on functional assessment rather than self-report.
All assessments within the main study protocol were conducted individually and in person. To reduce measurement burden and fatigue-related response bias, respondents were allowed to interrupt the assessment and rest for as long as they needed before continuing. During the assessment, the examiner could repeat or clarify instructions when respondents had difficulty understanding them, without suggesting specific answers. The Cornell Scale for Depression in Dementia is not a self-report instrument; it was completed by the researcher based on observation, interview with the participant, and information provided by medical staff. Thus, the main self-report battery included eight instruments, with a total of 99 items. On average, completion of the full protocol took approximately 1.5 h, usually with two or three breaks. Incomplete or ambiguously completed forms were excluded from further analysis.
A total of 452 individuals were screened. Of these, 98 declined participation, 79 did not meet the initial inclusion criteria, and 46 provided incomplete or ambiguously completed data. Thus, data were initially collected from 229 respondents with MMSE scores of 15 or higher.
For the purposes of the present study, the main analytical sample was restricted to respondents with MMSE scores of 20 or higher. This resulted in the exclusion of 12 participants with signs of moderate dementia. This decision was based on the use of self-report measures, which require sufficient comprehension of instructions and the ability to evaluate one’s own condition. The full sample of respondents with MMSE scores of 15 or higher was used only to examine the sensitivity of the profile solution to differences in cognitive status. These data were not used for the final latent profile model or for the subsequent regression analyses.
The final analytical sample included 217 participants aged 60–97 years (M = 76.45, SD = 10.14), comprising 40 men (18.4%) and 177 women (81.6%). The mean age was 77.10 years (SD = 10.60) among women and 73.58 years (SD = 7.26) among men. All main results of the latent profile analysis, between-class comparisons, and regression analyses are reported for the sample of respondents with MMSE scores of 20 or higher. Demographic characteristics are presented in Table 1.

2.2. Measures

2.2.1. Screening Methods

Cognitive status was assessed using the Mini-Mental State Examination (MMSE) (Folstein et al. 1975), a widely used screening instrument for detecting cognitive impairment of varying severity. MMSE scores range from 0 to 30, with higher scores indicating better cognitive functioning. The following score ranges were used to interpret cognitive status: 28–30 points indicated normal cognitive functioning, 25–27 points indicated mild cognitive impairment, 20–24 points indicated mild dementia, and scores below 20 indicated moderate dementia. A minimum score of 20 points was required for inclusion in the study, which allowed the exclusion of participants with more pronounced cognitive impairment while retaining respondents capable of understanding instructions and completing self-report measures. In addition, MMSE scores were used in sensitivity analyses to examine whether the identified classes were robust to differences in cognitive status. These analyses were conducted using the lower cut-off scores for each cognitive group, namely 28, 25, and 20 points.

2.2.2. Study Measures

The measures were selected to reflect both the core components of the PMT-based profile model and additional psychological characteristics potentially associated with fear of falling and avoidance behavior. Fear of falling, avoidance behavior, self-efficacy, and fear of loss of autonomy were used as latent profile indicators, whereas depressive symptoms, anxiety components, coping strategies, pain catastrophizing, loneliness-related indicators, functional independence, and postural balance were used to characterize the profiles and examine profile-specific associations.
  • Fear of falling was assessed using a Visual Analogue Scale (VAS) represented by a 10 cm line ranging from 0 (no fear) to 10 (maximum fear). This method has been previously used to assess fear of falling in older adults and has demonstrated adequate sensitivity (Scheffer et al. 2010). Fear levels were classified as low (0–2), moderate (3–6), and high (7–10). The VAS was selected because, at the time of the study, no validated multidimensional instrument for assessing fear of falling was available for the Russian-speaking population; it provided a direct numerical estimate of subjective fear intensity without requiring prior linguistic adaptation of a multi-item scale. Within the PMT framework, this measure was used as an indicator of subjective threat experience and perceived vulnerability to falling.
  • Avoidance behavior related to fear of falling was assessed using the modified questionnaire mFFABQ-R (Landers et al. 2011; Russian adaptation by Konovalchik and Strizhitskaya, forthcoming). The scale consists of 14 items rated on a five-point Likert scale, with total scores ranging from 14 to 70, reflecting the degree of avoidance of everyday activities. Within the PMT framework, avoidance behavior was considered a behavioral manifestation of protective responding to perceived fall-related threat.
  • Depressive symptoms were assessed using the Cornell Scale for Depression in Dementia (Alexopoulos et al. 1988; Levin 2023). The scale includes five domains covering behavioral, emotional, cyclic, ideational, and somatic symptoms of depression in older adults. Scores below 10 indicate absence of depression, scores from 10 to 18 indicate moderate depression, and scores above 18 indicate severe depression. In relation to PMT, depressive symptoms were examined as psychological characteristics that may be associated with reduced coping appraisal, pessimistic self-evaluation, and lower perceived ability to manage fall-related threat.
  • Anxiety structure and anxiety-related traits were measured using the Integrative Anxiety Test (Bizyuk et al. 2005). Results were interpreted using stanine scores: 1–3 stanines indicate low anxiety, 4–6 moderate anxiety, and 7 or higher high anxiety requiring clinical attention. The instrument comprises five subscales identified through factor analysis: emotional discomfort, asthenic component of anxiety, phobic component, anxious evaluation of future prospects, and social defense. Within the PMT framework, anxiety-related indicators were examined as variables reflecting heightened threat appraisal, perceived vulnerability, and anxious anticipation of negative consequences.
  • Self-efficacy was assessed using the General Self-Efficacy Scale (Schwarzer et al. 1996; Russian adaptation by Romek 1997). The scale consists of 10 items rated on a four-point response format, with total scores ranging from 10 to 40 and reflecting subjective confidence in one’s ability to cope with life challenges. Within the PMT framework, self-efficacy was used as an indicator of coping appraisal, reflecting perceived ability to manage difficult or potentially threatening situations.
  • Fear of loss of autonomy was assessed using the corresponding subscale of the Short Questionnaire on Fear of Disease Progression (Sirota and Moskovchenko 2014). This subscale reflects concerns related to loss of ability to independently manage daily activities, increased dependence on others, and restrictions in everyday functioning. Within the PMT framework, this measure was considered an indicator of perceived severity of fall-related consequences, particularly the anticipated loss of independence and personal agency.
  • Coping behavior was assessed using the E. Heim Coping Inventory (Heim 1988; Russian adaptation by Lugovskaya 2017), designed to evaluate strategies for coping with stress and difficult life situations. The instrument is based on a cognitive–behavioral framework and allows identification of stable patterns of responding to psycho-emotional stressors. Its use was motivated by the need for a comprehensive assessment of individual coping strategies related to age-related changes, physical limitations, and anxiety states, including fear of falling, in older and old-old adults. Within the PMT framework, coping strategies were examined as coping-related responses that may be associated with coping appraisal and with patterns of adaptive protective behavior or maladaptive avoidance.
  • Pain catastrophizing was assessed using the Pain Catastrophizing Scale (PCS) (Russian adaptation by Radchikova et al. 2020). The instrument includes 13 items rated on a five-point Likert scale (0–4) and comprises three subscales: Rumination, Magnification, and Helplessness. Total scores range from 0 to 52, with higher values indicating greater catastrophic thinking related to pain. In relation to PMT, pain catastrophizing was examined as a cognitive factor that may intensify threat appraisal by increasing attention to bodily sensations, perceived vulnerability, and expectations of harm.
  • Loneliness was assessed using the Differential Questionnaire on the Experience of Loneliness—Short Version (DOPO-3k) (Osin and Leontiev 2013). The 24-item instrument is rated on a four-point Likert scale and includes three subscales: General loneliness, Dependence on communication, and Positive loneliness. Higher scores reflect greater expression of the corresponding dimension. Within the PMT framework, loneliness-related indicators were examined as psychosocial characteristics that may be associated with perceived coping resources, dependence on interpersonal support, and the subjective costs of managing fall-related threat.
  • Functional independence was assessed using the Barthel Index (Mahoney and Barthel 1965). The scale evaluates independence in 10 basic activities of daily living, with total scores ranging from 0 (complete dependence) to 100 (complete independence). Higher scores indicate greater functional autonomy. In the present study, the Barthel Index was used to characterize the functional status of respondents. Within the PMT framework, this measure provided contextual information on actual functional autonomy, allowing comparison between subjective autonomy concerns and observed independence in daily activities.
  • Postural balance was assessed using the Berg Balance Scale (Berg et al. 1992). The instrument consists of 14 performance-based tasks rated on a five-point scale from 0 to 4, with a maximum total score of 56. Scores below 45 are commonly considered indicative of increased fall risk. In the present study, the Berg Balance Scale was used to characterize respondents’ objective balance performance. Within the PMT framework, this measure provided an objective fall-risk indicator against which subjective threat appraisal and avoidance behavior could be interpreted.

2.2.3. Statistical Analysis

Statistical analyses were conducted using IBM SPSS Statistics version 27 (IBM Corp., Armonk, NY, USA) and Mplus Demo Version (Muthén & Muthén, Los Angeles, CA, USA). IBM SPSS Statistics version 27 was used for descriptive statistics, assessment of distribution normality based on skewness and kurtosis indices, multiple linear regression analysis, and between-class comparisons using the Kruskal–Wallis test. Mplus was used to conduct latent profile analysis and identify data-derived latent classes of respondents.
Latent profile analysis was selected because it is a model-based approach designed to identify unobserved subgroups of participants with similar patterns across several indicators. In the present study, this approach was appropriate because we aimed to examine heterogeneity in the configuration of fear of falling, avoidance behavior, self-efficacy, and fear of loss of autonomy. The profile indicators were standardized before analysis to place them on a comparable metric. Competing models with different numbers of latent classes were estimated and compared using information criteria, classification quality, likelihood ratio tests, profile size, and substantive interpretability. The final class solution was selected by considering statistical fit, parsimony, adequate profile size, and theoretical interpretability within the PMT-based framework.
Regression analyses were exploratory and were conducted separately within each latent profile to identify psychological variables associated with fear of falling and avoidance behavior. Given the profile-specific sample sizes, especially in the maladaptive profile, these models were not intended as confirmatory predictive models. The results were therefore interpreted as preliminary patterns of associations. To reduce the risk of false-positive findings due to multiple testing, false discovery rate correction was applied.

3. Results

3.1. Latent Profile Analysis Procedure

The aim of the latent profile analysis was to identify relatively homogeneous subgroups of older adults differing in the severity of fear of falling, avoidance behavior, fear of loss of autonomy, and level of self-efficacy, followed by an assessment of the clinical and functional relevance of the identified latent classes. The analysis was conducted using data standardized by z-transformation.
The number of latent classes was selected by comparing alternative models. Two-, three-, and four-class models were estimated. The results of model comparison are presented in Table 2.
The number of latent classes was selected by comparing two-, three-, and four-class solutions across several model fit indices. Relative to the two-class model, the three-class solution showed improved fit, as reflected in lower AIC, BIC, and aBIC values. This improvement was also supported by the likelihood ratio tests. Both the LMR-LRT and BLRT were statistically significant for the three-class model, indicating that the three-class solution showed a relatively better fit than the two-class solution.
The four-class model showed a further decrease in the information criteria and also had significant LMR-LRT and BLRT values. However, this improvement was accompanied by a reduction in the size of the smallest class. In the four-class solution, the smallest class included 6.9% of the sample, whereas in the three-class solution the smallest class included 12.0% of respondents. Entropy was high across the tested models, and the difference between the three- and four-class solutions was not sufficient to offset the smaller class size in the four-class model.
Therefore, the three-class model was selected for further analysis. This solution was considered more parsimonious and substantively interpretable because it improved model fit compared with the two-class solution, retained acceptable classification quality, and avoided the very small additional class observed in the four-class model. It also provided a more balanced distribution of respondents across classes than the four-class solution.
The pattern of mean scores for fear of falling, avoidance behavior, self-efficacy, and fear of loss of autonomy within the identified latent classes allowed for a substantive interpretation of the resulting profiles. Class 1 showed the most relatively adaptive pattern of characteristics. It included respondents with low levels of fear of falling, avoidance behavior, and fear of loss of autonomy, along with relatively preserved self-efficacy. Class 2 can be interpreted as a more vulnerable profile. It was characterized by moderately expressed fear of falling, avoidance behavior, and fear of loss of autonomy, together with somewhat reduced self-efficacy. Class 3 generally corresponded to a relatively more maladaptive profile. It combined high fear of falling and markedly elevated avoidance behavior with moderate fear of loss of autonomy and slightly lower self-efficacy compared with the vulnerable profile. The data are presented in Table 3.
To assess the sensitivity of the class solution to differences in respondents’ cognitive status, an additional robustness analysis of the latent profiles was conducted. The analysis was performed sequentially in subsamples derived from the initial sample of 229 respondents by increasing the lower MMSE inclusion threshold. Cut-off scores of 15, 20, 25, and 28 points were used. This approach made it possible to examine whether the class solution remained relatively consistent after the gradual exclusion of participants with more pronounced cognitive impairment.
Sensitivity analysis suggested that the three-class solution was generally consistent across different MMSE cut-off scores. Entropy values remained high in all subsamples and ranged from 0.894 to 0.915, indicating acceptable classification accuracy. The smallest class included more than 10% of the sample in all subsamples, suggesting that the class structure was not driven by the presence of a very small or unstable profile.
The LMR-LRT was statistically significant in all subsamples except the MMSE ≥ 28 subsample. In the MMSE ≥ 28 subsample, the LMR-LRT did not reach statistical significance, which may be related both to the smaller sample size and to reduced variability in the indicators when the analysis was restricted to respondents with more preserved cognitive functioning. At the same time, the BLRT remained statistically significant in all subsamples. Overall, these results suggest relative consistency of the three-class solution across different MMSE cut-off scores. The data are presented in Table 4.
The substantive structure of the classes was also relatively consistent across MMSE cut-off scores. In all subsamples, Class 1 generally corresponded to a relatively more adaptive profile, characterized by lower fear of falling, lower avoidance behavior, lower fear of loss of autonomy, and relatively preserved self-efficacy. Class 2 reflected a more vulnerable profile, with moderately increased fear of falling, avoidance behavior, and fear of loss of autonomy, together with somewhat reduced self-efficacy. Class 3 generally corresponded to a relatively more maladaptive profile, with the highest fear of falling and the most pronounced avoidance behavior. The class sizes and standardized mean scores of the profiling variables across MMSE cut-off scores are presented in Table 5.
The sensitivity analysis suggested that the three-class structure remained relatively consistent when lower MMSE cut-off scores were used. This suggests that the identified profiles were unlikely to be an artifact of excluding respondents with lower cognitive scores. At the same time, given the self-report nature of the measures used in the study, we decided to use the MMSE ≥ 20 subsample for all subsequent analyses. This threshold allowed us to retain respondents with normal cognitive functioning, mild cognitive impairment, and mild dementia, while excluding participants with more pronounced cognitive impairment that could affect comprehension of instructions and the reliability of self-report responses.
These analyses provide within-sample evidence of relative stability of the three-class solution across different cognitive-status thresholds; however, they do not constitute external validation of the profile structure.

3.2. Between-Class Differences in Functional, Socio-Demographic, and Psychological Characteristics

To further characterize the identified profiles, between-class comparisons were conducted for indicators of functional independence, postural balance, and number of falls. The Kruskal–Wallis test did not reveal statistically significant differences between the profiles in the Barthel Index, Berg Balance Scale, or number of falls. These results suggest that the identified latent classes were not clearly differentiated by functional independence, objective balance performance, or fall history. The results are presented in Table 6.
Additional comparisons were also conducted for socio-demographic and psychological characteristics. Significant between-class differences were found for age, daily pain level, and state social defense. The maladaptive profile had the highest mean rank for age, indicating that respondents in this profile tended to be older. The vulnerable profile showed the highest mean rank for daily pain level, whereas the adaptive profile had the lowest mean rank. For state social defense, the highest mean rank was observed in the vulnerable profile and the lowest in the maladaptive profile. No statistically significant differences were found for the remaining socio-demographic and psychological indicators.

3.3. Regression Analysis

After the latent classes had been identified, separate regression models were constructed for each profile. This made it possible to examine which psychological characteristics were associated with fear of falling and avoidance behavior within different groups of respondents.
We first considered the regression model for psychological variables associated with fear of falling, presented in Table 7.
In the adaptive profile, the model explained a small proportion of the variance in fear of falling, 11.4%. Pain magnification, active avoidance, and suppression of emotions were variables significantly associated with fear of falling. Higher pain magnification and greater suppression of emotions were associated with higher fear of falling. At the same time, active avoidance was negatively associated with fear of falling. This finding may indicate that, in the adaptive group, cautious behavior was not necessarily accompanied by higher fear of falling and did not clearly reflect maladaptation.
In the vulnerable profile, the only variable significantly associated with fear of falling was the trait phobic component of anxiety. Higher phobic anxiety was associated with higher fear of falling. The model accounted for a relatively small proportion of variance, 8.4%, which suggests that fear of falling in this group may be related not only to the psychological variables included in the analysis.
In the maladaptive profile, the regression model showed a high R2 value and accounted for 82.3% of the variance in fear of falling. However, given the relatively small size of this subgroup, this estimate should be interpreted cautiously. The identified associations are better regarded as exploratory and hypothesis-generating rather than stable predictive effects. Fear of falling in this profile was associated with trait anxiety, depressive symptoms, dependence on communication, and several coping-related variables.
We then considered regression models for psychological variables associated with avoidance behavior. The detailed results are presented in Table 8.
In the adaptive profile, the model explained a small proportion of the variance in avoidance behavior, 3.6%. The only variable significantly associated with avoidance behavior was pain rumination. Stronger fixation on pain sensations was associated with more pronounced avoidance. This may indicate that even in the relatively adaptive group, bodily discomfort and persistent attention to pain may contribute to activity restriction.
In the vulnerable profile, the model accounted for 41.7% of the variance in avoidance behavior. A positive association was found only for state anxious appraisal of perspective. The more strongly respondents anxiously evaluated future events and possible consequences of their condition, the higher their level of avoidance. The remaining predictors had negative associations with avoidance. More pronounced disturbances of cyclic functions in depression, as well as more frequent use of the coping strategies “cooperation,” “emotional discharge,” and “withdrawal,” were associated with lower avoidance. These findings may be cautiously interpreted as suggesting that, in the vulnerable profile, avoidance is more closely related to anxious anticipation of future difficulties than to the general severity of emotional distress. Negative associations with coping strategies may indicate that when respondents have at least some ways of responding to tension, including turning to others, emotional discharge, or temporary withdrawal, avoidance of everyday activity is less pronounced.
In the maladaptive profile, the model explained 29.8% of the variance in avoidance behavior. The only variable significantly associated with avoidance behavior was ideational disturbances in depression. This indicator reflects ideas of guilt, failure, damage, and pessimistic evaluations of oneself and one’s condition. More pronounced ideational symptoms were associated with higher avoidance. This may indicate that, in the most maladaptive group, avoidance is related not only to fear of falling as such, but also to depressive beliefs about personal inadequacy, a sense of damage, reduced confidence in the ability to cope with the situation, and a negative evaluation of the future.
All reported associations remained statistically significant after false discovery rate correction for multiple testing.

4. Discussion

Contemporary studies of fear of falling in older adults often rely on a linear logic in which fear of falling and avoidance behavior are considered closely related components of the same process (Gagnon et al. 2005; Öztürk and Polat 2026; Adandom et al. 2026; Munir et al. 2026). This approach describes the general association between fear and activity restriction, but may insufficiently account for heterogeneity among older adults. In the present study, latent profile analysis was used to identify groups not on the basis of predefined categories, but according to the structure of the obtained data. Based on fear of falling, avoidance behavior, self-efficacy, and fear of loss of autonomy, three profiles were identified and provisionally labeled as adaptive, vulnerable, and maladaptive. These labels should be regarded as preliminary analytical categories that require further validation.
The data obtained through LPA address the first research question and suggest that older adults with fear of falling may not represent a homogeneous group. The adaptive profile was characterized by low fear of falling, low avoidance, lower fear of loss of autonomy, and relatively preserved self-efficacy. The vulnerable profile occupied an intermediate position, with moderately expressed fear of falling, avoidance behavior, and fear of loss of autonomy, together with reduced self-efficacy. The maladaptive profile was characterized by the highest fear of falling and the most pronounced avoidance behavior. These findings are consistent with studies indicating heterogeneity among older adults with fear of falling (Merchant et al. 2022; Goh et al. 2022; Finch et al. 2015), and further suggest that differences between groups involve not only fear and activity restriction, but also self-efficacy and fear of loss of autonomy.
The sensitivity analysis showed that the three-class solution remained relatively consistent across different MMSE cut-off scores. A similar profile structure was obtained both in the main analytical sample with MMSE ≥ 20 and in the broader sample with MMSE ≥ 15. This suggests that the identified profiles were not simply a consequence of the selected cognitive threshold. At the same time, the use of this profile structure in samples with lower MMSE scores requires further examination, especially given the self-report nature of the measures used.
The identified classes did not differ significantly in postural balance or number of falls. Therefore, the differences between the profiles could not be fully explained by objective indicators of fall risk or actual fall experience, but may reflect different psychological configurations of responding to perceived fall risk. At the same time, significant differences were found for age, daily pain level, and state social defense. The maladaptive profile included older respondents, whereas the vulnerable profile showed the highest daily pain level and the highest state social defense. These findings are consistent with evidence that pain appraisal, age-related changes, and psychosocial concerns about dependence and support may be associated with fear of falling and activity restriction (Creighton et al. 2017; Xiong et al. 2024; Payette et al. 2016; Savvakis et al. 2024; Shimizu et al. 2024). However, these differences should be interpreted as complementary characteristics rather than defining features of the profiles.
The second research question concerned how fear of falling and avoidance behavior were associated with psychological factors not used in profile construction. The regression models showed that these associations differed across profiles. Given the relatively small size of some latent profiles, especially the maladaptive profile, the regression findings should be interpreted as exploratory patterns of associations rather than stable predictive models.
In the adaptive profile, fear of falling was mainly associated with pain magnification and suppression of emotions, whereas active avoidance showed a negative association. Within the framework of Protection Motivation Theory, these associations may be considered in relation to subjective threat appraisal and coping-related responses. The positive association between pain magnification and fear of falling may reflect a higher subjective appraisal of the seriousness of possible fall-related consequences and greater sensitivity to potential physical harm. The negative association with the coping strategy “active avoidance” may indicate that limiting involvement in situations perceived as potentially dangerous is accompanied by a subjective reduction in perceived threat within this profile. At the same time, the positive association with suppression of emotions may reflect difficulties in emotional processing of experiences related to fall-related threat, with persistent internal tension despite relatively more adaptive overall functioning. Avoidance behavior in this profile was weakly explained and was associated only with pain rumination, which may also correspond to increased fixation on bodily sensations and cognitive processing of potential fall-related threat (Creighton et al. 2017; Xiong et al. 2024).
In the vulnerable profile, fear of falling was associated only with the trait phobic component of anxiety. From the perspective of Protection Motivation Theory, this may reflect heightened threat appraisal, in which potentially dangerous situations are perceived through a stronger sense of vulnerability and anticipation of unfavorable consequences. Avoidance behavior in this profile was associated with anxious appraisal of future prospects and coping strategies. The positive association with anxious appraisal of future events may be viewed as an expression of heightened threat appraisal, since anticipation of negative consequences is accompanied by a greater tendency to restrict activity. The negative association with cooperation may reflect the role of interpersonal support as a coping resource. When an older adult can rely on interpersonal support and receive assistance, their perceived ability to cope with the situation may be higher, and the need to avoid daily activity may be less pronounced. The negative association with emotional discharge may indicate the presence of an alternative way of responding to tension, in which anxiety and fear are expressed emotionally rather than through activity restriction. The negative association with the coping strategy “withdrawal” may be related to a temporary reduction in subjective threat through psychological distancing from an emotionally stressful situation, without necessarily shifting toward stable behavioral avoidance. The negative association with disturbances of cyclic functions in depression may indicate that sleep disturbances and diurnal mood fluctuations in this profile are not necessarily accompanied by stable activity restriction. These symptoms may reflect variability in emotional and physiological state rather than a persistent appraisal of one’s inability to cope with fall-related threat.
When the maladaptive profile is interpreted through Protection Motivation Theory, the divergent associations may reflect not only the intensity of threat, but also differences in coping-related responses to this threat. The negative association with the coping strategy “submission” may indicate that greater reliance on interpersonal support is associated with a lower need for independent control over the situation. In this case, the threat of falling remains significant, but coping may be experienced as more dependent on external support. Positive associations with trait anxiety, total depressive symptoms, and dependence on communication may reflect a combination of high threat appraisal, perceived vulnerability, and a need for reassurance from others.
The positive association with the coping strategy “optimism” may be related to the fact that, in this profile, optimism does not necessarily reflect genuine confidence in one’s own ability to cope with the threat. It may function as a form of cognitive self-reassurance in the context of high anxiety and perceived vulnerability. In this case, a positive attitude that “everything will be fine” may coexist with low confidence in one’s own actions and may not prevent avoidance. The negative association with physical symptoms of depression may be related to lower involvement in situations perceived as potentially risky. With reduced activity, fatigue, and hypodynamia, an older adult may encounter fewer situations that activate fall-related threat, and therefore subjective fear may be less pronounced. At the same time, avoidance behavior in this profile was associated with ideational disturbances in depression, that is, pessimistic evaluation of oneself, feelings of inadequacy, loss, or impairment. In the logic of PMT, this may reflect lower appraisal of one’s own coping ability, in which activity restriction becomes a more likely form of protective responding. This partially corresponds to previous findings linking depressive symptoms with activity restriction, while also suggesting that such associations may be profile-specific rather than universal (Gagnon et al. 2005).
Overall, the results support the view that fear of falling and avoidance behavior are related but not identical phenomena. Across profiles, they were associated with different psychological characteristics. This is consistent with previous studies showing that the behavioral repertoire of older adults varies depending on the presence and severity of fear of falling (Filiatrault and Desrosiers 2011), and with findings suggesting that fear of falling and avoidance should not be treated as fully identical constructs (Rider et al. 2026). From the perspective of Protection Motivation Theory, these differences may be related to the fact that the behavioral response to fall-related threat is associated not only with the intensity of fear, but also with the combination of threat appraisal, perceived vulnerability, coping resources, and the subjective costs of protective behavior. The findings also suggest that avoidance behavior may be better understood by taking into account anxious expectations, pain appraisal, depressive beliefs, interpersonal dependence, and coping strategies, rather than only the intensity of fear of falling.

5. Conclusions

The present study suggests that older adults with fear of falling may not represent a psychologically homogeneous group. Using latent profile analysis, three data-derived profiles were identified based on the combination of fear of falling, avoidance behavior, self-efficacy, and fear of loss of autonomy. These profiles were provisionally labeled as adaptive, vulnerable, and maladaptive. The profile structure remained relatively consistent across different MMSE cut-off scores, although further validation is needed, especially in samples with more pronounced cognitive impairment.
The findings suggest that fear of falling and avoidance behavior may represent related but not identical phenomena. Their psychological correlates differed across profiles, indicating that activity restriction in older adults may involve factors beyond the intensity of fear of falling or objective fall-related indicators such as postural balance or number of falls.
Class-specific regression models suggested that fear of falling and avoidance behavior were associated with different psychological factors across profiles. In the adaptive profile, fear and avoidance were mainly related to pain appraisal and emotion-related coping. In the vulnerable profile, phobic anxiety and anxious appraisal of future events were more prominent. In the maladaptive profile, fear of falling and avoidance were embedded in a broader pattern of anxiety-related, depressive, interpersonal, and coping-related characteristics.
Overall, the findings support the potential usefulness of a profile-based approach for studying fear of falling and avoidance behavior in older adults. From the perspective of Protection Motivation Theory, such an approach may help describe differences in threat appraisal, coping-related resources, and protective behavioral responses across older adults. It may also help identify more differentiated targets for psychological support, including pain-related beliefs, anxiety components, depressive appraisals, coping strategies, self-efficacy, and concerns about autonomy. Future studies should further validate the identified profiles in independent samples and examine their stability, replicability, and clinical relevance in longitudinal designs.

6. Limitations

The present study has several limitations. First, its cross-sectional design does not allow conclusions about the causal direction of the associations between fear of falling, avoidance behavior, and psychological factors. Avoidance may develop in response to fear, but activity restriction itself may also reduce successful movement experience and strengthen fear over time. Longitudinal studies are needed to clarify these relationships.
Second, the study relied mainly on self-report measures. Although the main self-report battery included eight instruments rather than all measures used in the study, and respondents were allowed to take breaks as needed, fatigue-related response bias and satisficing cannot be completely excluded. Although MMSE screening was used and participants with more pronounced cognitive impairment were excluded from the main analysis, responses could still be influenced by current emotional state, subjective appraisal, social desirability, reduced concentration, or tiredness. We did not directly assess fatigue or satisficing as a function of MMSE. Future studies would benefit from combining self-report data with behavioral, clinical, and performance-based assessments and from including more direct indicators of response quality.
Third, the main analyses were conducted in respondents with MMSE scores of 20 or higher. This increased the reliability of self-report data but limits the generalizability of the findings to older adults with more severe cognitive impairment. The profile structure may differ in samples with progressive dementia and should be examined separately.
Fourth, the latent profile analysis was exploratory and data-driven. Although the three-class solution was supported by several fit and classification indices and showed relative consistency across MMSE-based sensitivity analyses, the possibility of overfitting cannot be completely excluded. Its external replicability also remains untested. Therefore, the identified profiles should be regarded as provisional data-derived analytical groupings rather than stable psychological types. Future studies should examine whether the same profile structure can be reproduced in independent samples and different clinical or community settings.
Fifth, the sample had specific socio-demographic characteristics. Women were overrepresented, and a substantial proportion of respondents lived in long-term care institutions. These factors may limit the generalizability of the results to men and to community-dwelling older adults.
Sixth, multiple between-class comparisons and profile-specific regression analyses were conducted. Although FDR correction was applied to reduce the risk of false-positive findings, the possibility of Type I error inflation cannot be completely ruled out.
Finally, the profile-specific regression analyses should be interpreted with particular caution. The maladaptive profile included only 26 respondents, while several predictors were retained in the model of fear of falling. This low participant-to-predictor ratio may have increased the risk of model instability and inflation of explained variance. Therefore, these findings should be regarded as exploratory and hypothesis-generating rather than confirmatory, despite the application of FDR correction.

Author Contributions

Conceptualization, O.Y.S.; methodology, O.Y.S. and T.K.K.; validation, O.Y.S.; formal analysis, T.K.K.; investigation, T.K.K.; resources, T.K.K.; data curation, T.K.K.; writing—original draft preparation, T.K.K.; writing—review and editing, O.Y.S.; visualization, T.K.K.; supervision, O.Y.S.; project administration, T.K.K. All authors have read and agreed to the published version of the manuscript.

Funding

This research received no external funding.

Institutional Review Board Statement

The study was conducted in accordance with the Declaration of Helsinki and approved by the Ethical Committee (IRB) of St. Petersburg State University (protocol No. 115-02-1, 22 January 2025).

Informed Consent Statement

Informed consent was obtained from all subjects involved in the study.

Data Availability Statement

The data presented in this study are available upon a reasonable request from the corresponding author.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. Adandom, Henrietha C., Suha Damag, Michael E. Kalu, Israel I. Adandom, Adesola C. Odole, Lisa L. Cook, and Oluwagbohunmi A. Awosoga. 2026. Personality traits and coping with fear of falling: An interpretive description study of older adults in Canada. Journal of Health Psychology, 1–15. [Google Scholar] [CrossRef] [Scilit]
  2. Alexopoulos, George S., Robert C. Abrams, Robert C. Young, and Charles A. Shamoian. 1988. Cornell scale for depression in dementia. Biological Psychiatry 23: 271–84. [Google Scholar] [CrossRef] [Scilit]
  3. Almajid, Rania, Nuha Alharbi, Zainab Sharahili, and Dalal Alqahtani. 2025. The clinical characteristics of men and women living with a stroke: Influence of gender and fall self-efficacy. Physiotherapy Research International 30: e70054. [Google Scholar] [CrossRef] [Scilit]
  4. Auais, Mohammad, Beatriz Alvarado, Ricardo Guerra, Carmen Curcio, Ellen E. Freeman, Alban Ylli, and Nandini Deshpande. 2017. Fear of falling and its association with life-space mobility of older adults: A cross-sectional analysis using data from five international sites. Age and Ageing 46: 459–65. [Google Scholar] [CrossRef] [Scilit]
  5. Baek, Wonhee, Ari Min, Yoonjung Ji, Chang Gi Park, and Minkyung Kang. 2024. Impact of activity limitations due to fear of falling on changes in frailty in Korean older adults: A longitudinal study. Scientific Reports 14: 19121. [Google Scholar] [CrossRef] [Scilit]
  6. Bandura, Albert. 1977. Self-efficacy: Toward a unifying theory of behavioral change. Psychological Review 84: 191–215. [Google Scholar] [CrossRef]
  7. Berg, Katherine, Sharon Wood-Dauphine, J. I. Williams, and Brian Maki. 1992. Measuring balance in the elderly: Validation of an instrument. Canadian Journal of Public Health 83: S7–S11. [Google Scholar]
  8. Bizyuk, Alexander Pavlovich, Lyudvig Iosifovich Vasserman, and Boris Veniaminovich Iovlev. 2005. Application of the Integrative Anxiety Test. Saint Petersburg: V. M. Bekhterev National Medical Research Center of Psychiatry and Neurology. (In Russian) [Google Scholar]
  9. Creighton, Alexandra, Tanya Davison, and David Kissane. 2017. The correlates of anxiety among older adults in nursing homes and other residential aged care facilities: A systematic review. International Journal of Geriatric Psychiatry 32: 141–54. [Google Scholar] [CrossRef] [Scilit]
  10. Damar, Hale Turhan, and Canan Demir Barutcu. 2022. Relationship between hospitalised older people’s fear of falling and adaptation to old age, quality of life, anxiety and depression. International Journal of Older People Nursing 17: e12467. [Google Scholar] [CrossRef] [Scilit]
  11. Delbaere, Kim, Geert Crombez, Guy Vanderstraeten, Tine Willems, and Dirk Cambier. 2004. Fear-related avoidance of activities, falls and physical frailty: A prospective community-based cohort study. Age and Ageing 33: 368–73. [Google Scholar] [CrossRef] [Scilit]
  12. Denkinger, Michael D., Albert Lukas, Thorsten Nikolaus, and Klaus Hauer. 2015. Factors associated with fear of falling and associated activity restriction in community-dwelling older adults: A systematic review. The American Journal of Geriatric Psychiatry 23: 72–86. [Google Scholar] [CrossRef] [Scilit]
  13. Even-Zohar, Ahuva, Shulamith Kreitler, and Hanna Gendel Guterman. 2025. Falls of older adults: Which is worse, falling or fear of falling? Journal of Ageing and Longevity 5: 20. [Google Scholar] [CrossRef] [Scilit]
  14. Filiatrault, Johanne, and Johanne Desrosiers. 2011. Coping strategies used by seniors going through the normal aging process: Does fear of falling matter? Gerontology 57: 228–36. [Google Scholar] [CrossRef] [Scilit]
  15. Finch, Caroline F., Karen Stephan, Anna Wong Shee, Keith Hill, Terry P. Haines, Lindy Clemson, and Lesley Day. 2015. Identifying clusters of falls-related hospital admissions to inform population targets for prioritising falls prevention programmes. Injury Prevention 21: 254–59. [Google Scholar] [CrossRef] [Scilit]
  16. Folstein, Marshal F., Susan E. Folstein, and Paul R. McHugh. 1975. “Mini-mental state.” A practical method for grading the cognitive state of patients for the clinician. Journal of Psychiatric Research 12: 189–98. [Google Scholar] [CrossRef] [Scilit]
  17. Gagnon, Nadine, Alastair J. Flint, Gary Naglie, and Gerald M. Devins. 2005. Affective correlates of fear of falling in elderly persons. The American Journal of Geriatric Psychiatry 13: 7–14. [Google Scholar] [CrossRef]
  18. Goh, Choon-Hian, Kam Kang Wong, Maw Pin Tan, Siew-Cheok Ng, Yea Dat Chuah, and Ban-Hoe Kwan. 2022. Development of an effective clustering algorithm for older fallers. PLoS ONE 17: e0277966. [Google Scholar] [CrossRef] [Scilit]
  19. Heim, Edgar. 1988. Coping and psychological adaptation: Is there appropriate and inappropriate coping? Psychotherapie, Psychosomatik, Medizinische Psychologie 38: 8–18. [Google Scholar]
  20. Imani, Hadis, and Ben Godde. 2021. The association between falls and autonomy in older adults is mediated by self-efficacy. Discover Psychology 1: 8. [Google Scholar] [CrossRef] [Scilit]
  21. Jang, Yuri, Min-Kyoung Rhee, Juyoung Park, Jung Eun Ko, Nan Sook Park, Benjamin F. Henwood, Stacey L. Schepens Niemiec, and Soondool Chung. 2025. Falls as a mental health risk among Korean-American residents in subsidized senior housing: Mediating effect of fall-related self-efficacy. Aging & Mental Health 29: 542–48. [Google Scholar]
  22. Kongsted, Alice, and Anne Molgaard Nielsen. 2017. Latent class analysis in health research. Journal of Physiotherapy 63: 55–58. [Google Scholar] [CrossRef] [Scilit]
  23. Konovalchik, Tatyana Kirillovna. 2024. Fear of falling in older adults with various diseases. Clinical Psychology and Special Education 13: 76–95. (In Russian) [Google Scholar] [CrossRef] [Scilit]
  24. Konovalchik, Tatyana Kirillovna, and Olga Yurievna Strizhitskaya. Forthcoming. Russian adaptation of the Modified Fear of Falling Avoidance Behavior Questionnaire (mFFABQ-R). Counseling Psychology and Psychotherapy. Manuscript submitted for publication. (In Russian)
  25. Landers, Merrill R., Cortney Durand, D. Shalom Powell, Leland E. Dibble, and Daniel L. Young. 2011. Development of a scale to assess avoidance behavior due to a fear of falling: The Fear of Falling Avoidance Behavior Questionnaire (FFABQ). Physical Therapy 91: 1253–65. [Google Scholar] [CrossRef] [Scilit]
  26. Lee, Dayeon, and Sunghee H. Tak. 2023. A concept analysis of fear of falling in older adults: Insights from qualitative research studies. BMC Geriatrics 23: 651. [Google Scholar] [CrossRef] [Scilit]
  27. Lee, Eun Sook, and Boyoung Kim. 2024. The impact of fear of falling on health-related quality of life in community-dwelling older adults: Mediating effects of depression and moderated mediation effects of physical activity. BMC Public Health 24: 2459. [Google Scholar] [CrossRef] [Scilit]
  28. Levin, Oleg Semenovich, ed. 2023. Clinical Scales in Neurology: A Guide for Physicians. Moscow: MEDpress-inform. 272p, ISBN 978-5-907632-88-2. (In Russian) [Google Scholar]
  29. Li, Fuzhong, K. John Fisher, Peter Harmer, Edward McAuley, and Nicole L. Wilson. 2005. Fear of falling in elderly persons: Association with falls, functional ability, and quality of life. The Journals of Gerontology 58: P34–P40. [Google Scholar] [CrossRef] [Scilit]
  30. Lord, Stephe R., Catherine Sherrington, and Hylton B. Menz. 2007. Falls in Older People: Risk Factors and Strategies for Prevention, 2nd ed. Cambridge: Cambridge University Press. [Google Scholar] [CrossRef] [Scilit]
  31. Lugovskaya, Alena Aleksandrovna. 2017. Psychometric properties of the scale of constructive strategies of the coping behavior diagnostic technique. Journal of the Grodno State Medical University 15: 279–283. (In Russian) [Google Scholar] [CrossRef] [Scilit]
  32. Mahoney, Florence I., and Dorothea W. Barthel. 1965. Functional evaluation: The Barthel Index. Maryland State Medical Journal 14: 61–65. [Google Scholar]
  33. Merchant, Reshma Aziz, Yiong Huak Chan, Ivan Aprahamian, and John E. Morley. 2022. Patterns of participation restriction among older adults at risk of falls and relationship with intrinsic capacity: A latent cluster analysis. Frontiers in Medicine 9: 1023879. [Google Scholar] [CrossRef] [Scilit]
  34. Meyer, Marco, Andreas Arnold, Thomas Stein, Ulrich Niemöller, and Christian Tanislav. 2025. Fear of falling in older adults undergoing comprehensive geriatric care: Results of a prospective observational study. Journal of Clinical Medicine 14: 4366. [Google Scholar] [CrossRef] [Scilit]
  35. Munir, Shanza, Tehreem Mukhtar, Zunaira Ahmed, Zobia Saeed, Gohar Iqbal, Umer Kamran, and Zainab Imran. 2026. Interplay between fear of falling, balance deficits, and functional independence in older adults. The Healer Journal of Physiotherapy and Rehabilitation Sciences 6: 16–22. [Google Scholar] [CrossRef] [Scilit]
  36. Osin, Evgeny Nikolaevich, and Dmitry Alekseevich Leontiev. 2013. Differential Loneliness Experience Questionnaire: Structure and psychometric properties. Psychology. Journal of the Higher School of Economics 10: 55–81. (In Russian) [Google Scholar]
  37. Öztürk, Ümmühan Meltem, and Mümin Polat. 2026. Investigation of the relationship between pain, fear of movement, and falling in geriatric patients. PeerJ 14: e20661. [Google Scholar] [CrossRef] [Scilit]
  38. Payette, Marie-Christine, Claude Belanger, Vanessa Léveillé, and Sébastien Grenier. 2016. Fall-related psychological concerns and anxiety among community-dwelling older adults: Systematic review and meta-analysis. PLoS ONE 11: e0152848. [Google Scholar] [CrossRef] [Scilit]
  39. Preissner, Christian Erik, Navin Kaushal, Kathleen Charles, and Bärbel Knäuper. 2023. A protection motivation theory approach to understanding how fear of falling affects physical activity determinants in older adults. The Journals of Gerontology: Series B 78: 30–39. [Google Scholar] [CrossRef] [Scilit]
  40. Radchikova, Nataly P., Galina A. Adashinskaya, Tigranui R. Sanoyan, and Anastasia A. Shupta. 2020. Pain catastrophizing scale: Adaptation of the questionnaire. Clinical Psychology and Special Education 9: 169–87. [Google Scholar] [CrossRef] [Scilit]
  41. Rider, John V., Nirmala Lekhak, Daniel L. Young, and Merrill R. Landers. 2026. Psychological factors associated with fear of falling and fear of falling avoidance behavior in older adults: Results from a national sample. Clinical Gerontologist 49: 55–68. [Google Scholar] [CrossRef] [Scilit]
  42. Rogers, Ronald W. 1983. Cognitive and physiological processes in fear appeals and attitude change: A revised theory of protection motivation. In Social Psychophysiology: A Sourcebook. Edited by John T. Cacioppo and Richard E. Petty. New York: Guilford Press, pp. 153–176. [Google Scholar]
  43. Romek, Vladimir Georgievich. 1997. Self-Confidence as a Socio-Psychological Characteristic of Personality. Ph.D. thesis, Rostov State University, Rostov-on-Don, Russia; 23p. (In Russian) [Google Scholar]
  44. Sattler, Tjard, Sophie Gottschalk, Hans-Helmut König, Tobias Braun, Gisela Büchele, Michael Denkinger, Tim Fleiner, Corinna Nerz, Kilian Rapp, Martina Schäufele, and et al. 2025. Path model explaining the association between fear of falling and health-related quality of life in (pre-)frail older adults. BMC Geriatrics 25: 87. [Google Scholar] [CrossRef] [Scilit]
  45. Savvakis, Ioannis, Theodoula Adamakidou, and Christos Kleisiaris. 2024. Physical-activity interventions to reduce fear of falling in frail and pre-frail older adults: A systematic review of randomized controlled trials. European Geriatric Medicine 15: 333–44. [Google Scholar] [CrossRef] [Scilit]
  46. Scheffer, Alice C., Marieke J. Schuurmans, Nynke van Dijk, Truus Van Der Hooft, and Sophia E. De Rooij. 2010. Reliability and validity of the visual analogue scale for fear of falling in older persons. Journal of the American Geriatrics Society 58: 2228–30. [Google Scholar] [CrossRef] [Scilit]
  47. Schwarzer, Ralf, Matthias Jerusalem, and Vladimir Georgievich Romek. 1996. Russian version of the General Self-Efficacy Scale. Inostrannaya Psikhologiya 7: 71–77. (In Russian) [Google Scholar]
  48. Shimizu, Yuho, Kenichiro Sato, Susumu Ogawa, Daisuke Cho, Yoshifumi Takahashi, Daichi Yamashiro, Yan Li, Tomoya Takahashi, Keigo Hinakura, Akihiro Iizuka, and et al. 2024. Physically confident older adults are not afraid to fall, but only if they have positive images of older people: A cross-sectional study in Japan. Journal of Gerontology and Geriatrics 72: 66–75. [Google Scholar] [CrossRef] [Scilit]
  49. Sirota, Natalia Aleksandrovna, and Denis Vladimirovich Moskovchenko. 2014. Psychodiagnostics of fear of disease progression: Results of approbation of the Russian-language version of the questionnaire. V.M. Bekhterev Review of Psychiatry and Medical Psychology 4: 86–91. [Google Scholar]
  50. Tabacchi, Garden, Giovanni Angelo Navarra, Antonino Scardina, Ewan Thomas, Antonella D’Amico, Javier Gene-Morales, Juan C. Colado, Antonio Palma, and Marianna Bellafiore. 2025. A multiple correspondence analysis of the fear of falling, sociodemographic, physical and mental health factors in older adults. Scientific Reports 15: 6341. [Google Scholar] [CrossRef] [Scilit]
  51. Tinetti, Mary E., and Lynda Powell. 1993. Fear of falling and low self-efficacy: A cause of dependence in elderly persons. Journal of Gerontology 48: 35–38. [Google Scholar] [CrossRef] [Scilit]
  52. Tinetti, Mary E., Donna Richman, and Lynda Powell. 1990. Falls efficacy as a measure of fear of falling. Journal of Gerontology 45: P239–43. [Google Scholar] [CrossRef] [Scilit]
  53. Türkmen Keskin, Sultan, Ebru Sönmez Sari, Şahinde Canbulat, and Fatma Özlem Öztürk. 2025. Fear of falling avoidance behavior on daily living activities and physical activity in older adults: A cross-sectional study. Nursing & Health Sciences 27: e70060. [Google Scholar] [CrossRef] [Scilit]
  54. Wu, Kuan-Ying, Duan-Rung Chen, Chang-Chuan Chan, Yen-Po Yeh, and Hsiu-Hsi Chen. 2023. Fear of falling as a mediator in the association between social frailty and health-related quality of life in community-dwelling older adults. BMC Geriatrics 23: 421. [Google Scholar] [CrossRef] [Scilit]
  55. Xiong, Wanhong, Dan Wang, Wei Ren, Xinyi Liu, Renhui Wen, and Yu Luo. 2024. The global prevalence of and risk factors for fear of falling among older adults: A systematic review and meta-analysis. BMC Geriatrics 24: 321. [Google Scholar] [CrossRef] [Scilit]
  56. Yang, Yue, Meiying Li, Xiaoge Ding, Lei Zhang, Merxiang Jin, and Yinji Jin. 2025. Classification and influencing factors of fear of falling in elderly stroke patients: A latent profile analysis. Aging Clinical and Experimental Research 37: 329. [Google Scholar] [CrossRef] [Scilit]
  57. Zijlstra, G. A. Rixt, Jolanda C. M. van Haastregt, Erik van Rossum, Jacques T. M. van Eijk, Lucy Yardley, and Gertrudis I. J. M. Kempen. 2007. Interventions to reduce fear of falling in community-living older people: A systematic review. Journal of the American Geriatrics Society 55: 603–15. [Google Scholar] [CrossRef] [Scilit]
Table 1. Demographic characteristics of the sample (N = 217).
Table 1. Demographic characteristics of the sample (N = 217).
Sex and Age
GroupTotal (N)Older Men (N)Older Women (N)
Aged 60–74942371
Aged 75–901101694
Aged 91+13112
Socio-Demographic Characteristics of the Sample
CharacteristicN%
EducationIncomplete secondary10.5
Secondary146.5
Vocational secondary7635.0
Higher education12356.7
Academic degree41.8
EmploymentRetired, employed198.8
Retired, unemployed19891.2
Living conditionsLiving alone4621.2
Living with a spouse3415.7
Living with relatives4420.3
Living in a long-term care facility9342.9
Note. N = number of respondents.
Table 2. Fit indices for alternative latent profile models (N = 217).
Table 2. Fit indices for alternative latent profile models (N = 217).
Model Fit Index2 Classes3 Classes4 Classes
LogLik−2042.929−1997.791−1977.323
k131823
AIC4111.8574031.5834000.646
BIC4155.7964092.4214078.383
aBIC4114.6004035.3814005.499
Entropy0.9600.8990.924
Smallest profile size, %16.612.06.9
LMR-LRT p<0.0010.00230.0293
BLRT p<0.001<0.001<0.001
Note. LogLik—log-likelihood; k—number of estimated parameters; AIC—Akaike Information Criterion; BIC—Bayesian Information Criterion; aBIC—sample-size-adjusted Bayesian Information Criterion; Entropy—classification accuracy index, with higher values indicating better class separation; Smallest profile size = percentage of respondents in the smallest latent class; LMR-LRT-Lo–Mendell–Rubin adjusted likelihood ratio test; BLRT—bootstrap likelihood ratio test.
Table 3. Standardized mean scores of profiling variables by latent class.
Table 3. Standardized mean scores of profiling variables by latent class.
VariableClass 1Class 2Class 3
Class size, N1306126
Class size, %59.928.112.0
Fear of falling−0.5270.6881.093
Avoidance behavior−0.6040.3742.312
Self-efficacy0.129−0.210−0.239
Fear of loss of autonomy−0.2230.3970.476
Note. N = number of respondents.
Table 4. Sensitivity analysis of the three-class solution across MMSE cut-off scores.
Table 4. Sensitivity analysis of the three-class solution across MMSE cut-off scores.
Model Fit IndexMMSE ≥ 15MMSE ≥ 20MMSE ≥ 25MMSE ≥ 28
N229217179123
LogLik−1146.175−1085.691−885.900−599.671
k18181818
AIC2328.3502207.3831807.8001235.342
BIC2390.1572268.2211865.1731285.961
aBIC2333.1082211.1811808.1681229.046
Entropy0.9100.9050.8940.915
Smallest profile size, %11.81210.110.6
LMR-LRT p0.00020.00030.01120.1111
BLRT p<0.001<0.001<0.001<0.001
Note. LogLik—log-likelihood; k—number of estimated parameters; AIC—Akaike Information Criterion; BIC—Bayesian Information Criterion; aBIC—sample-size-adjusted Bayesian Information Criterion; Entropy—classification accuracy index, with higher values indicating better class separation; Smallest profile size = percentage of respondents in the smallest latent class; LMR-LRT-Lo–Mendell–Rubin adjusted likelihood ratio test; BLRT—bootstrap likelihood ratio test; MMSE—Mini-Mental State Examination.
Table 5. Standardized mean scores of profiling variables across MMSE cut-off scores.
Table 5. Standardized mean scores of profiling variables across MMSE cut-off scores.
SubsampleVariableClass 1Class 2Class 3
MMSE ≥ 15Class size, n1396327
Class size, %60.727.511.8
Fear of falling−0.5150.6721.109
Avoidance behavior−0.6130.3772.279
Self-efficacy0.137−0.207−0.230
Fear of loss of autonomy−0.2610.3960.435
MMSE ≥ 20Class size, n1306126
Class size, %59.928.112
Fear of falling−0.5270.6881.093
Avoidance behavior−0.6040.3742.312
Self-efficacy0.129−0.210−0.239
Fear of loss of autonomy−0.2230.3970.476
MMSE ≥ 25Class size, n1115018
Class size, %6227.910.1
Fear of falling−0.4600.7561.108
Avoidance behavior−0.6100.3542.314
Self-efficacy0.1−0.197−0.135
Fear of loss of autonomy−0.1460.4580.629
MMSE ≥ 28Class size, n793113
Class size, %64.225.210.6
Fear of falling−0.3550.6731.197
Avoidance behavior−0.6100.341.995
Self-efficacy0.067−0.067−0.028
Fear of loss of autonomy−0.0350.4580.37
Table 6. Between-class comparisons of functional, fall-related, demographic, and psychological indicators.
Table 6. Between-class comparisons of functional, fall-related, demographic, and psychological indicators.
VariableMean RankHp
Adaptive Profile
(n = 130) 		Vulnerable Profile
(n = 61) 		Maladaptive Profile
(n = 26)
Barthel Index108.080114.480100.7501.1520.562
Berg Balance Scale109.330112.59098.9400.8730.646
Number of falls105.950115.190109.7300.9950.608
Age98.300115.610149.64015.0020.001
Daily pain level100.860123.510117.1806.0240.049
State social defense108.890120.93080.9407.3110.026
Note. n = sample size; Mean Rank = average rank; H = Kruskal–Wallis test statistic; p = level of statistical significance.
Table 7. Class-specific regression models of psychological variables associated with fear of falling.
Table 7. Class-specific regression models of psychological variables associated with fear of falling.
Adaptive profile, N = 130, R2 = 0.114
PredictorsβtpFDR-adjusted p
Pain magnification0.2593.0880.0020.007
Coping strategy “Active avoidance”−0.224−2.5420.0120.020
Coping strategy “Suppression of emotions”0.1792.0390.0440.044
Vulnerable profile, N = 61, R2 = 0.084
PredictorsβtpFDR-adjusted p
Trait phobic component of anxiety0.2912.3130.0240.034
Maladaptive profile, N = 26, R2 = 0.823
PredictorsβtpFDR-adjusted p
Coping strategy “Submission”−0.779−6.448<0.0010.005
Trait anxiety0.6475.787<0.0010.005
Physical symptoms of depression−0.428−3.2510.0040.01
Total depression score0.3112.3340.0310.039
Coping strategy “Optimism”0.4083.1320.0050.01
Dependence on communication0.2922.2530.0360.04
Note. N—number of participants; R2—proportion of variance explained by the model; β—standardized regression coefficient; t—t statistic; p—two-tailed significance level, FDR-adjusted p—p-value adjusted using the false discovery rate procedure.
Table 8. Class-specific regression models of psychological variables associated with avoidance behavior.
Table 8. Class-specific regression models of psychological variables associated with avoidance behavior.
Adaptive profile, N = 130, R2 = 0.036
PredictorsβtpFDR-adjusted p
Pain rumination0.1892.1860.0310.031
Vulnerable profile, N = 61, R2 = 0.417
PredictorsβtpFDR-adjusted p
Disturbance of cyclic functions in depression−0.432−3.864<0.0010.002
Coping strategy
“Cooperation”		−0.427−3.974<0.0010.002
State anxious appraisal of perspective0.4413.774<0.0010.002
Coping strategy
“Emotional discharge”		−0.302−2.8060.0070.008
Coping strategy
“Withdrawal”		−0.318−2.8520.0060.008
Maladaptive profile, N = 26, R2 = 0.298
PredictorsβtpFDR-adjusted p
Ideational disturbances in depression0.5463.1890.0040.007
Note. N—number of participants; R2—proportion of variance explained by the model; β—standardized regression coefficient; t—t statistic; p—two-tailed significance level. FDR-adjusted p—p-value adjusted using the false discovery rate procedure.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Konovalchik, T.K.; Strizhitskaya, O.Y. Predictors of Avoidance Behavior in Fear of Falling Among Older Adults: A Latent Profile Analysis. Soc. Sci. 2026, 15, 379. https://doi.org/10.3390/socsci15060379

AMA Style

Konovalchik TK, Strizhitskaya OY. Predictors of Avoidance Behavior in Fear of Falling Among Older Adults: A Latent Profile Analysis. Social Sciences. 2026; 15(6):379. https://doi.org/10.3390/socsci15060379

Chicago/Turabian Style

Konovalchik, Tatyana K., and Olga Yu. Strizhitskaya. 2026. "Predictors of Avoidance Behavior in Fear of Falling Among Older Adults: A Latent Profile Analysis" Social Sciences 15, no. 6: 379. https://doi.org/10.3390/socsci15060379

APA Style

Konovalchik, T. K., & Strizhitskaya, O. Y. (2026). Predictors of Avoidance Behavior in Fear of Falling Among Older Adults: A Latent Profile Analysis. Social Sciences, 15(6), 379. https://doi.org/10.3390/socsci15060379

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop