Skip to main content
NCBI home page
Wiley Open Access Collection logoLink to Wiley Open Access Collection
. 2026 Jul 13;58(4):e70115. doi: 10.1111/jnu.70115

Longitudinal Trajectories and Determinants of Fear of Falling in Post‐Stroke Patients

Xiaomeng Han 1, Mingming Zhong 2,3, Wenxiu Niu 1,3, Fei Zheng 1,3, Yuhan Sha 1,3, Xin Feng 1, Ruiyuan Yin 1, Panpan Li 1, Fangli Ma 1,✉, Ying Ren 1,✉, Zhigang Zhang 1,✉
PMCID: PMC13360664  PMID: 42439480

ABSTRACT

Objective

Fear of falling is a common and serious psychological challenge after stroke, yet its longitudinal patterns remain poorly understood. Current evidence often fails to capture individual heterogeneity in recovery trajectories. This study aimed to identify distinct longitudinal trajectories of fear of falling and explore associated factors to inform targeted rehabilitation strategies in post‐stroke patients.

Design

A prospective longitudinal study.

Methods

From August 2023 to December 2024, we enrolled 166 stroke patients from the neurology wards of two hospitals in Lanzhou, China, forming a hospital‐based sample within an urban healthcare setting. Fear of falling was assessed using a single‐item question and the Short Falls Efficacy Scale‐International at four time points: 2 weeks, 1 month, 3 months, and 6 months post‐stroke. Data were analyzed using latent class growth modeling in Mplus 8.0.

Results

The prevalence of fear of falling was 48.8%, 49.4%, 45.8%, and 34.3% at 2 weeks, 1 month, 3 months, and 6 months, respectively. Mean Short Falls Efficacy Scale‐International scores were 11.30 (±4.74), 9.84 (±3.40), 9.61 (±3.13), and 8.00 (±1.22) at these time points. Three distinct fear of falling trajectories were identified: high‐level continuous decline (20.48%), medium‐level steady decline (26.51%), and low‐level fear (53.01%). Age, anxiety levels, and depression levels were significant factors associated with these trajectories.

Conclusion

Fear of falling after stroke demonstrates heterogeneous longitudinal trajectories that are significantly influenced by age, anxiety, and depression. These findings highlight the importance of individualized assessment and targeted psychological support during stroke rehabilitation.

Clinical Relevance

Fear of falling is a dynamic psychological challenge during stroke recovery. Routine screening for fear of falling, anxiety, and depression may help nurses identify patients at risk for persistent fear trajectories. Tailored rehabilitation and psychological interventions may improve rehabilitation participation and recovery outcomes.

Patient or Public Contribution

No patient or public engagement.

Keywords: factors, fear of falling, longitudinal trajectory, nurse, stroke

1. Introduction

Stroke is a leading cause of mortality and long‐term disability worldwide, with incidence rates continuing to rise. In 2019 alone, approximately 12.2 million new stroke cases were reported globally (GBD 2019 Stroke Collaborators 2021). During rehabilitation, stroke survivors are at high risk of falls due to muscle weakness, sensory deficits, impaired mobility, and balance dysfunction. Previous studies have reported that 11.5%–59% of stroke survivors experience falls after stroke (Xie et al. 2025). Given the potentially serious physical and psychological consequences of falls, many patients subsequently develop fear of falling, with reported prevalence rates ranging from 42% to 93.8% among stroke survivors (Tian et al. 2024).

Fear of falling is defined as a psychological state of excessive fear regarding potential physical injury, functional decline, or social embarrassment resulting from a loss of balance. This fear often leads to overly cautious behavior, such as deliberate reduction of daily activities, to avoid perceived fall risks. Such activity restriction directly contributes to declined physical function and reduced quality of life (Cumming et al. 2000; Lachman et al. 1998; Schoene et al. 2019). Furthermore, fear of falling profoundly impacts social behavior, exacerbating feelings of social isolation, anxiety, and depression (Painter et al. 2012; Scheffer et al. 2008). Consequently, the detrimental effects of fear of falling are considered comparable in importance to falls themselves.

While previous research indicates that fear of falling fluctuates over time (Delbaere et al. 2010), most studies have focused primarily on average group‐level trends. Such approaches assume that all patients follow a similar recovery pattern and may overlook substantial inter‐individual heterogeneity in psychological adaptation after stroke (Corbin 1998). In reality, fear of falling trajectories during stroke rehabilitation can vary significantly among patients. Identifying these heterogeneous developmental trajectories may help clinicians better understand the dynamic nature of fear of falling and facilitate targeted intervention strategies for high‐risk patients.

Latent class growth modeling (LCGM) is a person‐centered statistical approach that can identify distinct longitudinal trajectory subgroups within a population. Compared with traditional variable‐centered methods, LCGM is more suitable for capturing heterogeneity in longitudinal psychological and behavioral outcomes (Jung and Wickrama 2008). Therefore, this study aimed to identify distinct trajectories of fear of falling among stroke patients and explore associated influencing factors using LCGM. The findings may provide evidence to support individualized rehabilitation and nurse‐led fall prevention strategies during stroke recovery.

2. Methods

2.1. Study Design

This was a prospective longitudinal study approved by the Ethics Committee. The design of this observational study and the results are reported in accordance with the STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) statement (Skrivankova et al. 2021).

2.2. Setting

The study was conducted from August 2023 to December 2024 in neurology wards within two tertiary public hospitals in China (both located in urban areas).

2.3. Participants

Inclusion criteria: (1) age ≥ 18 years; (2) diagnosis of stroke according to established clinical criteria; (3) consultation within 2 weeks of onset; (4) ability to participate in out‐of‐bed rehabilitation activities; (5) preserved communication ability, defined as the capacity to understand and appropriately respond to simple verbal instructions and questions, as assessed by the attending neurologist or study nurse; and (6) no significant cognitive impairment according to education‐adjusted Mini‐Mental State Examination (MMSE) criteria (Zhang and Wang 1989). Exclusion criteria: (1) severe cardiopulmonary dysfunction; and (2) history of traumatic brain injury.

2.4. Sample

The sample size was estimated a priori for this longitudinal repeated‐measures study based on published sample size estimation guidelines (Jiang 2008). Considering four repeated assessments and a potential 20% attrition rate during follow‐up, the target sample size was set at 200 participants. After exclusions and losses to follow‐up, 166 participants completed all four assessments and were included in the final analysis, which satisfied the minimum sample size requirement for the planned analyses.

2.5. Variables and Measurement

A structured questionnaire was used to collect data on demographic factors (e.g., age, gender, education level) and disease‐related conditions (e.g., stroke type, time since onset, comorbidities, and history of falls).

2.5.1. Fear of Falling Assessment

Fear of falling was assessed using two complementary measures:

2.5.1.1. Single‐Item Fear of Falling Question

Participants were asked, “Are you afraid of falling?” with “yes” or “no” responses. This tool provides a direct, global assessment of fear of falling prevalence and has been widely used in post‐stroke populations for its clinical simplicity (Arfken et al. 1994).

2.5.1.2. Short Falls Efficacy Scale‐International (Short FES‐I)

This 7‐item questionnaire assesses concern about falling during various social and physical activities. Scores range from 7 to 28, with higher scores indicating greater fear. Based on established cut‐offs (Kempen et al. 2008), scores were interpreted as follows: 7‐8 (low concern), 9‐13 (moderate concern), and 14‐28 (high concern). The Short FES‐I was used to capture the multidimensional nature and severity of fear of falling, complementing the binary single‐item question. The Short FES‐I has demonstrated excellent internal consistency (Cronbach's α = 0.98) in Chinese stroke populations (Deng et al. 2015). According to Kempen et al. (2008), the psychometric properties and discriminative power of the Short FES‐I are nearly as good as the full 16‐item FES‐I. The authors further noted that while the full FES‐I provides more detailed information across a wider range of activities, the Short FES‐I may be more feasible in clinical settings when assessment time is limited or when respondents have difficulty completing longer questionnaires.

2.5.2. Psychological and Cognitive Measures

Cognitive function, anxiety, and depression were assessed at the baseline (T1, 2 weeks post‐stroke) during hospitalization. All baseline assessments were conducted through face‐to‐face interviews to ensure comprehension and data quality.

  1. Cognitive function was screened using the MMSE (Folstein et al. 1975). The MMSE total score ranges from 0 to 30, with higher scores indicating better cognitive function. Education‐adjusted cutoff scores were applied to define cognitive impairment: ≤ 17 for illiterate individuals, ≤ 20 for participants with primary school education (≤ 6 years), and ≤ 24 for those with secondary school education or above (> 6 years). The Chinese version of the MMSE has demonstrated good reliability, with an inter‐rater ICC of 0.99 and a test–retest ICC of 0.91 over 48–72 h (Zhang and Wang 1989).

  2. Anxiety symptoms were assessed using the 7‐item Generalized Anxiety Disorder scale (GAD‐7) (Wang et al. 2018). Total scores range from 0 to 21, with higher scores indicating more severe anxiety symptoms. Standard cut‐off scores were used to classify anxiety severity as minimal (0–4), mild (5–9), moderate (10–14), and severe (15–21). The Chinese version of the GAD‐7 has demonstrated good reliability and validity, with a Cronbach's α of 0.91 (Zeng et al. 2013).

  3. Depressive symptoms were assessed with the 9‐item Patient Health Questionnaire (PHQ‐9) (Xu et al. 2007). Total scores range from 0 to 27, with higher scores indicating more severe depressive symptoms. Standard cut‐off scores were used to classify depression severity as minimal (0–4), mild (5–9), moderate (10–14), and severe (15–27). The Chinese version of the PHQ‐9 has demonstrated good reliability and validity in stroke patients, with a Cronbach's α of 0.838 (Zheng et al. 2013).

  4. The Family Support Questionnaire (FSQ) is a 15‐item instrument adapted from the Perceived Social Support from Family scale developed by Procidano and Heller (1983). The original scale has demonstrated good reliability (Cronbach's α = 0.90) (Procidano and Heller 1983). The Chinese version was introduced by Zhang and Liu (2001) with yes/no responses and total scores ranging from 0 to 15. Based on the total score, family support was categorized as low (0–5), moderate (6–10), or high (11–15), with higher scores indicating greater perceived family support.

2.6. Procedures

Data were collected at four time points following stroke onset: 2 weeks (T1), 1 month (T2), 3 months (T3), and 6 months (T4).

The baseline assessment (T1) was conducted during hospitalization. Demographic and clinical data were obtained through electronic medical record review and structured patient interviews. Baseline assessments included cognitive function (MMSE), fear of falling (single‐item fear of falling question and Short FES‐I), anxiety (GAD‐7), depression (PHQ‐9), and family support (FSQ). All assessments were administered face‐to‐face by trained researchers to ensure comprehension and data quality.

Follow‐up assessments at T2, T3, and T4 focused on fear of falling and were conducted using the single‐item fear of falling question and the Short FES‐I. For participants who remained hospitalized, assessments were completed face‐to‐face at the bedside. For discharged participants, follow‐up data were collected through structured telephone interviews using standardized interview procedures to ensure consistency across time points. All questionnaires were reviewed for completeness immediately after data collection.

2.7. Data Analysis

Data were analyzed using IBM SPSS Statistics version 26.0 and Mplus version 8.0. Descriptive statistics were used to summarize participant characteristics and study variables. Continuous variables were described as mean ± standard deviation or median (interquartile range), while categorical variables were presented as frequencies and percentages.

LCGM was performed using Mplus 8.0 to identify distinct trajectories of fear of falling over time. Models with one to five latent classes were sequentially estimated. Model selection was based on a combination of statistical fit indices and clinical interpretability, including the Akaike Information Criterion (AIC), Bayesian Information Criterion (BIC), adjusted Bayesian Information Criterion (aBIC), entropy, Lo–Mendell–Rubin likelihood ratio test (LMR‐LRT), and bootstrap likelihood ratio test (BLRT). Lower AIC, BIC, and aBIC values indicated better model fit, while entropy values greater than 0.80 indicated acceptable classification accuracy. Significant LMR‐LRT and BLRT results (p < 0.05) suggested improved model fit compared with models containing one fewer class.

Differences among trajectory groups were examined using chi‐squared tests, one‐way analysis of variance, or nonparametric tests. Variables with statistical significance in the univariate analyses were entered into a multinomial logistic regression model to identify factors associated with fear of falling trajectory membership. Statistical significance was defined as p < 0.05.

3. Results

3.1. Baseline Characteristics of the Study Participants

A total of 200 questionnaires were distributed in this study, and 34 cases were lost in the study, and finally all 166 stroke patients completed the follow‐up. The participant flow diagram is shown in Figure S1.

Comparative analysis showed no significant differences at baseline between study completers (n = 166) and non‐completers (n = 34) across a range of demographic and clinical variables (Table S1). Since the groups were similar in all measured aspects, it can be inferred that the loss to follow‐up was non‐systematic and is not expected to have substantially biased the primary outcomes of this study.

3.2. Longitudinal Changes in Fear of Falling

The prevalence of fear of falling, as measured by the single‐item question, was 48.8% at 2 weeks, 49.4% at 1 month, 45.8% at 3 months, and 34.3% at 6 months, with the highest incidence observed at 1 month. Concurrently, the severity of fear of falling, assessed by the Short FES‐I, demonstrated a significant decreasing trend over time (Figure 1). The mean Short FES‐I scores were 11.30 (±4.74) at 2 weeks, 9.84 (±3.40) at 1 month, 9.61 (±3.13) at 3 months, and 8.00 (±1.22) at 6 months. This decline in severity was also reflected in the distribution of concern levels (Table S2), where the proportion of patients classified as having “high concern” (Short FES‐I score ≥ 14) was highest at 2 weeks (37.96%) and gradually decreased thereafter.

FIGURE 1.

FIGURE 1

Open in a new tab

Scatter plot of fear of falling over time.

3.3. Identification of Fear of Falling Trajectory Classes

A series of GMMs specifying one to five latent classes were estimated to identify distinct developmental trajectories. The fit indices for all models are presented in Table 1. The 3‐class solution was selected as the optimal model based on a combination of statistical evidence and theoretical rationale.

TABLE 1.

Model fitting results of 5 alternative latent categories of the developmental trajectory of fear of falling.

Category AIC BIC aBIC Entropy LMR BLRT Category probability (%)
1 3446.422 3465.094 3446.097 — — — 100.0
2 2799 2827.008 2798.513 0.932 0.2133 < 0.001 34.853/65.147
3 2248.603 2285.947 2247.954 0.997 < 0.001 < 0.001 20.482/53.012/26.506
4 2173.219 2219.899 2219.899 0.994 0.0748 < 0.001 23.494/16.787/6.627/53.012
5 2119.303 2175.319 2118.329 0.988 0.6349 < 0.001 12.886/4.238/21.687/53.012/8.434
Open in a new tab

Note: Bold values indicate the selected optimal model.

3.3.1. Significant Model Improvement

The LMR‐LRT showed a statistically significant improvement for the 3‐class over the 2‐class model (p < 0.001). In contrast, the test was non‐significant when comparing the 4‐class to the 3‐class model (p = 0.075), establishing the 3‐class model as the point of diminishing returns.

3.3.2. Exceptional Classification Accuracy

The model exhibited near‐perfect classification, as shown by an entropy of 0.997 and, crucially, the average latent class probability matrix (Table S3). The diagonal values of this matrix (1.000, 1.000, 0.997) indicate an almost 100% probability of correct class assignment, far exceeding the recommended threshold of 0.80 and ensuring high internal validity.

3.3.3. Substantial and Interpretable Classes

All three classes comprised substantial proportions of the sample (20.48%, 53.01%, 26.51%), supporting their substantive meaning. The 4‐ and 5‐class models, however, introduced small, likely spurious classes, which are indicative of overfitting and lack theoretical coherence.

Based on the parameters of the selected 3‐class model, the developmental trajectories of fear of falling are visualized in Figure 2. Each trajectory was named descriptively according to its distinct pattern of intercept (initial severity) and slope (rate of change), the estimated values for which are provided in Table S4. The resulting trajectories are: “low‐level fear” (C3, 53.01%, n = 88), “medium‐level steady decline” (C2, 26.51%, n = 44), and “high‐level continuous decline” (C1, 20.48%, n = 34).

FIGURE 2.

FIGURE 2

Open in a new tab

Latent categorical map of the developmental trajectory of fear of falling.

3.4. Univariate Analysis of Fear of Falling Trajectory Associations

Age group (χ 2 = 43.214, p < 0.001), educational level (χ 2 = 16.612, p = 0.034), medical payment method (χ 2 = 12.368, p = 0.015), mobility aid use (χ 2 = 6.573, p = 0.037), anxiety (H = 69.715, p < 0.001), depression (H = 63.605, p < 0.001), and family support (H = 19.971, p < 0.001) were significantly different in the distribution of the different developmental trajectory categories (Table 2). No significant differences were observed for the remaining demographic and clinical variables (all p > 0.05; Table S5).

TABLE 2.

Univariate analysis with different developmental trajectories.

Variable Category Low‐level fear (n = 88) Medium‐level stable decline (n = 44) High‐level continuous decline (n = 34) χ 2/H p
Age group (years), n (%) ≤ 55 41 (46.59) 8 (18.18) 1 (2.94) 43.214 < 0.001**
56–65 32 (36.36) 18 (40.91) 9 (26.47)
66–75 13 (14.77) 11 (25.00) 15 (44.12)
> 75 2 (2.28) 7 (15.91) 9 (26.47)
Educational level, n (%) Primary and below 40 (45.45) 6 (13.64) 11 (32.35) 16.612 0.034*
Junior high school 26 (29.55) 18 (40.91) 12 (35.29)
High school or technical secondary school 13 (14.77) 11 (25.00) 6 (17.65)
Junior college 3 (3.41) 4 (9.09) 4 (11.76)
Bachelor degree or above 6 (6.82) 5 (11.36) 1 (2.94)
Medical payment method, n (%) Employee medical insurance 26 (29.54) 26 (59.09) 16 (47.06) 12.368 0.015*
Medical insurance for urban and rural residents 61 (69.32) 17 (38.64) 17 (50.00)
Mobility aid use, n (%) No 76 (86.36) 30 (68.18) 25 (73.53) 6.573 0.037*
Yes 12 (13.64) 14 (31.82) 9 (26.47)
Anxiety, median (IQR) — 0.0 (0.0, 0.0) 3.0 (2.0, 4.0) 6.0 (0.0, 7.3) 69.715 < 0.001**
Depression, median (IQR) — 0.0 (0.0, 1.0) 5.0 (2.0, 6.0) 7.0 (1.8, 11.0) 63.605 < 0.001**
Family support, median (IQR) — 15.0 (14.0, 15.0) 15.0 (14.0, 15.0) 13.0 (12.0, 15.0) 19.971 < 0.001**
Open in a new tab
*

p < 0.05.

**

p < 0.001.

3.5. Multivariable Analysis of Fear of Falling Trajectory Predictors

Following a check for multicollinearity (all VIFs < 10), a multinomial logistic regression analysis was employed. This model was selected because the dependent variable—the three latent trajectory categories—is nominal rather than ordinal, a conclusion supported by a violated test of the parallel lines assumption (p < 0.05), indicating that the categories do not represent a hierarchy. The seven variables that were significant in the univariate analysis were entered as independent variables. The final model identified age group, anxiety, and depression as significant influences on the fear of falling trajectories (p < 0.05), with detailed results shown in Table 3.

TABLE 3.

Multiple logistic regression analysis with different developmental trajectories.

Dependent variables Independent variables β SE z Wald χ 2 p OR 95% CI
Low‐level fear vs. Medium‐ level stable decline Constant term −4.590 3.522 −1.303 1.698 0.193 0.010 0.000–10.109
Anxiety 0.602 0.155 3.895 15.174 < 0.001** 1.826 1.349–2.473
Depression 0.338 0.105 3.211 10.308 0.001* 1.401 1.141–1.722
Family support 0.095 0.216 0.44 0.194 0.660 1.100 0.720–1.680
Age group 0.818 0.302 2.706 7.324 0.007* 2.265 1.253–4.095
Educational level 0.341 0.253 1.349 1.819 0.177 1.406 0.857–2.308
Medical payment method −0.993 0.578 −1.717 2.946 0.086 0.371 0.119–1.151
Mobility aid −0.161 0.672 −0.24 0.058 0.810 0.851 0.228–3.175
Low‐level fear vs. High‐level continuous decline Constant term −2.890 3.823 −0.756 0.571 0.450 0.056 0.000–99.804
Anxiety 0.734 0.17 4.324 18.698 < 0.001** 2.083 1.493–2.904
Depression 0.401 0.121 3.314 10.983 0.001* 1.494 1.178–1.893
Family support −0.245 0.239 −1.022 1.045 0.307 0.783 0.490–1.251
Age group 1.639 0.392 4.176 17.435 < 0.001** 5.148 2.386–11.108
Educational level −0.016 0.317 −0.049 0.002 0.961 0.984 0.529–1.833
Medical payment method −0.576 0.69 −0.834 0.696 0.404 0.562 0.145–2.175
Mobility aid −0.757 0.817 −0.926 0.858 0.354 0.469 0.095–2.327
Open in a new tab
*

p < 0.05.

**

p < 0.001.

We found that for each unit increase in the patient's age group, the risk of changing from the “low‐level fear” to the “medium‐level steady decline” increased 2.265 times, and the risk of changing from the “low‐level fear” to the “high‐level continuous decline” increased 5.148 times.

Similarly, elevated levels of anxiety and depression were significantly associated with membership in more severe fear of falling trajectory groups. Using the “low‐level fear” group as the reference, each one‐level increase in anxiety was associated with 1.826 times greater odds of belonging to the “medium‐level steady decline” group (OR = 1.826, 95% CI [1.349–2.473], p < 0.001) and 2.083 times greater odds of belonging to the “high‐level continuous decline” group (OR = 2.083, 95% CI [1.493–2.904], p = 0.001). Similarly, each one‐level increase in depression was associated with 1.401 times greater odds of belonging to the “medium‐level steady decline” group (OR = 1.401, 95% CI [1.141–1.722], p = 0.001) and 1.494 times greater odds of belonging to the “high‐level continuous decline” group (OR = 1.494, 95% CI [1.178–1.893], p = 0.001).

4. Discussion

Our longitudinal analysis identified fear of falling prevalence rates of 48.8%, 49.4%, 45.8%, and 34.3% at four distinct phases of stroke rehabilitation. These findings align with Badrasawi et al.'s reported rate of 49.0% (Badrasawi et al. 2022) yet demonstrate marked divergence from Egyptian cohort data (64.4%) (Saleh et al. 2018), a discrepancy potentially attributable to variations in assessment timing.

Notably, the 1‐month post‐stroke period emerged as the peak fear of falling incidence, a nursing‐observed pattern likely reflecting the clinical reality that patients' expanding activity engagement during early nursing‐supervised recovery phases paradoxically elevates fall risk perception. This critical nursing intervention window may represent an optimal opportunity for nurse‐initiated fear mitigation strategies.

The observed temporal decline in fear of falling scores and proportion of high‐concern patients suggests progressive psychological adaptation throughout rehabilitation. Cross‐study comparisons reveal intriguing patterns: while Chen et al.'s Chinese cohort (Chen et al. 2023) documented 8.94% persistent high fear of falling levels and 18.21% moderate‐level persistence beyond 6 months post‐stroke, our cohort demonstrated complete resolution of high‐level fear of falling (0%) with moderate‐level persistence (34.34%) at this timepoint. Though sample characteristics may underlie these differences, both studies converge on the critical finding that moderate fear of falling substantially outweighs severe manifestations during rehabilitation.

Through LCGM modeling, we delineated three distinct fear of falling trajectory classes—a categorization schema consistent with prior work documenting 3–4 trajectory patterns (Liu et al. 2022; Noimontree 2017; Yang et al. 2025). The largest subgroup (53.01%) exhibited sustained low‐level fear trajectories, suggesting either inherent psychological resilience or the benefits of early nursing‐supported interventions. These patients maintained stable low fear of falling despite increasing rehabilitation demands, which may reflect successful integration of nursing‐based fear mitigation strategies, thereby promoting better therapeutic engagement and functional recovery.

The similarity between the “medium‐level steady decline” and the “high‐level continuous decline” lies in the fact that patients in both trajectories had higher levels of fear of falling in the early stages of recovery (compared to the “low‐level fear”), but their fear of falling declined over time, with the “high‐level continuous decline” experiencing a more rapid decline in fear of falling. These patients may have had more severe fear of falling in the early stage of rehabilitation due to their experience of falling or impaired muscle strength due to stroke, but as the rehabilitation and treatment progressed, the patients acquired more knowledge of rehabilitation exercises and were able to carry out limb function training in a more rational way, thus significantly reducing their fear of falling.

As mentioned earlier, age was identified as an important factor influencing fear of falling trajectories after stroke, which is consistent with the findings of Badrasawi et al. (2022) and Saleh et al. (2018). In our study, increasing age was associated with a substantially greater likelihood of belonging to the “medium‐level steady decline” and “high‐level continuous decline” trajectory groups compared with the “low‐level fear” group. One possible explanation is that age‐related declines in muscle strength, postural control, mobility, and sensory integration may impair balance confidence and increase perceived vulnerability to falling (Rodrigues et al. 2023). Older stroke patients may therefore require a longer period to regain functional independence during rehabilitation, contributing to persistently elevated fear of falling over time. Similar findings were reported by Noimontree (2017), who observed that older individuals were more likely to exhibit chronic or fluctuating fear of falling patterns.

Anxiety was also found to be significantly associated with fear of falling in previous studies (Schmid et al. 2011). Our study extends this evidence by demonstrating that higher anxiety levels are associated with increased odds of belonging to more severe fear of falling trajectory groups. A possible mechanism is that anxiety may heighten attentional focus on bodily instability and increase threat perception during movement, leading patients to overestimate their risk of falling (Young and Williams 2015). Such heightened vigilance may subsequently promote activity avoidance and reduced participation in rehabilitation exercises. Over time, reduced mobility experience and limited exposure to successful movement performance may reinforce low balance confidence and contribute to persistent fear of falling trajectories (Chen et al. 2023).

Depression was similarly associated with more severe fear of falling trajectories, which aligns with previous reports identifying depressive symptoms as predictors of fear of falling (Denkinger et al. 2015; Hoang et al. 2017; Kim and So 2013; Shin et al. 2010). Stroke patients with depressive symptoms may exhibit lower motivation, diminished self‐efficacy, and reduced engagement in rehabilitation activities, thereby limiting opportunities to rebuild physical confidence and functional independence (Gnanaprakasam et al. 2024). Furthermore, prolonged psychological distress may impair rehabilitation participation and functional recovery, indirectly maintaining elevated fear of falling levels over time.

Taken together, these findings suggest that fear of falling trajectories after stroke are influenced not only by physical recovery but also by dynamic psychological processes. Early identification of older patients and those with anxiety or depressive symptoms may help clinicians implement targeted psychological support and individualized rehabilitation strategies to prevent persistent fear of falling during stroke recovery.

5. Limitations

While elucidating critical fear of falling trajectory patterns, this study has several constraints. First, the study was conducted in only two urban tertiary hospitals, which may limit generalizability to rural populations or other healthcare systems. Future multi‐center studies are needed. Second, important functional and rehabilitation‐related variables, such as balance performance, gait ability, rehabilitation content, and detailed fall history, were not comprehensively assessed in this study. These factors may directly influence fear of falling trajectories and may also be associated with psychological symptoms such as anxiety and depression, potentially affecting the interpretation of the observed relationships. Third, the cohort was predominantly composed of ischemic stroke patients (91.6%), while hemorrhagic stroke patients accounted for only a small proportion of the sample. Given the distinct pathophysiological and recovery characteristics of different stroke subtypes, the identified trajectories may primarily reflect patterns among ischemic stroke patients. Therefore, caution is warranted when generalizing these findings to all stroke populations. Additionally, the prevalence of fear of falling was assessed using a single‐item binary question, which has limited reliability and cannot capture severity gradations. This may have influenced the reported prevalence estimates. However, trajectory modeling was based on the continuous Short FES‐I scores, which is a validated multi‐item scale. Future studies should consider using the full 16‐item FES‐I for more comprehensive psychometric assessment.

6. Conclusions and Implications

The fear of falling of stroke survivors during rehabilitation gradually decreased with time, but there was significant individual heterogeneity in the development trajectory of fear of falling. Nursing assessments should prioritize older patients and those presenting anxiety or depression symptoms, as these represent clinically significant predictors of persistent fear of falling that warrant nurse‐initiated interventions. The implementation of nursing‐led, age‐adjusted rehabilitation programs incorporating psychological support principles can effectively optimize fear reduction while sustaining therapeutic engagement throughout the recovery continuum. This approach aligns with contemporary nursing practice models emphasizing individualized, biopsychosocial care strategies for optimal post‐stroke outcomes.

7. Clinical Resources

American Stroke Association: Life After Stroke: Preventing Falls (https://www.stroke.org/en/life‐after‐stroke/preventing‐another‐stroke).

Centers for Disease Control and Prevention (CDC): STEADI—Older Adult Fall Prevention (https://www.cdc.gov/steadi/index.html).

Funding

This work was supported by the Health Industry Scientific Research Program of Gansu Province [grant number: GSWSHL2023‐02].

Disclosure

During the preparation of this manuscript, the authors used AI‐assisted tools for language refinement and grammar checking. All scientific content, data analysis, interpretation, and final manuscript approval were the sole responsibility of the authors.

Ethics Statement

All methods were performed in accordance with the declaration of Helsinki. The study was approved by the Ethics Committee of Lanzhou University School of Nursing (LZUHLXY20230035). Informed consent was obtained from all participants involved in the study.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Figure S1: Flowchart of study patients.

Table S1: Baseline characteristics by follow‐up status.

Table S2: Levels of concern about fear of falling over time.

Table S3: Quality of latent class classification.

Table S4: Mean intercept and slope estimates for fear of falling developmental trajectories.

Table S5: Univariate analysis with different developmental trajectories.

JNU-58-0-s001.docx (125.8KB, docx)

Contributor Information

Fangli Ma, Email: fang.mary@163.com.

Ying Ren, Email: 1192142874@qq.com.

Zhigang Zhang, Email: zzg3444@163.com.

Data Availability Statement

The datasets used during the study are available from the author upon reasonable request.

References

  1. Arfken, C. L. , Lach H. W., Birge S. J., and Miller J. P.. 1994. “The Prevalence and Correlates of Fear of Falling in Elderly Persons Living in the Community.” American Journal of Public Health 84, no. 4: 565–570. 10.2105/ajph.84.4.565. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Badrasawi, M. , Hamdan M., Vanoh D., Zidan S., Alsaeid T., and Muhtaseb T. B.. 2022. “Predictors of Fear of Falling Among Community‐Dwelling Older Adults: Cross‐Sectional Study From Palestine.” PLoS One 17, no. 11: e0276967. 10.1371/journal.pone.0276967. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Chen, Y. , Du H., Song M., et al. 2023. “Relationship Between Fear of Falling and Fall Risk Among Older Patients With Stroke: A Structural Equation Modeling.” BMC Geriatrics 23, no. 1: 647. 10.1186/s12877-023-04298-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Corbin, J. M. 1998. “The Corbin and Strauss Chronic Illness Trajectory Model: An Update.” Scholarly Inquiry for Nursing Practice 12, no. 1: 33–41. [PubMed] [Google Scholar]
  5. Cumming, R. G. , Salkeld G., Thomas M., and Szonyi G.. 2000. “Prospective Study of the Impact of Fear of Falling on Activities of Daily Living, SF‐36 Scores, and Nursing Home Admission.” Journals of Gerontology. Series A, Biological Sciences and Medical Sciences 55, no. 5: M299–M305. 10.1093/gerona/55.5.m299. [DOI] [PubMed] [Google Scholar]
  6. Delbaere, K. , Close J. C., Mikolaizak A. S., Sachdev P. S., Brodaty H., and Lord S. R.. 2010. “The Falls Efficacy Scale International (FES‐I). A Comprehensive Longitudinal Validation Study.” Age and Ageing 39, no. 2: 210–216. 10.1093/ageing/afp225. [DOI] [PubMed] [Google Scholar]
  7. Deng, N. , Zhang T., and Shi B. X.. 2015. “Reliability and Validity of Chinese Version of Short Falls Efficacy Scale International in Patients With Cerebral Infarction.” Chinese Journal of Rehabilitation Theory and Practice 21, no. 12: 1438–1442. [Google Scholar]
  8. Denkinger, M. D. , Lukas A., Nikolaus T., and Hauer K.. 2015. “Factors Associated With Fear of Falling and Associated Activity Restriction in Community‐Dwelling Older Adults: A Systematic Review.” American Journal of Geriatric Psychiatry 23, no. 1: 72–86. 10.1016/j.jagp.2014.03.002. [DOI] [PubMed] [Google Scholar]
  9. Folstein, M. F. , Folstein S. E., and McHugh P. R.. 1975. “‘Mini‐Mental State’. A Practical Method for Grading the Cognitive State of Patients for the Clinician.” Journal of Psychiatric Research 12, no. 3: 189–198. 10.1016/0022-3956(75)90026-6. [DOI] [PubMed] [Google Scholar]
  10. GBD 2019 Stroke Collaborators . 2021. “Global, Regional, and National Burden of Stroke and Its Risk Factors, 1990‐2019: A Systematic Analysis for the Global Burden of Disease Study 2019.” Lancet Neurology 20, no. 10: 795–820. 10.1016/s1474-4422(21)00252-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Gnanaprakasam, A. , Solomon J. M., Roy A. K., Deshmukh A. S., and Karthikbabu S.. 2024. “Association Between Depression and Adherence to Upper Limb Exercises Among Community‐Dwelling Stroke Survivors: A Cross‐Sectional Study.” Health Science Reports 7, no. 10: e70133. 10.1002/hsr2.70133. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Hoang, O. T. , Jullamate P., Piphatvanitcha N., and Rosenberg E.. 2017. “Factors Related to Fear of Falling Among Community‐Dwelling Older Adults.” Journal of Clinical Nursing 26, no. 1–2: 68–76. 10.1111/jocn.13337. [DOI] [PubMed] [Google Scholar]
  13. Jiang, Z. 2008. The Application of Statistical Analysis in Medical Research Projects. People's Medical Publishing House. [Google Scholar]
  14. Jung, T. , and Wickrama K. A. S.. 2008. “An Introduction to Latent Class Growth Analysis and Growth Mixture Modeling.” Social and Personality Psychology Compass 2, no. 1: 302–317. 10.1111/j.1751-9004.2007.00054.x. [DOI] [Google Scholar]
  15. Kempen, G. I. , Yardley L., van Haastregt J. C., et al. 2008. “The Short FES‐I: A Shortened Version of the Falls Efficacy Scale‐International to Assess Fear of Falling.” Age and Ageing 37, no. 1: 45–50. 10.1093/ageing/afm157. [DOI] [PubMed] [Google Scholar]
  16. Kim, S. , and So W. Y.. 2013. “Prevalence and Correlates of Fear of Falling in Korean Community‐Dwelling Elderly Subjects.” Experimental Gerontology 48, no. 11: 1323–1328. 10.1016/j.exger.2013.08.015. [DOI] [PubMed] [Google Scholar]
  17. Lachman, M. E. , Howland J., Tennstedt S., Jette A., Assmann S., and Peterson E. W.. 1998. “Fear of Falling and Activity Restriction: The Survey of Activities and Fear of Falling in the Elderly (SAFE).” Journals of Gerontology. Series B, Psychological Sciences and Social Sciences 53, no. 1: P43–P50. 10.1093/geronb/53b.1.p43. [DOI] [PubMed] [Google Scholar]
  18. Liu, M. , Liang F., Yin Y., et al. 2022. “Trajectory of Change in Fear of Falling and Its Influencing Factors Among Stroke Patients.” Journal of Nursing Science 37, no. 23: 67–71. [Google Scholar]
  19. Noimontree, W. 2017. “Fear of Falling in Older Adults: A Trajectory Analysis. [Doctoral Dissertation, Saint Louis University].”
  20. Painter, J. A. , Allison L., Dhingra P., Daughtery J., Cogdill K., and Trujillo L. G.. 2012. “Fear of Falling and Its Relationship With Anxiety, Depression, and Activity Engagement Among Community‐Dwelling Older Adults.” American Journal of Occupational Therapy 66, no. 2: 169–176. 10.5014/ajot.2012.002535. [DOI] [PubMed] [Google Scholar]
  21. Procidano, M. E. , and Heller K.. 1983. “Measures of Perceived Social Support From Friends and From Family: Three Validation Studies.” American Journal of Community Psychology 11, no. 1: 1–24. 10.1007/bf00898416. [DOI] [PubMed] [Google Scholar]
  22. Rodrigues, F. , Monteiro A. M., Forte P., and Morouço P.. 2023. “Effects of Muscle Strength, Agility, and Fear of Falling on Risk of Falling in Older Adults.” International Journal of Environmental Research and Public Health 20, no. 6: 4945. 10.3390/ijerph20064945. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Saleh, N. M. A. , Ibrahim H. S., Mohamed H., and El‐Gilany A.‐H.. 2018. “Predictors of Fear of Falling Among Community Dwelling Older Adults in Mansoura City, Egypt.” International Journal of Nursing Didactics 8, no. 8: 2327. 10.15520/ijnd.v8i08.2327. [DOI] [Google Scholar]
  24. Scheffer, A. C. , Schuurmans M. J., van Dijk N., van der Hooft T., and de Rooij S. E.. 2008. “Fear of Falling: Measurement Strategy, Prevalence, Risk Factors and Consequences Among Older Persons.” Age and Ageing 37, no. 1: 19–24. 10.1093/ageing/afm169. [DOI] [PubMed] [Google Scholar]
  25. Schmid, A. A. , Van Puymbroeck M., Knies K., et al. 2011. “Fear of Falling Among People Who Have Sustained a Stroke: A 6‐Month Longitudinal Pilot Study.” American Journal of Occupational Therapy 65, no. 2: 125–132. 10.5014/ajot.2011.000737. [DOI] [PubMed] [Google Scholar]
  26. Schoene, D. , Heller C., Aung Y. N., Sieber C. C., Kemmler W., and Freiberger E.. 2019. “A Systematic Review on the Influence of Fear of Falling on Quality of Life in Older People: Is There a Role for Falls?” Clinical Interventions in Aging 14: 701–719. 10.2147/cia.S197857. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Shin, K. R. , Kang Y., Kim M. Y., et al. 2010. “Impact of Depression and Activities of Daily Living on the Fear of Falling in Korean Community‐Dwelling Elderly.” Nursing & Health Sciences 12, no. 4: 493–498. 10.1111/j.1442-2018.2010.00567.x. [DOI] [PubMed] [Google Scholar]
  28. Skrivankova, V. W. , Richmond R. C., Woolf B. A. R., et al. 2021. “Strengthening the Reporting of Observational Studies in Epidemiology Using Mendelian Randomization: The STROBE‐MR Statement.” JAMA 326, no. 16: 1614–1621. 10.1001/jama.2021.18236. [DOI] [PubMed] [Google Scholar]
  29. Tian, X. , Mai Y. H., Guo Z. J., Chen J. W., and Zhou L. J.. 2024. “Contributing Factors and Interventions for Fear of Falling in Stroke Survivors: A Systematic Review.” Topics in Stroke Rehabilitation 31, no. 8: 772–787. 10.1080/10749357.2024.2333172. [DOI] [PubMed] [Google Scholar]
  30. Wang, Y. , Chen Y., and Zhang L.. 2018. “Evaluation of the Reliability and Validity of the Generalized Anxiety Disorder 7‐Item Scale Among Inpatients in General Hospital.” Journal of Clinical Psychiatry 28, no. 3: 168–171. [Google Scholar]
  31. Xie, M. , Taylor‐Piliae R. E., Yang C., Peng X., Yang Q., and Zhang Q.. 2025. “Risk Factors for Falls and Recurrent Falls in Older Stroke Survivors: A Systematic Review and Meta‐Analysis of Prospective Studies.” International Journal of Older People Nursing 20, no. 6: e70050. 10.1111/opn.70050. [DOI] [PubMed] [Google Scholar]
  32. Xu, Y. , Wu H., and Xu Y.. 2007. “The Reliability and Validity of Patient Health Questionnaire Depression Module (PHQ‐9) in Chinese Elderly.” Shanghai Archives of Psychiatry 05: 257–259. [Google Scholar]
  33. Yang, Y. , Li M., Ding X., Zhang L., Jin M., and Jin Y.. 2025. “Classification and Influencing Factors of Fear of Falling in Elderly Stroke Patients: A Latent Profile Analysis.” Aging Clinical and Experimental Research 37, no. 1: 329. 10.1007/s40520-025-03236-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Young, W. R. , and Williams A. M.. 2015. “How Fear of Falling Can Increase Fall‐Risk in Older Adults: Applying Psychological Theory to Practical Observations.” Gait & Posture 41, no. 1: 7–12. 10.1016/j.gaitpost.2014.09.006. [DOI] [PubMed] [Google Scholar]
  35. Zeng, Q. Z. , He Y. L., Liu H., et al. 2013. “Reliability and Validity of Chinese Version of the Generalized Anxiety Disorder 7‐Item (GAD‐7) Scale in Screening Anxiety Disorders in Outpatients From Traditional Chinese Internal Department.” Chinese Mental Health Journal 27, no. 3: 163–168. [Google Scholar]
  36. Zhang, J. P. , and Liu H. R.. 2001. “Family Support and Quality of Life Among Hemodialysis Patients.” Journal of Central South University (Medical Science) 26, no. 4: 359–362. [PubMed] [Google Scholar]
  37. Zhang, M. Y. , and Wang Z. Y.. 1989. “Application of the Chinese Version of the Mini‐Mental State Examination (MMSE).” Shanghai Archives of Psychiatry 7, no. 3: 108–111. [Google Scholar]
  38. Zheng, T. , Shi Y. Z., Zhang N., et al. 2013. “Reliability and Validity of the Patient Health Questionnaire‐9 in Stroke Patients With Depression.” Beijing Medical Journal 35, no. 5: 340–343. [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Figure S1: Flowchart of study patients.

Table S1: Baseline characteristics by follow‐up status.

Table S2: Levels of concern about fear of falling over time.

Table S3: Quality of latent class classification.

Table S4: Mean intercept and slope estimates for fear of falling developmental trajectories.

Table S5: Univariate analysis with different developmental trajectories.

JNU-58-0-s001.docx (125.8KB, docx)

Data Availability Statement

The datasets used during the study are available from the author upon reasonable request.

Articles from Journal of Nursing Scholarship are provided here courtesy of Wiley

RESOURCES