This study was a cross-sectional design. Convenience sampling was utilized to select eligible older patients who had undergone PCI from March 2024 to June 2025 at three comprehensive class a hospitals’ cardiology centers along the eastern coast of China. Inclusion criteria encompassed: (i) patients aged over 60 years; (ii) had undergone PCI with successful procedural outcomes; (iii) had signed informed consent. Exclusion criteria included concurrent malignancies, severe hepatic or renal impairment, active infections or other significant organ failures; communication barriers; activity limitations; cognitive impairment.

Based on conical model and literature reviews, we selected 27 possible influencing factors (23). According to Kendall’s cross-sectional sample size estimation method, the sample size should be 5 to 10 times the number of independent variables, considering a likely attrition rate of 10% and sampling error, the appropriate recommended sample size was range of 149 to 297 cases (25). Besides, to ensure the stability of the Latent Class Analysis (LCA) model (26), sample size exceeding 300 cases is recommended. In total, we successfully collected data from 315 participants.

Under stable vital sign in post-PCI patients, we conducted the research. Prior to distributing questionnaires, we explained the research purpose to ensure informed consent. Some data of demographic and physical were collected through face-to-face interviews supplemented by patient’s clinical case documentation. Six months after PCI, telephone follow-ups were conducted to complete the LSA, DASI, PHQ-9, CDSES, and SRSS questionnaires. In total, among 329 CHD patients we approached, data from 315 participants were collected successfully, representing a response rate of 95.7%.

This study has been approved by the Ethics Committee of Shanghai University of Traditional Chinese Medicine (approval number: 2024-1-16-10). All participants were informed of the study’s content and objectives prior to enrolment and provided written informed consent. All data collected in this research adhere to the principles of the Declaration of Helsinki, ensuring confidentiality and data integrity.

Latent class analysis (LCA) was performed with Mplus version 8.3 to identify subgroups of individuals based on life-space levels. The analysis began with a model specifying one latent class and incrementally increased the number of classes for model fitting estimation. Model fit was evaluated using the Akaike information criterion (AIC) (39), Bayesian information criterion (BIC) (40), and adjusted BIC (aBIC) (41), with lower values indicating better fit. The Lo–Mendell–Rubin adjusted likelihood ratio test (LMRT) and the bootstrapped likelihood ratio test (BLRT) were used to compare adjacent models (42); a p-value < 0.05 suggested a significantly improved fit with an additional class. Entropy was employed to assess classification accuracy, where higher values reflect better classification.

Statistical analyses of variables were conducted using SPSS version 26.0. Continuous variables are presented as mean ± standard deviation, and categorical variables as frequency and percentage. After identifying different latent classes via LCA, group comparisons were carried out using the χ² test (or Fisher’s exact test for cell counts < 5) for categorical variables, ANOVA for continuous variables. Effect analysis is used to determine the true effect size of difference analysis p-value, whether there is a true difference and its magnitude, effect size of χ² test is Cramér’s V, effect size of ANOVA is η². According to the standards proposed by Cohen (43), the criteria for effect sizes in ANOVA are as follows: 0.01 indicates a small effect, 0.06 indicates a medium effect, and 0.14 indicates a large effect. For χ² test, according to Cohen’s (44) guidelines for interpreting Cramér’s V coefficient, the standards are as follows: when df* = 2: 0.07 < V < 0.21 indicates a small effect, 0.21 < V < 0.35 indicates a medium effect, and V > 0.35 indicates a large effect, when df* = 3: 0.06 < V < 0.17 indicates a small effect, 0.17 < V < 0.29 indicates a medium effect, and V > 0.29 indicates a large effect, and so forth. Multinomial logistic regression was utilized to explore influence factors associated with latent class of life-space. To ensure the reliability of the regression model, a likelihood ratio test is conducted on the model, if p < 0.05 indicates that the model is effective as a whole. The explanatory ability of the model is achieved by referring to Pseudo R-Squared, which includes Cox & Snell R², McFadden R, and Nagelkerke R. The values of these metrics range from 0 to 1, the larger the value, the better the model can explain the variation of the dependent variable (45). p < 0.05 was considered statistically significant.

A total of 315 older patients undergoing PCI were included in the study. Demographically, the population was predominantly male (69.2%), primarily of Han ethnicity (98.7%), and mostly married (91.7%). Economically, the majority of participants reported a monthly household income ranging from 3,001 to 10,000 RMB (78.4%). Physically, hypertension prevalence was notably high at 72.7%. The proportion of patients undergoing emergency procedures was greater than those undergoing elective surgery (57.5% vs. 42.5%), with most receiving single-stent implantation (63.8%). The mean score of DASI was (34.30 ± 10.68). Psychologically, the mean scores of depressions and self-efficacy were (7.07 ± 3.82), (5.72 ± 1.78) respectively. Social and family environmentally, 40.3% of patients assumed caregiving responsibilities at home, and the mean score of social support was (41.24 ± 4.28).

Four models were ultimately fitted in this study, with fit indices presented in Table 1. When the number of classes n > 3, the decline in model fit indices became less pronounced as the class number increased. Among the models evaluated, the three-class model showed the lowest AIC and aBIC, significant LMR and BLRT p-values (<0.05), and an entropy of 0.942, thus being selected as the optimal solution.

Table 2 and Figure 1 show the restricted status of patients’ life-space, we named three latent classes according to their restriction levels. Class 1 (Daily Activity Group): Levels 1–4 are unrestricted; Level 5 indicates activities outside of town. Despite a high probability of restriction at Level 5, the mean scores across the three categories remain highest, indicating patients can generally perform routine activities; thus, designated as Daily Activity Group. Class 2 (Community Exercise Group): Levels 1–2 are unrestricted, with Level 3 representing intra-neighborhood activities, which have a restriction-free probability exceeding 90%, though some restrictions persist. Higher levels (4, 5) show lower scores and complete restrictions. Patients in this category primarily experience activity limitations within the community, hence classified as Community Exercise Group. Class 3(Home Maintenance Group): Only Levels 1–2, representing activities within the home and residential buildings, are unrestricted; Levels 3–5 outside the neighborhood are restricted, and all dimension scores in this class are the lowest among the three. The life-space is confined to the home, leading to the designation as the Home Maintenance Group. The proportions of these three classes are as follows: C1: 29.8% (n = 94), C2: 49.5% (n = 156), and C3: 20.6% (n = 65).

Table 3 presents an overview of demographic characteristics, physical health status, psychosocial factors, and environmental conditions across different patient subgroups, highlighting notable differences. Chi-square tests and univariate analysis of variance reveal that variables such as age, educational level, current employment status, diabetes history, DASI, self-efficacy, depression, social support and responsibility for family care, the number of places reachable within 15 min, safety hazards in life-space, residential address post-discharge significantly influence the distribution of patients across different life-space subgroups (p < 0.05). Based on the effect size criteria, in the univariate analysis, the effect sizes of age, Educational level, employment status, social support, number of places reachable within 15 min, safety hazards in the life space, residential address post-discharge are small effects, while the effect sizes of the other significant variables are medium effects.

Using the three latent classes as dependent variables (C1: Daily Activities Group; C2: In-Community Exercise Group; C3: Home Maintenance Group), multinomial logistic regression analysis was conducted with variables showing statistical significance in univariate analysis as independent variables. Although our effect size analysis showed that some variables that were significant in the univariate comparison did not have large effect sizes, considering the combination with the conical theory model, we still included them in the regression analysis. The chi-square value of the likelihood ratio test was 190.176 (p < 0.001), which indicates that the model as a whole is significantly effective. The Nagelkerke R² was 0.519, the McFadden pseudo-R² was 0.292, and the Cox & Snell R² was 0.453, suggesting moderate explanatory power of the model.

In the regression analysis, variables with smaller effect sizes were all excluded, except for age. The results revealed significant associations between different life-space subgroups and selected variables (Table 4). Compared to C1 (Daily Activities group), patients with reduced DASI (p = 0.01, OR = 0.95, 95% CI: 0.924 ~ 0.980), and lower self-efficacy (p = 0.022, OR = 0.771, 95% CI: 0.617 ~ 0.964) were more likely to belong to C2 (Community Exercise group). Conversely, younger age (p = 0.004, OR = 0.865, 95% CI:0.783 ~ 0.955), diabetes history (p = 0.036, OR = 0.368, 95% CI: 0.144 ~ 0.937), reduced DASI (p = 0.004, OR = 0.86, 95% CI: 0.783 ~ 0.955), lower self-efficacy (p = 0.030, OR = 0.695, 95% CI: 0.501 ~ 0.964), and higher depression (p = 0.003, OR = 1.253, 95% CI: 1.079 ~ 1.455) increased the likelihood of belonging to C3 (Home Maintenance group).

The results of this study indicate that total mean scores of life-space of older patients at 6 months after PCI was 68.39 ± 16.25, which proved a generally intermediate level because of exceeded 60 points. However, analysis across various dimensions revealed a decline in life-space with the expansion of the activity range, accompanied by notable restrictions during broader activities such as traveling to places outside town. Consistent with previous research (22), this gradient phenomenon, from near to far, indicates the possibility of resuming small to medium-range domains daily activities in older patients after PCI, yet with great challenge in engaging in long-distance activities. In our speculation, it may be attributed to the objective physiological status and subjective willingness of patients. Patients who undergo early rehabilitation may be able to perform small to medium-range domains activities more effectively at 6 months post-discharge as they experience significant improvement in cardiac function and quality of life after intervention (46). However, activities beyond the neighborhood, involving larger physical exertion and greater cardiovascular demands, pose substantial challenges for these patients. Furthermore, there is a need to consider patients’ subjective willingness to engage in activities, in addition to their objective physiological status (47). For instance, for remote activities such as community outings, it has been reported that patients often require accompaniment from family members or caregivers (48). The level of familial support for outdoor excursions may have an impact on the patient’s subjective motivation, with some patients potentially avoiding such activities due to absence of companionship.

Using LCA, this study identified three distinct latent subgroups of different dimensions life-space, including the daily activity group (C1), the community exercise group (C2), and the home maintenance group (C3). Patients in C1 group can perform activities outside the neighborhood relatively well, with only some restrictions within other districts of the town. For Chinese older adults, activities outside the community are mostly concentrated on medical appointments, routine shopping, and fee payments. Greater autonomy and social participation were observed in C1 group. Furthermore, the life-space in C2 group was mainly limited to within the neighborhood. In this range, older adults typically engage in leisure and rehabilitation activities (e.g., physical exercise, neighbourly visits, etc.), indicating a moderate level of life-space with room for improvement (49). In comparison, patients in C3 group were predominantly confined to their homes, with significantly reduced activity ranges below the average for patients with cardiovascular diseases (21). Moreover, these individuals also exhibited the lowest cardiopulmonary function and self-efficacy scores, but the highest depression levels, representing poorer rehabilitation outcomes. The rehabilitation outcome exemplified by patients in C1 group aligned more closely with the objective of cardiac rehabilitation, but approximately 70% of patients in this study were categorized into the C2 and C3 groups. Noticeably, older patients remain at elevated risk for recurrent CHD 6 months post-PCI (50). Therefore, Investigating the causes of varying life-space conditions and identifying strategies to enhance rehabilitation outcomes has become a primary focus of current research. In this context, our study also analyzes the associated factors influencing the differentiation among various life-space subgroups, with the objective of exploring possibility of facilitating adjustments within these subgroups and to support the efficacy of cardiac rehabilitation.

For the C3 group, age, diabetic history, and depression were independent factors affecting the transition of patients into the C1 group. Prior research (22) has documented the relationship of greater susceptibility to restricted life-space in older subjects. However, in our study, advanced-age patients demonstrated a higher likelihood of being classified into C1, which may be related to their frequent medical visits prompted by multimorbidity following PCI, which belongs to a passive expansion of life-space levels, rather than active engagement. As evidenced previously, the multimorbidity rate in the 60 ~ 69 age group was obviously lower than in those over 70 (51). Indirectly, compared to older counterparts, younger older patients have a lower outpatient medical consultation demand and may be influenced by their subjective activity intentions, favoring home-based maintenance. Therefore, it is critical to monitor older patients during assessment and follow-up for finding the real reason caused activity limitations, thereby facilitating a transition toward active engagement. However, the effect size of age in the univariate analysis is relatively small, which indicates that the impact of age on life-space remains a subject of debate, necessitating further verification. History of diabetes and depression were also key factors influencing the likelihood of classification into C3. Patients with diabetes experience impaired activity capacity as they are prone to develop complications such as motor dysfunction, diabetic foot ulcers, and medication-induced hypoglycemia. Approximately 45% of older diabetic patients had mobility impairments, (52) with an increased risk of falls. Additionally, patients usually occur recurrent symptoms (e.g., chest tightness, dull pain, and fatigue) within 6 months post-PCI. The physiological alterations after symptoms can precipitate psychological fluctuation (53), then progressively evolving anxiety and depressive states. Depression, in turn, can reduce physical activities and social participation (54). This interaction creates a bidirectional cycle of physiological and psychological impairment that may lead patients to avoid social and community involvement, further constricting patients’ life-space. Therefore, during six-month post-PCI, it is essential to conduct psychological assessments and interventions, in addition to addressing physical symptom management. Above conclusions and relevant literature, which underscores the requirement for linking physiological recovery, depression risk reduction, and life-space resumption. Medical staff may try to start by alleviating patients’ fears related to extensive activities to promote the formation of a positive cycle in all three aspects.

Simultaneously, DASI and self-efficacy were found to be common influencing factors for the transition from C2 and C3 to C1. DASI is a physiological indicator of cardiopulmonary function. In this study, lower scores on the index correlated with a higher likelihood of classification into C3 or C2 groups and greater propensity for categorizing into C3 than C2. Patients with reduced exercise tolerance tend to have limited activity ranges; and Tsai et al. (55) confirmed in their research that moderate to high-intensity activities occurred predominantly outside the home, and insufficient activity would exacerbate spatial confinement. In general, broader activity ranges are desirable goals during the six-month post-PCI rehabilitation, necessitating a foundational level of cardiopulmonary function. Concerning the mean DASI of 39.40 ± 10.75 points among Chinese patients with cardiovascular diseases reported previously (56), in our study, patients in C2 and C3 groups scored below this threshold, indicating significant functional deficits In general, broader activity ranges are desirable goals during the six-month post-PCI rehabilitation but necessitating a foundational level of cardiopulmonary function (21). Meanwhile, actively expanding the scope of life space is also conducive to improving cardiopulmonary function. This is indirectly that the three subgroups (C1, C2, and C3) can be transformed into each other. Chronic disease self-efficacy refers to the subjective confidence and perceived ability of patients to effectively control their condition and address disease-related challenges (57). In this study, patients who had lower scores were more likely to be categorized into C2 and C3 compared with C1. In other study (58) demonstrated positive correlations of self-efficacy with both mobility and activity levels; According to previous exploration, self-efficacy would exert a remarkable impact on health behaviors (59), quality of life, and psychological well-being among older patients with cardiovascular diseases. Patients exhibiting higher self-efficacy are more likely to engage in regular exercise and adhere to a healthy diet, which may subsequently affect their life-space. Therefore, self-efficacy may be a predictor of spatial heterogeneity in life space. In future care, perhaps the assessment of self-efficacy can be focused on in the early stage of rehabilitation, thereby having a certain grasp of the development of the patient’s future life space. In conclusion, both of them have certain influences on the heterogeneity of life space. The DASI score represents the patient’s physiological base level, while self-efficacy represents the patient’s intrinsic confidence in rehabilitation. During the subsequent cardiac rehabilitation process, this indicates that it is appropriate for medical staff to consider the rehabilitation after PCI from the perspective of balancing the patient’s physiological basis and psychological state, which can avoid overconfidence or excessive fear.