Psychological resilience and self-control in individuals at high risk of stroke: A mental health-focused study

Abstract

BACKGROUND

For individuals at a high risk of stroke, cardiovascular risk comprises both biomedical and psychological factors. The relationship between health risk and sustained health risk can be driven by behavior and self risk.

AIM

To investigate whether psychological resilience mediates relationships between active health behaviors and self-control in individuals at high risk of stroke.

METHODS

A cross-sectional study was conducted from January 2024 to December 2025 involving 1000 individuals at a high risk of stroke recruited from several area institutions. Participants completed the Scale for Proactive Health Behavior in Chronic Diseases, 10-point Connor-Davidson Resilience Scale, Stroke Self-monitoring Scale, and 7-point Generalized Anxiety Disorder Scale to assess proactive health behavior, psychological stability, self-control, and anxiety symptoms, respectively. Mediation analysis was performed using the PROCESS Model 4 macro, considering demographic variables.

RESULTS

The average age of the participants was 58.34 ± 9.76 years (54.2%). Proactive health behaviors were positively correlated with both psychological resilience (r = 0.562, P < 0.001) and self-control (r = 0.647, P < 0.001). A total of 34.6% of the participants showed mild anxiety (7-point Generalized Anxiety Disorder Scale score 5-9), and psychological distress was inversely correlated with resilience (r = -0.412, P < 0.001) and self-control (r = -0.389, P < 0.001). Furthermore, psychological resilience significantly mediated the association between proactive health behaviors and self-control (indirect effect = 0.198, 95% confidence interval: 0.165-0.234), accounting for 28.2% of the overall effect. The direct effect of an active health approach on self-control remained significant (β = 0.374, P < 0.001).

CONCLUSION

Psychological resilience functions as a pathway to mental health, proactive health behaviors, and risk-based self-control. Resilience-building, routine mental health screening, and development of stress tolerance are important for at-risk populations.

Key Words: Psychological stability; Subclinical psychological stress; Results of mental functioning; Stroke prevention; Self-management/control; Executive functions; Anxiety; Psychological resilience

Core Tip: This study measured psychological resilience, proactive health behaviors, self-control abilities, and the presence of anxiety in participants at a high risk of stroke. The results showed positive correlations between proactive health behaviors and resilience and between proactive health behaviors and self-control. Minimal-mild anxiety was common among the participants. Mediation analysis showed that resilience partially explained this relationship. These results suggest that stroke prevention strategies should go beyond traditional health education and incorporate evidence-based resilience-enhancing interventions for individuals at a high risk of stroke.

INTRODUCTION

According to statistics from the World Health Organization. The global incidence of stroke is high. In addition, stroke carries a high mortality rate and often results in major disability. Individuals at a high risk of stroke have long-term risk factors which require ongoing monitoring. These individuals may experience anxiety caused by the need to monitor their symptoms, make treatment decisions, and cope with family pressures. Moreover, individuals who have experienced a stroke essentially have daily chores involving performing the same stroke-related control actions repeatedly. Patients must remember to measure/monitor their blood pressure, glucose levels, weight, smoking behaviors, medication intake, and clinic visits, while managing the fears that are likely to occur when considering the risks of experiencing a future stroke. These somewhat daunting tasks turn routine prevention into a psychological burden rather than a simple medical prescription[1-4]. The concept of proactive health behavior has become a cornerstone of the current stroke prevention strategies. Proactive health behaviors include a range of deliberate, self-initiated actions aimed at maintaining health, preventing disease progression, and managing risk factors before clinical manifestations occur[5]. When this behavior goes beyond passive compliance with medical guidelines and includes active information retrieval, regular health monitoring, lifestyle changes, adherence to medication, and interaction with medical services[6], the same pattern is observed. Data from longitudinal studies have shown that individuals who are constantly engaged in proactive behavioral care are more likely to experience a negative impact on their health. However, a previous study showed that the incidence of stroke in those individuals who are proactive in their care is reduced by 25%-30% compared with that in those who adhere to a reactive or passive approach to health[7]. The American Heart Association’s “Life is 8 Essential Components” concept, which includes behavioral components such as a healthy diet, physical activity, and smoking cessation, has been validated as a powerful predictor of cardiovascular health outcomes, including stroke prevention[8].

Despite the generally recognized benefits of proactive health behaviors, individuals’ abilities to transform these behaviors into sustainable self-control practices differ significantly. Notably, the achievement of self-control, also known as self-control ability, is defined as the ability to monitor one’s health, make informed decisions, take appropriate action, and use available resources to effectively manage health. Self-control is not the ability to manage health. However, the achievement of self-control is a critical outcome of mental functioning in stroke prevention. Conceptually, self-control reflects the external outcome of basic psychological processes, including executive functions, attention control, and the ability to cope with stress[9]. Therefore, essentially the same patterns have been observed in impacted patients, as previous studies have shown that patients at a high risk of stroke who have higher self-control ability demonstrate better control of modifiable risk factors, higher rates of treatment adherence, and improved clinical outcomes[10]. The psychological pathways linking proactive health behaviors to higher self-control abilities remain unelucidated, which limits the development of targeted interventions to improve this pathway.

Psychological resilience, an indicator that is central to modern psychiatry and the science of mental health, offers a useful framework for understanding how individuals cope with chronic health problems[11]. Defined as the dynamic ability to adapt successfully to adversity, stress, or significant health threats, resilience is increasingly recognized as a modifiable psychological resource that can be developed through targeted intervention rather than being a fixed personality trait[12]. Psychological resilience differs from the stress buffering model, a fundamental theoretical framework in psychiatric research, which is commonly used as a tool for improving mental health. This suggests that resilience acts as a protective psychological pathway that mitigates the impact of chronic stressors on mental and physical health[13]. Within this model, resilience functions along several pathways, including higher emotional regulation, cognitive reassessment of health threats, sustained motivational engagement, and retention of meaning and purpose despite adversity. Therefore, determining and understanding the role of stress in an individual’s life is important. Patterns in these processes have been observed, as empirical evidence from populations with cardiovascular diseases shows that individuals with higher levels of resilience exhibit better adaptation to stress, more effective problem solving in managing their conditions, and greater self-efficacy in managing disease characteristics that are fundamentally psychological in nature and amenable to treatment. Psychiatric and psychotherapeutic intervention can also have roles in these processes[14].

Despite the theoretical foundations provided by current psychiatric and psychological science, significant gaps remain in understanding the mental health pathways underlying effective stroke risk management. Although resilience has been extensively studied in psychiatric rehabilitation after stroke, its role as a major mental health resource in the context of prevention remains poorly understood. Furthermore, empirically elucidating the specific psychological pathways by which resilience enhances the ability to transform healthy behaviors into sustained self-control may reveal similar patterns among high-risk individuals. Existing research has often considered resilience a secondary variable rather than a primary psychiatric indicator requiring targeted clinical attention. Quantifying the mediating contribution of resilience is essential to inform the integration of mental health services into cardiovascular disease prevention programs and guide the allocation of resources between behavioral and psychological interventions[15].

Therefore, to address these research gaps, in this study we aimed to investigate proactive health behaviors/self-control practices by examining psychological resilience, the relationship between active health behaviors and self-control, and the psychological pathway of prevention.

MATERIALS AND METHODS

Study design and participants

This cross-sectional observational study was conducted from January 2024 to December 2025 at the Faculty of Health and Well-being and the Faculty of Basic Medical Sciences of Panzhihua University and at Panzhihua Central Hospital (Panzhihua City, Sichuan Province, China) which serve urban and suburban areas. The study protocol was approved by the Research Ethics Committee of Panzhihua Central Hospital, and all participants provided written informed consent prior to participation in the study. The recruitment strategy included a targeted selection of individuals classified as high risk for stroke based on established criteria.

The inclusion criteria for participation were: (1) Age between 40 and 75 years; (2) Having at least two major stroke risk factors, including hypertension (systolic blood pressure ≥ 140 mmHg or diastolic blood pressure ≥ 90 mmHg, or current use of antihypertensive drugs), diabetes mellitus (fasting glucose ≥ 7.0 mmol/L or current use of hypoglycemic drugs), dyslipidemia (total cholesterol ≥ 6.2 mmol/L or low-density lipoproteins ≥ 4.1 mmol/L, or current use of lipid-lowering drugs), obesity [body mass index (BMI)] ≥ 28 kg/m²), smoking (defined as smoking at least one cigarette per day for the last year) or having a history of a transient ischemic attack; (3) The ability to read and understand survey materials in the local language; (4) No history of a diagnosed stroke or severe cognitive impairment (Mini-Mental State Examination score ≥ 24); (5) No serious mental health problems that could prevent participation; and (6) Stable health status without recent hospitalization (in the last 3 months). Exclusion criteria included ongoing participation in other behavioral therapy studies, the presence of an incurable medical condition with a life expectancy of less than one year, severe visual or hearing impairments that could prevent or impair completion of the assessment, and failure to provide informed consent.

Initially, 1247 individuals were selected to participate in the study. Among these, 168 did not meet the inclusion criteria, 52 declined to participate, and 27 dropped out during the data collection phase because of personal reasons or incomplete responses to the questionnaire. The final analytical sample included 1000 participants, the equivalent of 80.2% of the total number of participants who had met the inclusion criteria (the participant selection process is shown in Figure 1).

Figure 1 Participant flowchart of screening, exclusion, and final inclusion in the study.

A power analysis performed using G*Power 3.1 showed that this sample size would provide sufficient statistical power (> 0.95) to detect small- to medium-sized effects in the mediation analysis, with the alpha significance level set at 0.05.

Measures

Demographic and clinical characteristics: A structured demographic questionnaire was used to collect comprehensive information on the characteristics of the participants. Demographic variables included age (measured in full years), sex (male or female), ethnicity (Han Chinese, ethnic minorities, or others), marital status (married, divorced/separated, widowed, or never married), educational level (primary school or lower, secondary school, high school, and higher education), employment status (employed, retired, or unemployed), and monthly household income (grouped as < 3000 yuan, 3000-5000 yuan, 5001-8000 yuan, and > 8000 yuan). The clinical and behavioral variables were ascertained from the medical records or based on the participants’ self reports. These variables included height and weight, BMI, smoking status (current smoker, former smoker, or never smoked), alcohol consumption [classified as non-drinker, occasional drinker (less than once a week), or regular drinking (at least once a week)], current medication intake, and the presence of diagnosed comorbidities, including hypertension, diabetes mellitus, and dyslipidemia.

Proactive health behavior: Proactive health behavior was measured using the Chronic Disease Proactive Health Behavior Scale, a validated tool specifically designed to evaluate health-promoting behaviors in populations at risk of developing chronic diseases. The scale consists of 20 items distributed across five conceptually different dimensions: (1) Health information retrieval (four items), which evaluates the active pursuit of health-related knowledge from various sources; (2) Health monitoring (four items), which evaluates regular self-monitoring practices such as measuring and tracking blood pressure; (3) Lifestyle change (five items), which measures engagement in lifestyle changes that promote health, including diet adjustments and physical activity; (4) Medication adherence (three items), which measures consistency in taking medications as prescribed; and (5) Use of health services (four items), which evaluates appropriate and timely access to preventive medical care. Each item is rated on a five-point Likert scale ranging from 1 (never) to 5 (always), with higher total scores indicating greater involvement in proactive health behaviors. The higher total scores suggest persistent behavioral patterns. The total score can range from 20 to 100. The scale showed excellent internal consistency in the present population sample, with a Cronbach’s alpha coefficient of 0.91. The five-dimensional structure was confirmed by confirmatory factor analysis, which showed good indicators of model compliance (comparative compliance index, 0.94; root-mean-square approximation error, 0.06).

Psychological stability: Psychological resilience was measured using the Connor-Davidson 10-point Resilience Scale (CD-RISC-10), a widely used tool for assessing resilience. The CD-RISC-10 is a shortened version of the original 25-point scale which has demonstrated reliable psychometric properties in various population groups, including individuals with chronic diseases. The tool evaluates the main characteristics of resilience, such as the ability to adapt to changes, tendency to see the humorous side of things, ability to cope with stress, ability to recover from illness or difficulties, belief that everything happens for a reason, ability to cope with unpleasant feelings, and personal strength. Each item is rated on a five-point Likert scale from 0 (completely wrong) to 4 (almost always true), giving an overall score ranging from 0 to 40, with higher scores reflecting greater stability. The CD-RISC-10 scale was comprehensively validated in a Chinese population, with Cronbach’s alpha coefficients ranging from 0.85 to 0.91 and repeat-testing reliability coefficients ranging from 0.80 to 0.87 at intervals of 2 weeks to 4 weeks. In the present study, the scale showed high internal consistency with a Cronbach’s alpha of 0.89. The Chinese version of the CD-RISC-10 has been validated in previous studies with satisfactory psychometric properties[16]. Subclinical psychological distress and mild anxiety symptoms were measured using the 7-point Generalized Anxiety Disorder Scale (GAD-7), a well-validated screening tool for anxiety symptoms in the medical population. The GAD-7 questionnaire evaluates the frequency of anxiety symptoms over the past 2 weeks using seven items (feeling nervous, unable to stop worrying, excessive anxiety, difficulty relaxing, restlessness, irritability, and fear of something terrible). Each item is scored on a four-point Likert scale from 0 (not at all) to 3 (almost every day), which gives a total score ranging from 0 to 21. Using standard thresholds, the severity of symptoms is classified as minimal (0-4), mild (5-9), moderate (10-14), and severe (15-21). Mild anxiety, defined as a GAD-7 score of 5-9, is defined as clinically significant but subclinical anxiety that does not meet diagnostic thresholds but may affect functioning. The Chinese version of the GAD-7 has excellent psychometric properties, with Cronbach’s alpha coefficients of 0.89-0.92 and high reliability in repeated testing. In the present study, the internal consistency was excellent (Cronbach’s alpha = 0.91).

Ability to self-organize: Self-control ability, defined in this study as a measure of mental functioning that reflects the ability to perform executive functions and cope with stress, was measured using the Stroke Self-Management Scale, a disease-specific tool designed to assess self-control abilities in individuals at risk of stroke or in the recovery stage. Although originally developed for stroke survivors, this scale has been adapted and validated for use in high-risk populations to assess readiness to manage risk factors for cardiovascular disease. The scale includes 28 points in four areas that reflect key abilities for mental functioning: (1) Symptom management (eight items), which evaluates the ability to recognize and respond to warning signs through attention monitoring; (2) Risk factor control (eight items), which evaluates the ability to manage/regulate modifiable risk factors through executive planning and behavioral assessment; (3) Information and resource use (six items), which measures cognitive flexibility in finding and using health information and community resources; and (4) Emotional regulation (six items), which measures coping with stress and managing emotional responses to health problems. The points are evaluated on a five-point Likert scale from 1 (strongly disagree) up to 5 (strongly agree), with the overall score ranging from 28 to 140; higher scores indicate a higher self-control ability. The scale has demonstrated excellent psychometric properties, with Cronbach’s alpha values exceeding 0.90 in numerous validation studies. The reliability of the internal consistency in the present population was 0.93, and the four-factor structure was confirmed by confirmatory factor analysis with acceptable compliance indicators. The scale has been culturally adapted and validated for use in a Chinese population, demonstrating its reliability and validity[17].

Data collection procedures

Data were collected by trained researchers who had completed standardized training covering research protocols, ethical aspects, and procedures for conducting all assessments. Recruitment was conducted through several channels, including referrals from outpatient clinics, community health-screening programs, and information sessions at community centers. The participants were identified through systematic screening programs specifically for populations at a high risk of stroke conducted at these recruitment sites. After confirming eligibility for inclusion and informed consent, the researchers asked the participants to complete the questionnaires. Participants were provided with a quiet, private environment, either in a medical facility or community center. The assessment tool package was completed in a fixed order, starting with demographic and clinical information, followed by the scales of proactive health behavior, psychological stability, and self-control. During this process, researchers were available to answer questions and provide explanations as needed. The survey procedures were standardized, meaning that the questionnaire completion scheme remained unchanged. The completion time ranged from 25 minutes to 35 minutes. Quality control procedures included the immediate verification of completed questionnaires to rule out missing data. In such cases, the participants were asked to answer the missed questions, if possible. Data were entered into a secure electronic database by trained personnel using double checking to minimize rewriting errors.

Statistical analysis

All statistical analyses were performed using SPSS (version 27.0; IBM Corporation, Armonk, NY, United States) and PROCESS version 4.1 macro developed by Hayes. Prior to the main analyses, the data were checked for accuracy, missing values, and univariate outliers, which were defined as values greater than three SD from the mean. The amount of missing data was minimal (less than 2% for any single variable) and was determined to be completely random based on Little’s Missing Completely at Random test; therefore, data with missing values were excluded. Descriptive statistics, including frequencies, percentages, averages, and standard deviations were used to describe the results. The normality of the distribution of continuous variables was measured using the Kolmogorov-Smirnov test and visual analysis of histograms and QQ graphs, and all key variables showed acceptable normality.

Pearson’s bivariate correlation coefficients were calculated to examine the zero-order relationships between the main study variables, including proactive health behavior, psychological resilience, and self-control. Correlation coefficients were interpreted in accordance with generally accepted recommendations; absolute values from 0.10 to 0.29 were considered small, from 0.30 to 0.49: Moderate, and 0.50 or more: Large. Statistical significance was determined using two-way t-tests with an alpha significance level of 0.05.

The main hypothesis regarding mediation was tested using the PROCESS macro Model 4, which implements path analysis based on least squares regression to estimate the direct and indirect effects measured in the population data. This approach has several advantages over the traditional Baron and Kenny methods, including estimating indirect effects using bootstrap confidence intervals (CIs) that do not assume normality of the data distribution, simultaneously evaluating all paths in the mediation model, and calculating effect size indicators for indirect effects. In this model, proactive health behavior was the independent variable (x), psychological stability was the mediator (m), and self-control ability was the dependent variable (y). The demographic and clinical covariates, including age, sex, education level, income, BMI, and the number of comorbidities, were included as control variables to account for potential distorting effects.

The significance of the indirect effect (the mediated path through resilience) was measured using bias-adjusted bootstrap CIs based on 10000 bootstrap samples, with non-zero intervals indicating significant mediation. The proportion of the mediated total effect was calculated as the ratio of the indirect effect to the total effect measured in the sample, while the direct effect (pathway from proactive health behavior to ability to self-control, considering the mediator) was investigated to determine whether the mediation was partial (the direct effect remained significant) or complete (the direct effect became insignificant). However, all regression coefficients are presented as non-standardized values with SEs, standardized beta coefficients, t-values, and P-values. This process ensures that the same pattern persists because the effect sizes for the general model were measured using R²and adjusted R² values.

RESULTS

Demographic and clinical characteristics

The demographic and clinical characteristics of the 1000 included participants are shown in Table 1. The average age of the participants was 58.34 years (SD = 9.76 years, range 40-75 years), and most (38.4%) were 55-64 years of age. The population included 54.2% males and 45.8% females. Most participants identified as Han Chinese (93.6%), whereas 6.4% represented ethnic minorities or other ethnic groups. A similar pattern applied to marital status: 81.7% were married, 7.3% were divorced or separated, 6.8% were widowed, and 4.2% were never married. The level of education in the sample varied: 18.5% completed primary education or lower, 31.2% - secondary education, 28.7% - senior education, and 21.6% - higher education. Regarding employment, 42.3% of the participants were employed, 43.9% were retired, and 13.8% were unemployed. The monthly household income distribution showed that 23.4% of the participants earned less than 3000 yuan, 35.7% earned between 3000 and 5000 yuan, 26.8% earned between 5001 and 8000 yuan, and 14.1% earned more than 8000 yuan.

Table 1 Demographic and clinical characteristics of participants (n = 1000).

Characteristic	Category	n (%) or mean ± SD
Age (years)		58.34 ± 9.76
	40-54	312 (31.2)
	55-64	384 (38.4)
	65-75	304 (30.4)
Sex	Male	542 (54.2)
	Female	458 (45.8)
Ethnicity	Han Chinese	936 (93.6)
	Minority/other	64 (6.4)
Marital status	Married	817 (81.7)
	Divorced/separated	73 (7.3)
	Widowed	68 (6.8)
	Never married	42 (4.2)
Education	Primary or below	185 (18.5)
	Middle school	312 (31.2)
	High school	287 (28.7)
	College or above	216 (21.6)
Employment	Employed	423 (42.3)
	Retired	439 (43.9)
	Unemployed	138 (13.8)
Monthly income (yuan)	< 3000	234 (23.4)
	3000-5000	357 (35.7)
	5001-8000	268 (26.8)
	> 8000	141 (14.1)
BMI (kg/m2)		26.34 ± 3.45
Smoking status	Current	316 (31.6)
	Former	248 (24.8)
	Never	436 (43.6)
Alcohol use	Non-drinker	427 (42.7)
	Occasional	335 (33.5)
	Regular	238 (23.8)
Hypertension	Yes	743 (74.3)
Diabetes	Yes	389 (38.9)
Dyslipidemia	Yes	526 (52.6)
Number of risk factors	1	314 (31.4)
	2	427 (42.7)
	≥ 3	259 (25.9)

BMI: Body mass index.

Clinical and behavioral characteristic analyses revealed an average BMI of 26.34 kg/m² (SD = 3.45), indicating that the average study participant was overweight according to the Chinese BMI classification. Regarding smoking status, 31.6% were current smokers, 24.8% were former smokers, and 43.6% had never smoked. For alcohol consumption, 42.7% did not drink alcohol, 33.5% drank alcohol occasionally, and 23.8% drank alcohol regularly. A high prevalence of diagnosed comorbidities (74.3% had hypertension, 38.9% had diabetes, and 52.6% had dyslipidemia) was observed. Many participants had multiple comorbidities/risk factors; 31.4% of the participants had one comorbidity/risk factor, 42.7% had two, and 25.9% had three or more, highlighting the high-risk characteristics of this study population.

Descriptive statistics and correlations

Descriptive statistics for the main variables of the study and their mutual correlations are shown in Table 2. The average score for proactive health behavior was 67.23 (SD = 12.45, range 28-98), and the scores were distributed relatively normally within the possible range. The average psychological resilience score was 28.76 (SD = 6.34, range 8-40), indicating a moderate or high level of resilience. Self-control ability showed an average score of 98.45 (SD = 16.78, range 45-137), indicating a moderate level of self-control among the participants with high variability. For subclinical psychological distress, the average GAD-7 score was 5.82 (SD = 4.23, range 0-19). According to the standard selection criteria, 45.3% of the participants reported minimal anxiety (scores 0-4), 34.6% reported mild anxiety (scores 5-9), 15.2% reported moderate anxiety (scores 10-14), and 4.9% reported severe anxiety symptoms (scores 15-21). More than half of the participants (54.7%) reported experiencing some symptoms of anxiety, with minimal and mild anxiety being the most common categories among those with anxiety symptoms.

Table 2 Descriptive statistics and correlations among study variables (n = 1000).

Variable	Mean	SD	Range	1	2	3
Proactive health behaviors	67.23	12.45	28-98	-
Psychological resilience	28.76	6.34	8-40	0.562c	-
Self-management ability	98.45	16.78	45-137	0.647c	0.534c	-

^cP < 0.001.

Pearson’s correlation analysis revealed significant positive associations between all three of the major variables measured, with proactive health behaviors showing a strong positive correlation with self-control ability (r = 0.647, P < 0.001), indicating that the participants who were more likely to exhibit proactive health behaviors had higher self-control. In addition, proactive health behaviors were significantly and positively correlated with psychological resilience (r = 0.562, P < 0.001), suggesting that more active participation in health-promoting behaviors is associated with higher resilience. This indicates the persistence of similar patterns because psychological resilience showed a moderate to strong positive correlation with self-control ability (r = 0.534, P < 0.001), suggesting that individuals with higher levels of resilience have better self-control abilities. The GAD-7 scores indicating mainly minimal to mild anxiety levels showed significant inverse correlations with all main variables of the study: Proactive health behavior (r = -0.356, P < 0.001), psychological stability (r = -0.412, P < 0.001), and self-control ability (r = -0.389, P < 0.001), which indicates that higher levels of subclinical psychological distress were associated with less involvement in health-related behaviors, reduced psychological resilience, and poorer self-control abilities. All correlations were statistically significant at P < 0.001 and represented moderate or large effect sizes in accordance with the generally accepted interpretation guidelines. These bivariate associations provide preliminary support for the proposed mediation model and highlight the relevance of subclinical mental health vulnerabilities in this population.

Results of mediation analysis

The results of the mediation analysis examining psychological resilience as a mediating factor between proactive health behaviors and self-control are shown in Table 3. The analysis considered demographic covariates, including age, sex, education level, income, BMI, and number of comorbidities. The overall mediation model was statistically significant and explained a significant proportion of variance in the dependent variables.

Table 3 Mediation analysis results: Psychological resilience as mediator (n = 1000).

Path and effect	B	SE	β	t	95%CI	P value
Total effect (c)
PHB → SMA	0.702	0.032	0.521	21.938	0.639-0.765	< 0.001
Path a
PHB → PR	0.286	0.016	0.564	17.875	0.255-0.317	< 0.001
Model R2	0.342
Path b and c’
PR → SMA	0.692	0.067	0.261	10.328	0.561-0.823	< 0.001
PHB → SMA (direct, c’)	0.504	0.034	0.374	14.824	0.437-0.571	< 0.001
Model R2	0.528
Indirect effect (a × b)	0.198	0.018	-	-	0.165-0.2341	-
Proportion mediated	28.2%

¹Bootstrap confidence interval (10000 samples).

All models were controlled for age, sex, education, income, body mass index, and the number of comorbidities.CI: Confidence interval; PHB: Proactive health behavior; SMA: Self-management ability; PR: Psychological resilience.

In the first regression equation predicting psychological resilience based on proactive health behavior (pathway a), proactive health behavior showed a significant positive effect (B = 0.286, SE = 0.016, β = 0.564, t = 17.875, P < 0.001). This indicates that for each one-unit increase in proactive health behavior, psychological resilience increased by 0.286 units under covariate control. The model explained 34.2% of the variance in psychological resilience (R² = 0.342, F (7, 992) = 73.56, P < 0.001).

In the second regression equation predicting self-control ability based on both proactive health behavior (pathway c’) and psychological resilience (pathway b), both predictors showed significant independent effects, while psychological resilience significantly predicted self-control ability (B = 0.692, SE = 0.067, β = 0.261, t = 10.328, P < 0.001), indicating that for each one-unit increase in resilience, self-control capacity increased by 0.692 units when controlling for proactive health behaviors and covariates. The direct effect of proactive health behaviors on self-control behavior remained significant (B = 0.504, SE = 0.034, β = 0.374, t = 14.824, P < 0.001), indicating partial rather than complete mediation, that is, the same pattern was observed, as this model explained 52.8% of the variance in self-control ability (R² = 0.528, P < 0.001, F (8, 991) = 138.42, P < 0.001).

The overall effect of proactive health behavior on self-control ability (pathway c, excluding mediator) was significant (B = 0.702, SE = 0.032, β = 0.521, t = 21.938, P < 0.001). The indirect effect through psychological resilience (pathway a × pathway b) was calculated as 0.198, with an SE of 0.018 measured in the sample, whereas the bias-adjusted bootstrap CIs based on 10000 samples (95%CI: 0.165-0.234) did not contain zero, confirming the statistical significance of the indirect effect. The proportion of the overall effect mediated by psychological resilience was 28.2% (indirect effect divided by the overall effect: 0.198/0.702 = 0.282), indicating that approximately 28% of the effect of proactive health behaviors on self-control was realized through higher psychological resilience. This indicates the persistence of a similar pattern, as the remaining 71.8% represents a direct effect of proactive health behavior on self-control, independent of resilience. These results support the hypothesis that psychological resilience partially mediates the link between active health behaviors and self-control.

Additional tests

Additional analyses examined whether the mediating effect differed in key demographic subgroups, including age groups (40-54 years, 55-64 years, 65-75 years), sex (male vs female), and burden of comorbidities (1, 2, or 3+ risk factors). Indirect influence analysis using the PROCESS 59 model showed that the indirect effect of psychological resilience did not differ significantly in the different age groups (indirect influence index = 0.012, 95%CI: -0.023 to 0.047) or sexes (indirect influence index = -0.034, 95%CI: -0.089 to 0.021), indicating that there was no significant difference between the age groups (indirect influence index = 0.012, 95%CI: -0.023 to 0.047). This suggests that the mediating pathways functioned similarly in these demographic segments. A slight, albeit, statistically significant trend was noted in the burden of comorbidities (moderation index of mediation = 0.045, 95%CI: 0.003-0.091), and the indirect effect was slightly stronger among individuals with a large number of risk factors. Notably, the effect was small in magnitude.

DISCUSSION

The present study provides new evidence from a mental health perspective, demonstrating that psychological resilience serves as a significant mediating factor in the relationship between active health behaviors and self-control among individuals who have a high risk of stroke. The results also showed a high prevalence of subclinical symptoms in this population, with more than one-third of the participants showing mild anxiety. Resilience can serve as a psychological resource for psychosomatic problems in individuals concerned about stroke risk.

The strong positive association between proactive health behaviors and self-control abilities observed in this study is consistent with and extends previous research demonstrating the fundamental role of health-promoting behaviors in the prevention and treatment of chronic diseases[18]. The correlation coefficient of 0.647 represents a large effect size and shows that individuals who actively search for health information, monitor their physiological parameters, change lifestyle risk factors, follow a medication regimen, and make appropriate use of health services clearly demonstrate superior capabilities in managing their overall cardiovascular risk profile. This finding is consistent with the results of the Chinese Longitudinal Health and Retirement Study, which found that participation in multiple preventive health interventions was associated with a 35% reduction in the likelihood of developing cardiovascular events over a five-year period[19]. The pathways underlying this link are likely to include several branches, including better health literacy, which helps in making informed decisions, the formation of healthy habits that become habitual over time, improved physiological parameters as a result of lifestyle changes, and stronger therapeutic alliances with health professionals. These behaviors support ongoing management efforts[20].

The average score for proactive health behaviors in our high-risk population (67.23/100) indicates moderate rather than optimal engagement, suggesting greater opportunities for improvement through targeted interventions measured in the participants. However, a moderate level of engagement is of particular concern given that these individuals face an increased risk of stroke and will benefit the most from intensive preventive measures. In addition, qualitative studies in similar populations have identified many barriers to proactive health behaviors, including insufficient knowledge of stroke risk factors and warning signs, competing life circumstances that reduce the priority of maintaining health, fatalistic beliefs about the inevitability of illness, financial constraints that prevent access to healthy food and sports facilities, and a lack of availability of healthy food and sports facilities. In addition, social support for beneficial behavioral changes may be insufficient[21]. These similar patterns between previous studies and the results of the present study indicate that addressing these multilayered barriers through comprehensive interventions that combine education, skills training, environmental modification, and social support mobilization is a critical priority in stroke prevention programs.

The mediating role of psychological resilience identified in this study provides important insights into the psychological pathways through which proactive health behaviors lead to an increased capacity for self-control. A significant indirect effect through resilience (accounting for 28% of the overall effect) suggests that engaging in proactive health behaviors not only directly improves self-control through the acquisition of knowledge and skills but also increases psychological resources that further contribute to effective management. This finding is consistent with theoretical perspectives that postulate resilience as a dynamic process rather than a static trait. Thus, engaging in mastery experiences, such as successful health behavioral changes, builds confidence and adaptive abilities that extend to different areas of management[22]. The resource caravan theory provides an appropriate framework for understanding this process, suggesting that an initial increase in resources (e.g., improved health as a result of behavioral changes) creates an impulse to acquire additional resources, including psychological resources such as resilience, which in turn allows for more effective management of current health problems[23].

Several pathways may explain how proactive health behaviors, as measured in our study population, increase resilience in high-risk stroke populations. Successful participation in health-related behaviors provides mastery experiences that increase self-efficacy, a key component of resilience. For example, self-efficacy is increased when an individual successfully controls their blood pressure. Adhering to exercise regimens or communicating effectively with health professionals can enable individuals at risk of a stroke to accumulate evidence of their ability to influence health outcomes, which reinforces beliefs in controllability and subjectivity[24]. This pattern is observed when proactive health behaviors involve information-seeking or proactive interaction, such as visiting a doctor. Group sports, participation in health education sessions, and collaboration with medical teams can strengthen social connections and perceived support; these benefits represent important protective factors for sustainability[25]. Physiological improvements resulting from health-related behaviors, including reduced inflammation, improved cardiovascular endurance, and better metabolic control, may have direct neurobiological effects on the stress response systems and emotional regulation abilities that underlie resilience[26]. In addition, the cognitive interaction required to plan health-related behaviors is important because implementation can strengthen executive functions, including cognitive flexibility and problem-solving skills, which are crucial for sustainable adaptation to health problems[27]. These executive functions represent the neurocognitive substrate underlying the ability to self-control, explaining why we view self-control as a result of mental functioning, rather than just a behavioral model.

The fact that psychological resilience independently predicts self-control, which is understood as a result of mental functioning that reflects executive functions and the ability to cope with stress, even when controlling for active health behaviors, demonstrates the unique contribution of resilience as a psychological resource distinct from behavioral patterns. In the present study, the participants with higher levels of resilience demonstrated superior self-discipline in several areas, including symptom recognition and response, risk factor control, information use, and emotional regulation. This is consistent with evidence from chronic disease research showing that resilience improves adaptive coping strategies, enables healthy behavior maintenance during stressful periods, promotes flexible problem solving when management difficulties occur, and supports an emotional balance that prevents catastrophic responses to health problems[28]. The link between resilience and self-discipline likely works through a variety of psychological pathways, including better stress management that prevents behavioral disorders during challenging periods, more optimistic assessments of health situations that motivate further effort, greater cognitive flexibility in identifying alternative management strategies when initial approaches fail, and higher self-efficacy in disease management tasks[29]. The same pattern was observed in our sample (34.6% of the participants showed mild anxiety), indicating that these individuals require special clinical attention. However, the participants with mild anxiety did not meet the diagnostic criteria for generalized anxiety disorders. Subclinical psychological distress, including mild anxiety and emotional dysregulation, can impair executive function, concentration, and the ability to cope with stress. These are the fundamental psychological processes underlying effective self-control. Previous research has shown that even mild symptoms of anxiety can impair working memory, attention control, and decision-making, all of which are critical for comprehensive health behavior management. Our results expand this body of research by demonstrating significant inverse relationships between mild anxiety and psychological resilience as well as self-control ability in high-risk stroke populations, suggesting that subclinical mental health vulnerabilities may represent an underestimated barrier to effective cardiovascular risk management.

The partial rather than complete mediated model in this study shows that while resilience is a critical pathway for mental health, the direct pathways from proactive health behavior to self-control remain high. From the perspective of psychiatry and psychosomatic medicine, this conclusion has significant implications for clinical practice. Psychological resilience in this analysis is considered not just a behavioral correlate but also a functional psychological dimension that operates through emotion regulation, stress adaptation, and cognitive flexibility, which are generally amenable to psychiatric and psychological interventions. Thus, clinical stroke prevention protocols should go beyond traditional health education and include the same systematic psychological assessment and evidence-based resilience-enhancing interventions that we measured in this study. Furthermore, cognitive behavioral therapy methods that target maladaptive cognitions in relation to health management and stress reduction programs should be based on the following criteria: (1) Mindfulness approaches that address chronic health-related anxiety and acceptance; and (2) Commitment therapy approaches that promote psychological flexibility. Both are empirically proven methods that can be integrated into the prevention of cardiovascular diseases. The integration of such psychological interventions with medical treatment reflects the principles of psychosomatic medicine and collaborative care, in which mental health professionals work together with primary care teams and cardiologists to address the psychological factors that determine the outcomes of chronic diseases[30]. The patterns observed in our results suggest the potential benefits of psychosomatic medicine and collaborative care. From a practical perspective, this finding suggests that optimal stroke prevention programs should integrate components of behavioral modification, such as goal setting, action planning, and environmental restructuring, with strategies for increasing resilience, including cognitive reassessment training, mindfulness-based stress reduction, problem-solving skills training, and meaning-making interventions[31].

Several limitations should be considered when interpreting the results of this study and planning future studies. The cross-sectional design does not allow unambiguous causal conclusions to be drawn regarding the direction of the relationship between proactive health behavior, resilience, and self-control. Although the proposed model assumed that behavior increases resilience, which in turn improves self-control, alternative directional models are also possible. For example, individuals with initially higher resilience may be more likely to initiate proactive health behaviors, and a higher level of success in self-control may enhance engagement in behavioral activities and resilience. Longitudinal studies with several stages of evaluation are necessary to identify time sequences and establish causal relationships more accurately. All measurements were based on self-reports, which creates the potential for systematic error associated with the method of data collection, desire for social desirability, and inaccuracies in the participants’ recall. Future studies will benefit from the inclusion of objective behavioral indicators, such as electronic monitoring of medication intake, physical activity assessed using an accelerometer, documentation of the use of medical services, and assessments of self-control skills conducted by observers or based on task results. Furthermore, the present study population was drawn from a specific geographic region and may not fully reflect the diversity of high-risk stroke populations worldwide, particularly given the cultural differences in health perceptions, resilience, and self-monitoring approaches. Although similar patterns were observed among our participants, relevant studies in different cultural contexts and healthcare systems are needed to establish generalizability. Finally, although we controlled for many demographic and clinical covariates, confounding variables, such as personality traits, cognitive function, health literacy, and social support, may partially explain the observed relationships. Therefore, more comprehensive models that include these additional factors are needed in future studies. Such additional work will promote a more complete understanding of the nomological networks that surround sustainability and self-control.

CONCLUSION

The results of this study suggests that psychological resilience functions as a pathway to mental health, promoting proactive health management and reducing the risk of developing mental disorders. The high prevalence of subclinical psychological problems shows the importance of psychological assessments and resilience-building interventions.