Family Harmony Mediates the Association Between Subjective Well-Being and Health-Promoting Behaviors Among Middle-Aged Women in Northern China

Abstract

Purpose

Health-promoting behaviors (HPB) are essential for disease prevention and healthy aging but remain suboptimal among middle-aged women (aged 45–59 years), who may face perimenopausal changes and work–family demands. Subjective well-being (SWB) is associated with HPB, but the role of family relational factors in this association remains unclear. Family harmony may represent a key relational pathway linking SWB to HPB; however, evidence from Northern China remains limited, and this pathway may differ by ethnicity because women’s family roles and responsibilities vary across sociocultural contexts. This study examined whether family harmony mediates the SWB–HPB association among middle-aged women in Northern China and explored variation across HPB domains and by ethnicity.

Patients and Methods

A cross-sectional survey was conducted from June to October 2022 in eight communities in Yantai and Yanji, Northern China. A total of 529 middle-aged women, comprising 415 Han majority women (78.45%) and 114 Chinese Korean minority women (21.55%), completed validated measures of SWB, family harmony, and HPB. Descriptive, correlation, and mediation analyses using the Hayes PROCESS macro (Model 4) were performed, together with domain-specific and ethnicity-stratified analyses.

Results

Family harmony partially mediated the SWB–HPB association (indirect effect = 0.047, 95% confidence interval [0.014, 0.081]). The indirect effect was significant for health responsibility, physical activity, nutrition, stress management, and spiritual growth, but not interpersonal relationships. Among Han participants, the association was primarily direct; among Chinese Korean participants, an indirect-only mediation pattern through family harmony was observed.

Conclusion

These findings highlight the importance of considering family relational contexts, alongside individual psychological resources, in efforts to support healthy lifestyles among middle-aged women. Community health programs may benefit from combining individual-level support with family-engaged components and tailoring the emphasis of these strategies to the sociocultural contexts of different population groups.

Introduction

Health-promoting behaviors (HPB) are a cornerstone of disease prevention and healthy aging,¹ yet they remain suboptimal among middle-aged women, a group experiencing concurrent physiological and role-related transitions. HPB refers to a multidimensional set of modifiable lifestyle practices, including health responsibility, physical activity, nutrition, interpersonal relationships, stress management, and spiritual growth.^2,3 Globally, the number of women who are menopausal or postmenopausal is projected to exceed one billion by 2030,⁴ highlighting the growing public health relevance of women’s health in midlife. During this life stage, women often undergo the perimenopausal transition while simultaneously balancing work and family caregiving responsibilities,^5,6 which may constrain their ability to initiate and sustain healthy routines.⁷ Moreover, women frequently assume a disproportionate share of household health-related decision-making and caregiving tasks,^8,9 suggesting that suboptimal HPB in this life stage may have broader implications for household health environments.¹⁰ Despite well-established benefits, a systematic review and meta-analysis indicated that health-promoting behaviors among menopausal women are, on average, at a moderate level, with physical activity consistently scoring the lowest, suggesting substantial room for improvement in midlife women’s engagement in HPB.¹¹ Accordingly, identifying actionable correlates and potential pathways of HPB among middle-aged women to inform culturally appropriate interventions has become an important public health priority. Among these potential determinants, positive psychological resources, particularly subjective well-being, have received increasing attention in recent research.

Subjective well-being (SWB) is increasingly recognized as a positive psychological resource associated with adaptive functioning and healthier lifestyle patterns.¹² Beyond reflecting psychological health, SWB may facilitate sustained engagement in HPB by supporting positive appraisal, self-regulation, and resilience when individuals face chronic stressors.¹³ Empirical studies have reported positive associations between SWB and multiple health-related behaviors, including physical activity and other lifestyle practices.¹⁴ Longitudinal evidence further suggests that higher SWB can prospectively predict more favorable health behavior profiles over time.¹⁵ However, much of the existing literature remains descriptive or focuses primarily on direct associations, leaving the potential pathways linking SWB with health behaviors insufficiently specified.¹³ This limitation may be particularly salient for midlife women, whose daily health practices are often embedded in family roles and relational contexts.¹⁶ Accordingly, it remains unclear whether family-level relational processes, particularly family harmony, represent a potential pathway linking SWB with sustained HPB in this population.

Given documented regional heterogeneity in family organization across China,^17,18 family harmony may represent a particularly important relational resource for middle-aged women in Northern China. Family harmony broadly reflects cohesive, supportive, and low-conflict family functioning, and has been operationalized in Chinese populations using a validated multidimensional instrument, the 24-item Family Harmony Scale (FHS) and its 5-item short form.¹⁹ In principle, a harmonious family climate can facilitate both practical and emotional support, reduce interpersonal friction, and enable more stable routines for maintaining health-promoting behaviors. More broadly, family-level processes—such as shared norms, support, supervision, modelling, and the household environment—are increasingly recognized as potential pathways through which families shape members’ health behaviors,¹⁶ suggesting that family harmony may be relevant to HPB beyond individual-level correlates. Evidence from a large Chinese sample further indicates that higher family harmony weakens the association between life stress and depressive symptoms,¹⁹ indicating that family harmony may function as a protective relational context for midlife women. However, family harmony has rarely been examined as a potential mediator between positive psychological resources such as SWB and health-promoting behaviors. In addition, adult women remain underrepresented in the broader family health-promotion literature,¹⁶ warranting focused examination in community-based samples of middle-aged women.

Building on these perspectives, theoretical accounts and emerging evidence suggest that the association between SWB and healthier lifestyles may be partly accounted for by family harmony. The broaden-and-build theory posits that positive affect broadens individuals’ thought–action repertoires and helps build enduring resources, including social and relational resources, which can facilitate constructive interactions and more supportive relational processes relevant to family harmony.²⁰ Consistent with this theoretical account, higher SWB may be associated with more adaptive coping and communication patterns within the household, which may be linked to greater family harmony.²¹ In turn, family harmony may support the initiation and maintenance of health routines through multiple family-level processes, such as shared norms, encouragement, modelling, coordinated daily schedules, and the provision of instrumental and emotional support.^16,22 Moreover, consistent with the stress-buffering hypothesis,²³ a harmonious family climate may mitigate the adverse impact of daily stressors, thereby making sustained engagement in health-promoting routines more feasible. Importantly, HPB is multidimensional,^2,3 and the relevance of family processes may differ across behavioral domains. Therefore, it is necessary to examine both overall HPB and domain-specific pathways within the proposed mediation model linking SWB, family harmony, and HPB.

Women’s family roles and health-related practices are embedded in sociocultural contexts, including household structure, gender-role expectations, intergenerational relationships, and the division of domestic and caregiving work.^17,18,24 Accordingly, the role of family harmony in women’s health-related behaviors may not be uniform across ethnic groups. Research among Chinese Korean women has highlighted women’s substantial household responsibilities and the importance of spousal support.²⁵ Comparative research involving Han and Chinese Korean adults aged 40–59 has further suggested that the factors associated with marital adjustment may differ between the two groups, including gender-role attitudes, intergenerational relationships, elder-care burden, and whether both spouses are employed.²⁶ These differences in family roles and support structures may shape both how SWB is reflected in family harmony and how family harmony supports HPB, thereby contributing to different mediation patterns across ethnic groups. However, it remains unclear whether this mediating pattern differs between Han majority and Chinese Korean minority women. Therefore, this study examined whether family harmony mediates the association between SWB and HPB among middle-aged women in Northern China and explored whether this mediation pattern varies across HPB domains and between Han majority and Chinese Korean minority women.

This study proposes the following hypotheses, as illustrated in Figure 1:

Hypothesis 1: Subjective well-being, family harmony, and overall health-promoting behaviors are positively correlated among middle-aged women in Northern China.

Hypothesis 2: Subjective well-being shows a significant positive association with overall health-promoting behaviors among middle-aged women in Northern China.

Hypothesis 3: Family harmony mediates the association between subjective well-being and overall health-promoting behaviors among middle-aged women in Northern China.

Figure 1.

Hypothesized mediation model.

Exploratory aims: To explore whether (i) the mediation pattern differs by ethnicity, comparing the Han majority and the Chinese Korean minority, and (ii) the mediation effects vary across HPB domains.

Materials and Methods

Study Design and Setting

This study employed a community-based cross-sectional survey conducted between June and October 2022 in two purposively selected sites in Northern China: Yantai, Shandong Province, and Yanji, Jilin Province (Yanbian Korean Autonomous Prefecture). The two sites were selected to capture contextual heterogeneity relevant to the study aims. Yantai was chosen as a predominantly Han urban setting with relatively stronger regional economic development,²⁷ whereas Yanji was chosen as a setting with a high concentration of Chinese Korean minority residents within an autonomous prefecture and comparatively different socioeconomic and sociocultural conditions.²⁸ Together, these two sites provided meaningful variation in ethnic composition and broader family and community context, allowing exploratory examination of whether the associations among subjective well-being, family harmony, and health-promoting behaviors differed across ethnic and regional settings.

Ethical Consideration

The study was approved by the Institutional Review Board of Yanbian University School of Medicine (Approval No. 20220516-1088) and conducted in accordance with the Declaration of Helsinki. Written informed consent was obtained from all participants prior to data collection. Participation was voluntary, and anonymity and confidentiality were assured. This study is reported in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines.²⁹

Participants and Sampling

This study used purposive sampling with a typical-case approach to capture variation in socioeconomic and sociocultural contexts within Northern China. The two study sites, Yantai and Yanji, were selected to reflect variation in ethnic composition and regional context. Within each site, communities were stratified by type (urban, suburban, and rural) to ensure coverage across different residential settings. Eligible communities were then identified through a multistage, feasibility-based site selection process at the district/county, township/subdistrict, and community/village levels, taking into account accessibility and willingness to participate.

Eligible participants met the following criteria: (i) women aged 45–59 years who had resided in the community for at least three months; and (ii) willingness and ability to provide informed consent. The exclusion criteria were as follows: (i) cognitive impairment, significant hearing or visual impairment, or communication difficulties that precluded valid participation; and (ii) severe physical illness, including major somatic diseases, that prevented completion of the survey. Eligibility was determined primarily based on participants’ self-reports, supplemented by on-site screening conducted by trained investigators.

The sample size was estimated using the standard formula for estimating a population mean in a cross-sectional survey: ,³⁰ where σ represents the estimated standard deviation of the health-promoting lifestyle score among middle-aged women, reported as 18.20 in a previous study.³¹ The level of significance was set at α = 0.05, corresponding to , and the allowable absolute error (d) was set at 10% of the expected standard deviation (d = 1.82). Based on these parameters, the minimum required sample size was calculated to be N = 385. Considering an anticipated non-response or incomplete-questionnaire rate of 10%, the final required sample size was adjusted to 428 participants.

Within the selected communities, eligible women were recruited through community-based on-site convenience sampling. A total of 177 participants were recruited from urban communities, 172 from suburban communities, and 180 from rural communities. The final analytic sample comprised 529 participants.

Measurements

Sociodemographic Variables

A self-administered questionnaire was developed a priori to capture sociodemographic and health-related variables based on the study objectives and prior research. The instrument comprised 14 items assessing the following domains: age, ethnic group, marital status, residence, education level, employment status, living arrangement, children, medical insurance, average monthly household income, diagnosed chronic disease, current medication use, physical examination in the past three years, and use of community health services.

Health-Promoting Behaviors

Health-promoting behaviors were assessed using the revised Chinese version of the Health-Promoting Lifestyle Profile II (HPLP-II-R), originally developed by Walker et al and subsequently adapted and revised for the Chinese population by Cao et al,^2,3 The scale consists of 40 items across six dimensions: health responsibility (11 items), physical activity (8 items), nutrition (6 items), interpersonal relationships (5 items), stress management (5 items), and spiritual growth (5 items). Each item is rated on a 4-point Likert scale (1 = “never,” 2 = “occasionally,” 3 = “often,” 4 = “always”). Total scores range from 40 to 160, with higher scores indicating a healthier lifestyle. According to established classification criteria, scores of 40–69, 70–99, 100–129, and 130–160 indicate poor, fair, good, and excellent levels of HPB, respectively. In the Chinese validation study, Cronbach’s α coefficients for the six subscales ranged from 0.63 to 0.81; in the present study, Cronbach’s α for the total scale was 0.71, indicating acceptable internal consistency.

Family Harmony

Family harmony was measured using the Family Harmony Scale–Short Form (FHS-5), developed by Kavikondala et al within the context of Chinese culture.¹⁹ The scale contains five items that assess emotional connectedness and interpersonal harmony within the family. Each item is rated on a 5-point Likert scale (1 = strongly agree, 5 = strongly disagree). Item scores were subsequently reverse-coded so that higher total scores indicate greater levels of family harmony (range: 5–25). The Cronbach’s α was 0.79 in the original study and 0.76 in the present study, indicating acceptable internal consistency.

Subjective Well-Being

Subjective well-being was evaluated using the Happiness Scale for Middle-Aged Women (HAS-MW), developed by Shin et al,³² which is grounded in existence, relation, and growth theory. The Chinese version of the HAS-MW was culturally adjusted by Liu et al.³³ The scale includes 24 items across four dimensions: self-worth (8 items), positive outlook (8 items), self-management (5 items), and family relationship (3 items). Each item is rated on a 5-point Likert scale (1 = “strongly disagree” to 5 = “strongly agree”), yielding a total score ranging from 24 to 120. Higher scores indicate higher levels of subjective well-being. The scale has demonstrated good psychometric properties, with Cronbach’s α coefficients of 0.92 for the original scale, and the same value (0.92) was obtained in the present study. The survey questionnaire is provided in Supplementary Appendix 1.

Data Collection

Prior to the formal survey, a pilot study was conducted with a small sample to ensure the clarity and feasibility of the procedures. During scheduled survey days in each selected community, trained researchers approached potentially eligible women in public areas of the community. Women who expressed interest were screened according to the eligibility criteria. All participants provided written informed consent before enrollment. Trained researchers explained the study purpose, procedures, and precautions using a standardized script. To ensure accurate communication, especially in the Chinese Korean minority communities, the research team included investigators proficient in Korean who provided language support when necessary. Participants completed the questionnaires independently and anonymously in a relatively quiet area at the survey site. Upon completion, the questionnaires were checked on-site for completeness and accuracy.

A total of 550 questionnaires were distributed to eligible women who agreed to participate. Of these, 529 valid questionnaires were returned and included in the analysis, yielding a valid response rate of 96.18%. The average time required to complete the questionnaire was approximately 15–20 minutes.

Data Analysis

All statistical analyses were performed using IBM SPSS Statistics version 26.0 (IBM Corp., Armonk, NY, USA) and the PROCESS Macro version 4.2. The normality of the data distribution was assessed using both Probability-Probability (P-P) plots and Quantile-Quantile (Q-Q) plots. Data were considered to approximate a normal distribution when the points lay close to the diagonal line. Due to the low proportion of missing data (< 5%), all questionnaires were retained. Missing values were handled as follows: for continuous variables, they were imputed using the mean or median, while the mode was used for categorical variables. For continuous variables that were approximately normally distributed, data were summarized using the mean and standard deviation (SD). For continuous variables that deviated substantially from normality or for ordinal variables, the median and interquartile range (IQR) were reported. Categorical variables were expressed as frequencies and percentages (n, %).

Data analysis proceeded in several steps to examine the relationships among the key variables and test the hypothesized mediation model. First, group differences in health-promoting behaviors across various sociodemographic characteristics were examined using independent samples t-tests (for two-group comparisons) and one-way analysis of variance (ANOVA) (for comparisons involving three or more groups). Second, bivariate correlations among the key continuous variables (subjective well-being, family harmony, and health-promoting behaviors) were assessed using Pearson’s correlation analysis to provide a preliminary understanding of their interrelationships. Subsequently, multiple linear regression was performed to identify significant predictors of health-promoting behaviors. Prior to interpretation of the regression models, multicollinearity among predictors was assessed using tolerance and variance inflation factor (VIF). A tolerance value > 0.20 and a VIF value < 5.0 were considered to indicate no problematic multicollinearity. Sociodemographic variables that showed significant group differences in the initial analyses were considered potential confounders and were included as covariates in the regression and mediation models. This approach enabled examination of the associations among subjective well-being, family harmony, and health-promoting behaviors after accounting for the influence of covariates. Finally, the mediating role of family harmony in the relationship between subjective well-being and health-promoting behaviors was tested using Model 4 of the PROCESS Macro for SPSS. The analysis utilized 5000 bootstrap samples to estimate the indirect effect with a bias-corrected 95% confidence interval (CI). The mediation effect was considered statistically significant if this confidence interval did not include zero. All statistical tests were two-tailed, and a P-value < 0.05 was considered statistically significant unless otherwise specified.

Results

Participant Characteristics

The general characteristics of the study participants are summarized in Table 1. The final sample consisted of 529 middle-aged women, of whom 78.45% were from the Han majority and 21.55% were from the Chinese Korean minority. Most participants (74.10%) were married, and residence was evenly distributed across urban (33.46%), suburban (32.51%), and rural (34.03%) areas. Regarding educational attainment, nearly three-quarters of the sample (74.67%) had attained a high school education or above.

Table 1.

Sociodemographic Characteristics of Middle-Aged Women (N=529)

Variables (%)	Family Harmony		Subjective Well-Being		Health-Promoting Behaviors
	t/F	P	t/F	P	t/F	P
Ethnic Group
Han majority (78.45)	−0.515	0.607	−0.956	0.339	−0.122	0.903
Chinese Korean minority (21.55)
Marital Status
Married (74.10)	2.118	0.035	3.120	0.002	4.236	<0.001
Not married (never married/divorced/widowed) (25.90)
Residence
Urban (33.46)	1.226	0.221	0.644	0.520	2.870	0.005
Suburban (32.51)
Rural (34.03)
Education Level
Elementary school (5.29)	5.126	0.002	1.268	0.285	16.179	<0.001
Middle school (20.04)
High school (37.24)
College or above (37.43)
Employment Status
Employed (92.82)	0.485	0.630	−0.163	0.871	1.464	0.151
Unemployed (7.18)
Living Arrangement
Living alone (15.50)	−2.654	0.008	−3.760	<0.001	−3.762	<0.001
Living with others (84.50)
Children
No children (14.18)	0.843	0.400	0.662	0.508	1.752	0.080
Has children (85.82)
Medical Insurance
Self-paid (10.96)	0.379	0.684	2.274	0.104	1.096	0.335
Urban employee insurance (27.98)
Urban and rural resident insurance (61.06)
Average Monthly Household Income (RMB)
≤5000 (14.18)	2.420	0.065	4.577	0.004	3.561	0.014
5001–10,000 (41.59)
10,001–15,000 (34.40)
>15,000 (9.83)
Diagnosed Chronic Disease
No (49.53)	3.103	0.002	3.833	<0.001	6.852	<0.001
Yes (50.47)
Current Medication Use
Not currently taking medication (61.06)	3.009	0.003	3.951	<0.001	5.124	<0.001
Currently taking medication (38.94)
Physical Examination in the Past Three Years
Never (21.93)	5.090	0.006	6.748	0.001	19.178	<0.001
Only when feeling unwell (52.74)
Once a year (25.33)
Use of Community Health Services
Yes (93.95)	4.885	<0.001	4.898	<0.001	7.978	<0.001
No (6.05)

Notes: Independent-samples t tests were used for two-group comparisons, and one-way analysis of variance was used for comparisons involving three or more groups.

Abbreviations: HPB, Health-promoting Behaviors; RMB, renminbi.

Correlation Analyses

Descriptive statistics and bivariate correlations for family harmony, SWB, and HPB are presented in Table 2. Mean scores were 14.91 ± 1.87 for family harmony, 74.43 ± 10.56 for SWB, and 118.99 ± 6.81 for HPB. Pearson correlation analyses revealed that all three variables were significantly and positively intercorrelated. Specifically, family harmony was strongly positively correlated with HPB (r = 0.540, P < 0.01). SWB was also positively associated with HPB (r = 0.372, P < 0.01) and with family harmony (r = 0.238, P < 0.01).

Table 2.

Correlation Among Family Harmony, SWB, and HPB

Variables	Mean ± SD	Family Harmony	SWB	HPB
Family Harmony	14.91 ± 1.87	1
SWB	74.43 ± 10.56	0.238**	1
HPB	118.99 ± 6.81	0.540**	0.372**	1

Note: **P < 0.01.

Abbreviations: SD, standard deviation; SWB, Subjective Well-being; HPB, Health-promoting Behaviors.

Mediation Analysis Results

Prior to mediation analysis, collinearity diagnostics indicated no evidence of problematic multicollinearity among the predictors, with tolerance values ranging from 0.537 to 0.927 and VIF values ranging from 1.079 to 1.861 (Supplementary Appendix 2, Table S1). Based on the results of the univariate analyses, nine sociodemographic and health-related variables were significantly associated with HPB: marital status, residence, education level, living arrangement, average monthly household income, diagnosed chronic disease, current medication use, physical examination in the past three years, and use of community health services. These variables were therefore treated as potential confounders and included as covariates in subsequent mediation analyses to control for their potential effects.

The total-sample mediation model was then estimated after adjusting for these covariates. The final regression model for overall HPB explained 53.2% of the variance in HPB (R² = 0.532, adjusted R² = 0.522, F = 53.455, P < 0.001; Supplementary Appendix 2 and Table S2). The results revealed a significant total effect of SWB on HPB (Effect = 0.157, 95% CI [0.085, 0.228]), indicating that higher levels of SWB were associated with greater engagement in HPB. After accounting for family harmony, the direct effect of SWB on HPB remained statistically significant (Effect = 0.109, 95% CI [0.044, 0.174]). The indirect effect of SWB on HPB through family harmony was also significant (Effect = 0.047, 95% CI [0.014, 0.081]), indicating a partial mediation effect. These findings indicate a significant indirect association between SWB and HPB via family harmony, as illustrated in Figure 2 and detailed in Supplementary Appendix 2, Tables S2 and S3.

Figure 2.

Mediation model for the total sample.

Notes: **P < 0.01; ***P < 0.001. a, the effect of SWB on family harmony; b, the effect of family harmony on HPB; c, the total effect of SWB on HPB; c′, the direct effect of SWB on HPB.

Abbreviations: SWB, Subjective Well-being; HPB, Health-Promoting Behaviors.

To further explore the mediating role of family harmony in the relationship between SWB and specific dimensions of HPB, separate path analyses were conducted for each of the six subdimensions of the HPLP-II-R. The results indicated that family harmony significantly mediated the association between SWB and HPB across most domains. As shown in Table 3 and Figure 3, the indirect effects of SWB on HPB through family harmony were statistically significant for five subdimensions: health responsibility (Effect = 0.030, 95% CI [0.007, 0.043]), physical activity (Effect = 0.028, 95% CI [0.006, 0.039]), nutrition (Effect = 0.017, 95% CI [0.004, 0.033]), stress management (Effect = 0.017, 95% CI [0.003, 0.030]), and spiritual growth (Effect = 0.020, 95% CI [0.004, 0.034]). These findings indicate significant indirect associations of SWB with these subdimensions via family harmony. However, the indirect effect for the interpersonal relationships dimension did not reach statistical significance (Effect = 0.012, 95% CI [−0.0002, 0.030]). Detailed regression and mediation results for the six HPB domains are provided in Supplementary Appendix 2, Tables S4–S15.

Table 3.

Path Analysis Results of HPB Among Middle-Aged Women

Path	Effect	BootSE	Boot 95% CI
HPB
Total effect	0.157	0.037	0.085, 0.228
SWB → HPB	0.109	0.033	0.044, 0.174
SWB → Family Harmony → HPB	0.047	0.017	0.014, 0.081
Health Responsibility
Total effect	0.162	0.039	0.067, 0.187
SWB → Health Responsibility	0.132	0.038	0.045, 0.162
SWB → Family Harmony → Health Responsibility	0.030	0.009	0.007, 0.043
Physical Activity
Total effect	0.157	0.046	0.050, 0.182
SWB → Physical Activity	0.129	0.045	0.030, 0.161
SWB → Family Harmony → Physical Activity	0.028	0.008	0.006, 0.039
Nutrition
Total effect	−0.010	0.041	−0.097, 0.066
SWB → Nutrition	−0.027	0.041	−0.113, 0.050
SWB → Family Harmony → Nutrition	0.017	0.008	0.004, 0.033
Interpersonal Relationships
Total effect	0.005	0.046	−0.086, 0.096
SWB → Interpersonal Relationships	−0.007	0.047	−0.098, 0.085
SWB → Family Harmony → Interpersonal Relationships	0.012	0.008	−0.0002, 0.030
Stress Management
Total effect	−0.017	0.038	−0.083, 0.067
SWB → Stress Management	−0.034	0.038	−0.097, 0.053
SWB → Family Harmony → Stress Management	0.017	0.007	0.003, 0.030
Spiritual Growth
Total effect	0.044	0.040	−0.040, 0.116
SWB → Spiritual Growth	0.024	0.040	−0.057, 0.099
SWB → Family Harmony → Spiritual Growth	0.020	0.008	0.004, 0.034

Abbreviations: BootSE, bootstrap standard error; CI, confidence interval; SWB, Subjective Well-being; HPB, Health-Promoting Behaviors.

Figure 3.

Mediation models for the six domains of health-promoting behaviors.

Notes: ①–⑥ denote the six HPB domains: ① Health Responsibility; ② Physical Activity; ③ Nutrition; ④ Interpersonal Relationships; ⑤ Stress Management; and ⑥ Spiritual Growth. *P < 0.05; **P < 0.01; ***P < 0.001. Solid lines indicate statistically significant paths, whereas dashed lines indicate nonsignificant paths. a, the effect of SWB on family harmony; b, the effect of family harmony on the respective HPB domain; c, the total effect of SWB on the respective HPB domain; c′, the direct effect of SWB on the respective HPB domain.

Abbreviations: SWB, Subjective Well-being; HPB, Health-Promoting Behaviors; HR, Health Responsibility; PA, Physical Activity; NU, Nutrition; IR, Interpersonal Relationships; SM, Stress Management; SG, Spiritual Growth.

This study further examined the pathways linking SWB to HPB using ethnicity-stratified analyses. As shown in Table 4 and Figure 4, among the Han majority group, SWB showed a significant total effect on HPB (Effect = 0.177, 95% CI [0.096, 0.259]) and a significant direct effect (Effect = 0.149, 95% CI [0.075, 0.222]), whereas the indirect effect via family harmony was not statistically significant (Effect = 0.029, 95% CI [−0.008, 0.068]). These findings suggest that, for the Han majority group, the association between SWB and HPB was primarily direct rather than indirect through family harmony. In contrast, among the Chinese Korean minority group, the total effect of SWB on HPB was not statistically significant (Effect = 0.092, 95% CI [−0.063, 0.247]) and the direct effect was also non-significant (Effect = −0.015, 95% CI [−0.165, 0.134]), whereas family harmony demonstrated a significant indirect effect (Effect = 0.107, 95% CI [0.033, 0.203]). These results indicate an indirect-only statistical mediation pattern in the Chinese Korean minority group, characterized by a significant indirect effect via family harmony and a nonsignificant direct effect. Detailed ethnicity-stratified regression and mediation results are provided in Supplementary Appendix 2, Tables S16–S19.

Table 4.

Ethnicity-Stratified Mediation Results for the Association Between SWB and HPB

Path	Effect	BootSE	Boot 95% CI
Han majority
Total effect	0.177	0.042	0.096, 0.259
SWB → HPB	0.149	0.038	0.075, 0.222
SWB → Family Harmony → HPB	0.029	0.019	−0.008, 0.068
Chinese Korean minority
Total effect	0.092	0.078	−0.063, 0.247
SWB → HPB	−0.015	0.075	−0.165, 0.134
SWB → Family Harmony → HPB	0.107	0.043	0.033, 0.203

Abbreviations: BootSE, bootstrap standard error; CI, confidence interval; SWB, Subjective Well-being; HPB, Health-Promoting Behaviors.

Figure 4.

Comparison of the mediation models between the Han majority and Chinese Korean minority groups.

Notes: **P < 0.01; ***P < 0.001. Solid lines indicate statistically significant paths, whereas dashed lines indicate nonsignificant paths. The vertical dashed line separates the two ethnicity-stratified models. a, the effect of SWB on family harmony; b, the effect of family harmony on HPB; c, the total effect of SWB on HPB; c′, the direct effect of SWB on HPB.

Abbreviations: SWB, Subjective Well-being; HPB, Health-Promoting Behaviors.

Discussion

This study examined the relationship between SWB and HPB among middle-aged women in Northern China and assessed whether family harmony may serve as a relational pathway in this association. The findings indicate that family harmony may be a relevant psychosocial correlate and mediator in the association between SWB and HPB, suggesting that midlife women’s health routines may be shaped not only by individual psychological resources but also by family relational contexts.^16,34 Beyond the overall pattern, results revealed heterogeneity across HPB domains. The indirect pathway through family harmony was observed for most domains, but not for interpersonal relationships. Additionally, ethnicity-stratified analyses suggested different patterns between Han majority and Chinese Korean minority women, underscoring the potential importance of sociocultural context in understanding how well-being is associated with health-related behaviors.^35,36

Middle-aged women reported a good level of overall HPB, indicating favorable engagement and ongoing opportunities for improvement in maintaining health routines. Considered alongside prior surveys and related evidence,^11,37 this level was somewhat higher than the moderate levels of health-promoting behaviors reported in comparable populations. This pattern may reflect a balance between increasing attention to health and persistent constraints on the adoption and maintenance of healthy routines during midlife.³⁸ Differences in sampling frames, settings, and measurement approaches should nevertheless be considered when making cross-study comparisons. At the individual level, SWB showed a positive association with HPB, supporting the view that SWB may function as a psychological resource for proactive health management during midlife, a period often marked by physiological changes and competing role demands.^1,3,7 Notably, SWB was also positively associated with family harmony, suggesting that family relational context is relevant when considering how psychological well-being is reflected in everyday health-promoting practices.²² This observation supports the relevance of considering both individual psychological resources and relational environments when promoting HPB in midlife women.

Mediation analyses indicated that family harmony partially mediated the association between SWB and HPB among middle-aged women. One plausible explanation is that individuals with higher SWB may be more likely to engage in constructive family interactions, including supportive communication and collaborative problem-solving, which can foster a more harmonious family climate conducive to maintaining health routines.²¹ This interpretation aligns with broaden-and-build theory,²⁰ which posits that positive affect helps build enduring social resources, including supportive exchanges within close relationships. A harmonious family environment may, in turn, reduce interpersonal friction and facilitate both emotional and practical support, such as shared responsibilities and co-participating in healthy routines, thereby lowering the effort required to sustain daily HPB.³⁹ Moreover, consistent with the stress-buffering hypothesis,²³ family harmony may mitigate the impact of daily stressors and help preserve self-regulatory capacity, enabling sustained engagement in HPB during midlife. The indirect effect through family harmony accounted for 30.22% of the total association between SWB and HPB, highlighting the practical relevance of family relational contexts to the observed association. Accordingly, strategies that support family harmony may be incorporated into multicomponent community-based health-promotion programs. In practice, health-promotion programs could integrate individual-level support—such as interventions to enhance well-being and motivation—with straightforward family-based strategies. For instance, programs could facilitate shared goal setting for diet and physical activity, support collaborative planning and equitable distribution of household responsibilities, and encourage co-participation in activities such as walking, meal preparation, and stress-management routines.⁴⁰

Domain-specific findings suggest that the family-harmony pathway may be more evident for certain HPB domains. In the present study, family harmony significantly mediated the association between SWB and five HPB domains—health responsibility, physical activity, nutrition, stress management, and spiritual growth—supporting the view that close family processes may contribute to multiple self-regulatory and routine-based health practices.⁴⁰ This pattern is broadly consistent with evidence that household relationship contexts may support health behaviors through practical assistance, encouragement, and the reduction of everyday friction, thereby lowering the “cost” of sustaining routines over time.^41,42

By contrast, family harmony did not mediate the interpersonal relationships domain. One plausible explanation lies in construct mismatch. In the HPLP-II, interpersonal relationships domain reflects broader interpersonal connections and supportive exchanges, including discussing concerns with people close to them, and is not necessarily limited to within-family functioning.^2,3 Family harmony, however, is conceptually and operationally centered on intra-family functioning; therefore, it may be less predictive of relational functioning that depends more heavily on extra-familial ties such as friendships, peers, and community engagement.^19,39 This distinction is also meaningful in midlife, when women may experience work–family conflict and spousal caregiving demands.^43,44 These findings suggest that promoting HPB may require attention to both supportive family involvement in daily routines and opportunities for social connection beyond the household. Community programs could provide low-threshold group activities—such as walking groups, community dance classes, and group stretching—as well as peer-support meetups and interest-based clubs. Simple “buddy” systems with achievable participation goals may further enhance adherence and contribute to sustainable social support over time.⁴⁵

Ethnicity-stratified analyses suggested different patterns in the SWB–HPB association. Among Han majority women, the direct association between SWB and HPB was significant, whereas the indirect effect via family harmony was not significant. By contrast, among Chinese Korean minority women, the indirect effect via family harmony was significant, whereas the direct association was not significant. Previous research suggests that family roles, spousal support, intergenerational relationships, caregiving responsibilities, and employment arrangements may differ between Han and Chinese Korean families.^25,26 One possible interpretation is that differences in family roles and support structures may shape both how SWB is reflected in family harmony and how family harmony translates into practical and emotional support for maintaining healthy routines. However, because these processes were not directly assessed in the present study, this interpretation remains tentative and should be regarded as a hypothesis for future research. Future studies should directly examine how family roles and support structures may shape the role of family harmony across ethnic groups. From a practical perspective, health-promotion programs should assess local family contexts and participant preferences and offer both individual-level and family-engaged components, rather than applying fixed ethnicity-specific prescriptions.^40,46

Limitations

Several limitations should be acknowledged. First, the cross-sectional design precludes causal inference; longitudinal studies are needed to clarify temporal directionality and potential dynamic pathways. Second, all variables were self-reported and thus susceptible to recall and social desirability biases. However, data collection followed a standardized protocol administered by trained investigators, and all instruments were established measures with demonstrated reliability and validity, which may help mitigate measurement error. Third, the sample was drawn from two cities in Northern China, and the Chinese Korean minority subsample was relatively small; therefore, generalizability—particularly of ethnicity-stratified results—should be interpreted with caution. Future multi-site studies with broader regional and ethnic representation, along with tests of measurement invariance across groups, are warranted to enhance external validity.

Conclusion

Family harmony partially mediated the association between subjective well-being and health-promoting behaviors among middle-aged women in Northern China, with indirect associations observed across most behavioral domains and different patterns in ethnicity-stratified analyses. These findings highlight the potential value of combining individual psychological support with family-engaged components in community health promotion and tailoring their emphasis to the sociocultural contexts of different population groups.

Funding Statement

National Natural Science Foundation of China funded project (72264020); Yunnan Philosophy and Social Science Innovation Team (2024CX08); First-Class Discipline Team of Kunming Medical University (2024XKTDTS16); Yunnan Graduate Tutor Team Construction Project (2025-14-63).

Data Sharing Statement

The datasets generated and analyzed during the current study are not publicly available due to privacy considerations. When collecting the data, participants were informed that their data would not be made publicly available to maintain confidentiality. However, the data are available from Hailian Zhang, a corresponding author, upon reasonable request, subject to compliance with any applicable privacy regulations and consent agreements.

Ethics Approval and Consent to Participate

This study was approved by the Institutional Review Board of Yanbian University School of Medicine (Approval No. 20220516-1088) and conducted in accordance with the ethical principles of the Declaration of Helsinki. Informed consent was obtained from all participants prior to data collection. Participants were assured of voluntary participation, anonymity, and confidentiality, and all procedures complied with institutional and academic ethical standards.

Author Contributions

All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.

Disclosure

The author(s) report no conflicts of interest in this work.