Introduction
The global trend of population ageing presents critical challenges worldwide, with China being no exception to this demographic transformation. The projected steep increase in the number of Chinese older adults (aged 60 and above), from 254 million in 2019 to 402 million in 2040, highlights the urgency of this issue. This substantial growth represents a formidable 28% of the total population aged 60 or above [
1]. China has already entered an ageing society and will continue to age rapidly in the future [
2]. In analogous situations, most developed countries have implemented vigorous measures to address the issues arising from an ageing population. Chinese policymakers and society need to adapt promptly to this demographic shift. The prospect of continued rapid ageing highlights the importance of necessitating preventive actions aimed at safeguarding well-being. The transition from focusing “how to live longer” to prioritizing “how to live better” marks a crucial juncture in the current ageing process in China. Improving the quality of life (QOL) of older adults is of key significance for raising healthy life expectancy [
3].
Studies have suggested that the QOL of older adults differs significantly from that of the general population [
4], thereby highlighting their unique needs and challenges. Hakka older adults not only face health and QOL challenges similar to those of other aging populations but are also deeply influenced by their unique cultural background and family socioeconomic status (SES). There are approximately 80 million Hakka people worldwide, with 50 million residing in China. The term “Hakka” not only refers to this ethnic group but also represents a rich cultural heritage developed over thousands of years [
5], particularly about wine culture. Due to the humid climate of their traditional settlements, Hakka people have developed Hakka wine, which is believed to dispel cold and nourish the blood, making it a common part of daily life. During festivals, wine serves as a customary gift and plays a vital role in honoring gods and ancestors. It is an essential medium for expressing affection and strengthening social bonds. Notably, Hakka older adults place great importance on preserving and practicing this cultural tradition. Despite extensive exploration of QOL dynamics across diverse age groups, including younger [
6], children and adolescents [
7,
8], and specific special populations [
9,
10], a notable gap persists in understanding of QOL specifically among older adults, particularly in the context of family SES. Recognizing the paramount significance and frequently overlooked aspect of QOL deprivation in older adults, it is imperative to delve into the modifiable factors that can positively influence and raise their living standards. Family SES plays a crucial role in shaping health-related behaviors, which contribute significantly to an individual’s overall health status [
11,
12]. Moreover, the efficacy of interventions designed to modify these behaviors exhibits a distinct population-specific profile [
13,
14], emphasizing the necessity for tailored strategies. Research focusing on older adults has demonstrated that targeted interventions aimed at health-related behaviors can effectively improve their QOL and overall health status [
15], thereby underscoring their potential for a positive shift.
Despite extensive inquiry into the individual associations between specific health-related behaviors and QOL, including physical exercise [
16], a healthy diet [
17], and sleeping [
18]. These available studies have focused on the relationship between specific health-related behaviors and QOL, overlooking the comprehensive impact of multiple health-related behaviors on QOL. Importantly, various health-related behaviors are interconnected and mutually influence each other, not occurring in isolation [
19].
Therefore, this research endeavors to bridge the existing gaps by meticulously exploring the associations between specific health-related behaviors, the number of healthy behaviors, behavior patterns, and QOL, considering the moderating effect of family SES.
Methods
Design and participants
Participants were recruited from Ninghua, China, where Hakka people gathered, through trained interviewers and the standard questionnaire during the 2018 Spring Festival using convenience sampling. The data collection was nested in a large cross-sectional survey, the 2018 China Health-related Quality of Life Survey of Older Adults (CHRQLS-OA 2018) [
20]. The data employed in this study encompassed participants’ social demographic information, health-related behaviors, and QOL. Several inclusion criteria were considered during the implementation of data collection: (1) participants aged 60 years old or above, (2) participants with local household registration, and (3) participants who voluntarily participated in this survey. The following participants were excluded: (1) participants who did not conduct normal conversation due to aphasia, deafness, or other critical body illnesses, (2) participants with severe mental disorders or had been diagnosed with cognitive impairment, and (3) participants who lost their daily living abilities. Consequently, we collected 1,500 questionnaires and 1,262 were valid after controlling data quality, with an effective rate of 84.13%.
Measures
QOL is an individual’s perception of their position in life in the context of the culture and value systems in which they live and concerning their goals, expectations, standards, and concerns [
21]. The assessment of QOL was conducted utilizing the Chinese version of the World Health Organization Quality of Life Instrument-Older Adults Module (WHOQOL-OLD) [
22,
23]. The Chinese version of the WHOQOL-OLD has demonstrated good internal consistency and reliability, as previous research shows [
22]. Appendix 1 presents the final version of the WHOQOL-OLD module.
QQL was assessed using the Chinese version of the WHOQOL-OLD is a five-point Likert scale, that consists of 24 items divided into 6 domains (each of the domains has 4 items), including Sensory Abilities (SAB), Autonomy (AUT), Past, Present, Future Activities (PPF), Social Participation (SOP), Death and Dying (DAD), and Intimacy (INT). QQL ranges from 24 to 120, and higher scores suggest better QOL.
Health-related behaviors were considered in three dimensions: specific health-related behaviors, the number of healthy behaviors, and behavior patterns. Specific health-related behaviors included “healthy diet”, “sleep regularity”, “physical exercise”, “smoking”, and “drinking”. The number of healthy behaviors is calculated as the sum of the incidences of “healthy diet”, “regular sleep”, “physical exercise”, “non-smoking”, and “non-drinking”. The number of health-related behaviors with a minimum count of 0 (10.70%) and a maximum count of 5 (16.96%) behaviors per individual. The detailed criteria and results concerning these health-related behaviors have been reported in our prior research [
24].
Family socioeconomic status (SES)
Family SES is typically indicated by the education and employment of family members, as well as family income [
25‐
27]. In this study, six indicators were selected to assess the family SES of older adults: education level, average annual household incomes (AHI, Chinese Yuan, CNY), current residence, employment before retirement, current employment, and personal savings were considered as confounder factors. Education level was classified as follows: illiterate, literacy class/home school, primary school, and junior high school or above. AHI were categorized as 15,000 or below, 15,001–30,000, 30,001–45,000, 45,001–60,000, and 60,001 or above. Current residence was categorized into village, town, and county. Employment before retirement was categorized as farming and non-farming. Current employment status was categorized as yes and no. Personal savings were categorized as 10,000 or below, 10,001–30,000, 30,001–50,000, 50,001–70,000, and 70,001 or above. Concurrently, the principal component analysis (PCA) method was employed to construct a family SES for a comprehensive evaluation, following established precedents in prior research. The specific family SES variables and their corresponding assignment protocols for this study are outlined in detail in Appendix 2.
Covariates
In this study, age, sex, education level, marital status, current residence, living arrangements, AHI, and current employment status were considered as confounder factors. Age was classified into the following categories: 60–64 years, 65–69 years, 70–74 years, 75–79 years, and 80 years and above. Sex was categorized as male or female. Education level was classified as follows: illiterate, literacy class/home school, primary school, and junior high school or above. Marital status was categorized as married / cohabitation, widowed, and others (unmarried, divorced, separated, etc.). Current residence was categorized into village, town, and county. Living arrangements were categorized as follows: living alone, living with spouse only, living with children, mixed habitation, and others (including living only with grandchildren, living with others, nursing home, etc.). AHI were categorized as 15,000 or below, 15,001–30,000, 30,001–45,000, 45,001–60,000, and 60,001 or above. Current employment status was categorized as yes and no.
Statistical analyses
Categorical variables were described as frequency (proportion). Continuous variables were described as mean (SD) because they are normally distributed. Two-Step Cluster Analysis (TCA) was employed in five health-related behaviors to determine the behavioral patterns among Hakka older adults. The TCA process serves as an exploratory tool that automatically identifies the optimal number of clusters [
28]. It consists of two main stages: preclustering and hierarchical clustering, and hierarchical algorithms utilize silhouette width to determine the best clustering number [
29]. For measuring clustering distance, the log-likelihood method is employed, while Schwarz’s Bayesian criterion (BIC) serves as the basis for clustering evaluation [
28,
29].
We used
t-tests and analysis of variance (ANOVA) to explore whether participants differed between specific health-related behaviors, the number of healthy behaviors, behavior patterns, and QOL. Generalized linear regression model was used to explore the relationship between specific health-related behaviors, the number of healthy behaviors, behavior patterns, and QOL. Two models were considered. Model 1 was the crude model. Model 2 was adjusted model, adjusted for age, sex, education level, marital status, current residence, living arrangements, AHI, and current employment status. The comprehensive score coefficients of the six items of education level, AHI, current residence, employment before retired, current employment, and personal savings were calculated by the method of principal component analysis, and then normalized as the weight coefficients of each item to construct the family SES (The weights of family SES indicators based on principal component analysis in this study are shown in Appendix 3). The PROCESS macro of Statistical Package for Social Sciences (SPSS) was used to analyze the moderation model [
30]. Model 1 was used to evaluate whether the family SES played a moderating role between the behavior patterns of Hakka older adults and QOL. All analyses were performed using SPSS 26.0 and Mplus (version 8.3). Two-sided
P < 0.05 was considered statistically significant.
Discussion
In this study, the association between specific health-related behaviors, the number of healthy behaviors, and behavior patterns and QOL were examined in Hakka older adults (aged 60 +). Sleep regularity and physical exercise were positively associated with QOL. The number of healthy behaviors (from 4 to 5) served as a protective factor for QOL. Compared with the moderate-health pattern, the risk-selective pattern and family SES were positively related to QOL. And the family SES exclusively moderated the risk-selective pattern on QOL.
Healthy diet [
17], sleep regularity [
31], and physical exercise [
32] significantly improve QOL among older adults. Prior research found a positive correlation between healthy dietary patterns and QOL [
33]. Dietary patterns of healthful plant-based food were considered a protective factor for cognitive ability [
34]. Sleep may positively influence QOL among older adults with type 2 diabetes [
35]. Light physical activity has been associated with improved overall cognitive function among older adults, particularly in domains such as orientation, attention, concentration, and language abilities [
36], and is beneficial to older adults’ physical well-being [
37]. Importantly, our findings showed the accumulation of health-related behaviors was positively correlated with higher QOL. Interaction roles in multiple health-related behaviors, affecting each other, not isolation [
19]. For example, eating habits influence sleep quality [
38], such as the Mediterranean diet, a widely acknowledged beneficial dietary pattern [
39,
40], has been associated with high-quality sleep [
41]. The combination of healthy eating and physical activity exhibits a preventive effect against certain diseases [
42,
43]. Meanwhile, appropriate physical exercise has been demonstrated to improve sleep quality [
44], leading to a higher QOL [
45]. The findings of this study were also confirmed in the Hakka older adults: participants with a higher number of healthy behaviors scored higher on the QOL domain (from 3.13 ± 0.32 to 3.74 ± 0.78), and were more likely to develop higher QOL (
P < 0.05).
In this study, there was no significant correlation between smoking and QOL, but it is worth noting that there was a significant positive correlation between smoking and SAB in QOL among Hakka older adults, which is consistent with previous studies [
46]. The age-related physical limitations in smokers are more pronounced compared to those in non-smokers, and the disparity increases with age [
47]. Notably, drinking was negatively related to with SAB in QOL. That is, the lower the probability of drinking, the more sensory abilities is affected, which is consistent with previous research [
48,
49]. Some studies also reported health benefits of moderate drinking [
50,
51], this notion remains controversial in the global literature [
52]. For example, drinking has been linked to improved mental health [
53], especially among those experiencing anxiety. On the other hand, excessive drinking has been associated with significant health risks, including higher alcohol consumption, episodic binge drinking, and the intake of high-alcohol-content beverages, all of which have been linked to increased mortality rates [
54]. In China, alcohol (such as Chinese Baijiu, Huangjiu, and Rice wine) plays a unique cultural role. In the cultural context, individuals tend to drink to alleviate their anxiety [
55]. In Hakka areas, in which malaria prevalence contributes to the custom of wine-making and medicinal ingestion. Moreover, health policies in China do not prohibit drinking, and the Chinese Centre for Disease Control and Prevention (CDC) indicates that moderate drinking (1–2 standard drinks/day) may confer health benefits, including a reduced risk of ischemic heart disease, stroke, and diabetes mellitus, among adults (aged 40 +). In summary, this policy framework aligns with our findings, which indicate a positive relationship between drinking and higher QOL, supporting the notion that moderate alcohol consumption may have health-promoting effects. When formulating drinking-related policies, it is essential to enhance public awareness by promoting moderate drinking. Additionally, improving the packaging and labeling of alcoholic beverages with clear guidelines on responsible consumption can help mitigate risks. Lastly, preserving local and traditional drinking cultures while balancing public health concerns is crucial. Policymakers should consider regional and ethnic drinking customs to ensure that regulations both protect public health and respect cultural heritage.
In SAB, family SES significantly buffered the negative impact of risk-selective pattern. These results align with previous research suggesting that higher SES can buffer the adverse consequences of health-risk behaviors through resource accumulation [
56,
57]. Our results suggest that family SES negatively moderated the positive effect of risk-selective pattern on PPF. The rewarding effect of alcohol temporarily enhances an individual’s sense of self-affirmation, especially when recognized as a “drinker status” in social interactions [
58]. A large number of studies have also demonstrated the positive effects of alcohol on mental health [
59,
60]. The negative moderating effect of family SES can be thought of as an alternative reward mechanism in social learning theory [
61], whereby high SES groups derive satisfaction from occupational achievements, social activities, etc., thus reducing alcohol dependence [
62].
In this study, family SES negatively moderated the positive effect of risk-selective pattern on SOP as well as INT, and past research suggests that alcohol consumption may act as a social enjoyment good that enhances social ties and interactions among people [
63]. Individuals who also had regular drinking places had more emotionally supportive friends and a stronger sense of community belonging [
64]. These findings are consistent with our findings. High SES individuals typically have broader and more diverse social networks that include not only family and friends, but also community leaders, educational institutions, and business executives, etc., and as a result, they show higher levels of motivation in social participation [
65]. Drinking behaviors high SES individuals are perceived negatively in some settings, which can reduce their effective participation and exert a negative effect [
66].On the other hand, our study showed that family SES buffered the negative effect of paradoxical-health pattern on INT. Consistent with our results, it has been shown that individuals who smoking and drinking for a long period of time have a higher rate of chronic diseases, require frequent medical or home recuperation, and have fewer opportunities to socialize [
66]. In contrast, high SES individuals often have access to higher quality healthcare services, including more advanced medical equipment, more professional healthcare teams, and more comprehensive treatment plans [
67], which can provide physiological safeguards for sustained social participation.
The results of this study suggest that family SES negatively moderated the positive effect of risk-selective pattern on DAD. Individuals with higher family SES who engage in smoking and drinking may mitigate the physical damage caused by these habits through access to better medical treatment [
68]. Consequently, despite these bad habits, their fear of death may not be significantly diminished [
69]. For total score, family SES primarily functioned as a risk buffer rather than a health amplifier, further reinforcing its role as a “buffer” against risky behaviors rather than an “amplifier” for health behaviors. This finding is consistent with the core mind of the Social Determinants of Health (SDH) theory which posits that SES improves overall well-being by reducing health disparities [
70]. The nonsignificant moderation of the paradoxical-health pattern implies that socioeconomic resources alone cannot resolve intrinsic behavioral contradictions.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.