Introduction

Depression is one of the most common mental and psychiatric disorders among older adults [1]. Globally, the prevalence of depression reaches a peak in older adulthood, with a prevalence of 7.5% among women and 5.5% among men [2]. Depressive disorder is a considerable public health concern for older adults in China. It is estimated that the overall prevalence of depressive symptoms among older Chinese people is 20.0% [3]. The burgeoning ageing trend and severe outcomes of late-life depression (e.g., increased mortality rate and suicidal behaviors) [4, 5] have led studies to investigate the multiple risk factors of depression among older adults, including the psychosocial (e.g., social stress), behavioral (e.g., adverse lifestyles), and physiological (e.g., poor physical health) factors [6]. Researchers have discovered that the quality of intergenerational relationships with children plays a critical role in sha** the mental health status of older Chinese adults [7,8,9].

Derived from Confucianism, the norm of filial piety is culturally expected and practiced cross generations in traditional Chinese society [10]. Under this social norm, adult children have an obligation to provide care for and respect their older parents [11]. Therefore, intergenerational interactions are critical sources of social support for older people, and affectual bonds between older parents and adult children comprise a key kinship network within Chinese families. However, the considerable evolution of Chinese society, such as the narrowing family structure, altered intergenerational living arrangements, and extended educational period of adult children [12], may bring more tensions, disagreements, and pressure to the relations with offspring, threatening older people’s subjective well-being [13]. Therefore, the extent to which and the manner in which intergenerational relationships influence mental health of older adults have novel implications within the context of these transformed social norms.

From the perspective of solidarity [14], conflict [15], and ambivalence [16], the features and structure of intergenerational relationships are characterized by multidimensional properties. Intergenerational relationship quality should not be measured only through binary propositions, but rather multiple aspects, including intergenerational contacts and interactions, similarities, affectual closeness, and intergenerational conflict [17]. However, few studies have employed a multifaceted perspective to evaluate the implications of relationships between older parents and adult children, or employed multiple domains to holistically evaluate intergenerational relationships. Most of the studies have focused only on the influences of intergenerational support and transfers [18,19,20] and contact frequency [21, 41]. IADL scores ranged from 8 to 24, with a higher score indicating poorer functional health status.

Data analysis

Data analyses were performed using SPSS 26 [42]. Descriptive analyses were conducted first to obtain the mean and standard deviation (SD) for each key study variable, and intercorrelations between study variables were subsequently examined. The mediating role of sense of loneliness in the association between intergenerational relationship quality and each of its four subdomains on depressive symptoms was further tested using the PROCESS macro in SPSS [43]. Age, gender, marital status, employment status, educational attainment, self-perceived economic status, number of children, number of chronic diseases, and IADL scores were included as covariates. The total effect (c), direct effect (c’), and bootstrap-based confidence intervals (CIs) of the indirect effect (ab) were calculated using 5,000 bootstrap samples. Significant mediating effects were indicated by 95% CIs that did not contain zero [44].

Results

Table 1 lists the socio-demographic characteristics of the participants (n = 801). The average age of the participants was 68.58 years, and 43% of the participants were men. Approximately 60% of the participants were married and living with their spouses, whereas others were divorced or separated and widowed or never married. Three-fourths were retired or no longer working, and 25% of them still had part-time or full-time jobs. The average educational attainment (SD) of participants was 2.43 (1.27), which was between primary school and middle school. The mean (SD) of the self-perceived economic status score was 2.95 (0.60). Participants had an average of 2.43 children. The mean (SD) number of chronic diseases was 1.07 (1.12). The mean (SD) score of the 8-item IADL Scale was 8.77 (2.20), indicating that participants had a relatively high ability to live independently.

Table 1 Characteristics of participants (n = 801)
Full size table

Table 2 presents the means (SD) and bivariate correlations among the study variables. Compared with other domains, the average score for consensual–normative solidarity (mean = 8.49) was the lowest, indicating that older parents and adult children struggle to achieve similarity in their social values or beliefs in contemporary Hong Kong. This substantial disagreement between generations was also reflected in the perception of intergenerational conflicts (mean = 11.98), with scores indicating that older parents frequently perceive or experience tensions with their adult children. Similarly, scores for affectual closeness were low (mean = 11.01), indicating that parents struggle to get along with and maintain intimate relationships with their adult children. Regarding correlations, the results demonstrated that depressive symptom scores were negatively correlated with overall intergenerational relationship quality (r = -0.411) and the four subdomains (consensual-normative solidarity: r = -0.226, structural-associational solidarity: r = -0.276, affectual closeness: r = -0.386, and intergenerational conflict: r = -0.164) and were positively associated with sense of loneliness (r = 0.462). P-values of all above-mentioned correlations were lower than 0.001.

Table 2 Correlations among study variables (n = 801)
Full size table

A basic mediation model was employed using the PROCESS macro to investigate the role of loneliness in the associations between intergenerational relationships and depressive symptoms. The model (Fig. 2) explained 38.9% (F = 46, p < 0.001) of the variance in depressive symptoms after adjustment for age, gender, marital status, employment status, educational attainment, self-perceived economic status, number of children, number of chronic diseases, and IADL scores. Both the total (B = -0.058 (0.006), t = -9.020, CI [-0.071, -0.045]) and direct effects (B = -0.034 (0.007), t = -5.017, CI [-0.047, -0.020]) of overall intergenerational relationship quality on depressive symptoms were significant. Furthermore, sense of loneliness partially mediated this relationship, with the significant indirect effect (B = -0.024 (0.003), CI [-0.031, -0.018]).

Fig. 2
figure 2

Intergenerational Relationship Quality and Depressive Symptoms Mediated by Sense of Loneliness. Note. Age, gender, marital status, employment status, educational attainment, self-perceived economic status, number of children, number of chronic diseases, and IADL score were controlled. Indirect effect (ab): B = -.024, SE = .003, CI 95% = -.031 to -.018. *** p < 0.001

Full size image

Further analyses were performed using the four subdomains of IRQS as independent variables in the mediation model. As presented in Fig. 3, after controlling for aforementioned nine covariates, the total effects of consensual–normative solidarity (B = -0.072 (0.018), t = -3.973, CI [-0.107, -0.036]), structural–associational solidarity (B = -0.066 (0.013), t = -5.132, CI [-0.092, -0.041]), affectual closeness (B = -0.194 (0.019), t = -10.388, CI [-0.231, -0.158]), and intergenerational conflict (B = -0.059 (0.018), t = -3.322, CI [-0.094, -0.024]) on depressive symptoms were all significant. The results also indicated that among the four subdomains of the IRQS, affectual closeness produced a larger influence on mental health among older adults.

Fig. 3
figure 3

Intergenerational Relationship Sub-Domains and Depressive Symptoms Mediated by Sense of Loneliness. Note. Age, gender, marital status, employment status, educational attainment, self-perceived economic status, number of children, number of chronic diseases, and IADL score were controlled. a. Indirect effect (ab): B = -.033, SE = .008, 95% CI = -.048 to -.018. b. Indirect effect (ab): B = -.026, SE = .005, 95% CI = -.037 to -.016. c. Indirect effect (ab): B = -.065, SE = .009, 95% CI = -.084 to -.048. d. Indirect effect (ab): B = -.059, SE = .008, 95% CI = -.077 to -.044. *** p < 0.001, ** p < 0.01, * p < 0.05

Full size image

The direct effects of consensual–normative solidarity (B = -0.039 (0.017), t = -2.304, CI [-0.072, -0.006]), structural–association solidarity (B = -0.040 (0.012), t = -3.281, CI [-0.064, -0.016]), and affectual closeness (B = -0.130 (0.019), t = -6.709, CI [-0.167, -0.092]) on depressive symptoms were significant, only that of intergenerational conflict (B = 0.000 (0.017), t = -0.013, CI [-0.034, 0.034]) was insignificant and zero. This result suggested that sense of loneliness completely mediated the effect of intergenerational conflict on depressive symptoms (indirect effect: B = -0.059 (0.008), CI [-0.077, -0.044]), indicating that loneliness can fully explain the process through which intergenerational conflict influenced depression [45]. The indirect effects of consensual–normative solidarity (B = -0.033 (0.008), CI [-0.048, -0.018]), structural–associational solidarity (B = -0.026 (0.005), CI [-0.037, -0.016]), and affectual closeness (B = -0.065 (0.009), CI [-0.084, -0.048]) were significant, indicating that loneliness partially mediated the associations between these domains of parents–adult children relationships and depressive symptoms [45].

Discussion and Implications

This study employed multiple domains of older parents–adult child relationships to systematically investigate the direct and indirect influence of intergenerational relationships on older people’s depressive symptoms through the mediating role of sense of loneliness. All proposed research hypotheses were supported. The associations between the four domains of intergenerational relationships and depressive symptoms were mediated by sense of loneliness. Among the four subdomains, affectual closeness showed a relatively stronger association with depressive symptoms. Regarding the quality of perceived parent–adult child relationships, older adult participants rated consensual–normative solidarity the lowest. The results of this study highlight the importance of harmonious parent–child relationships in preventing mental problems among older adults.

Among the subdomains of intergenerational relationship quality, consensual–normative solidarity received the lowest score. This reflected the divergence of generational attitudes toward social norms and social issues, which was consistent with the findings of a recent survey in Hong Kong where age-related differences in public attitudes towards welfare policy were significant [46]. From the perspective of intergenerational cultural transmission [47], one generation transmits cultural values, beliefs, knowledge, and practices to the next generation. However, the increasing prevalence of individualization, economic independence, and migration [48], have prevented the traditional social norms from being easily transmitted to younger generations. Social values may be altered by generational traits and social experiences that become attached to generationally distinct “collective memories” [49]. Moreover, in a dynamically changing society, such as that of Hong Kong, intergenerational consensus can influence both family harmony and personal well-being, and social stability [50]. Therefore, it is crucial to reinforce mutual understanding between generations at both family and community levels. Service providers could organize intergenerational programs involving both younger and older adults to increase dialogue across generations and improve mutual understanding.

We further found that all four domains of intergenerational relationship quality were significantly associated with depressive symptoms in older adults. This result supplemented the work of Fu and Ji [7] by finding that structural–associational solidarity was also associated with depressive symptoms in Hong Kong older people. Moreover, a perceived intimate affectual connection with adult children demonstrated the strongest association with depressive symptoms, supporting that forming close affectional bonds is a universal human need [51]. Affectual closeness is characterized as a positive emotional sentiment or feeling toward intergenerational solidarity [34], and manifests as intimacy in parent–adult child relationships; getting along well; and providing money or gifts to parents, which are core expectations in traditional filial obligation [10]. In other words, conventional filial expectations remain robust among older Chinese in Hong Kong, and the fulfillment of these core filial norms in the interactions with adult children substantially influences older adults’ loneliness experiences and mental health. As suggested by the study findings, improving the various domains of parent–adult children relationship quality, especially the affectual closeness between generations is pivotal to reducing the risks of suffering depressive symptoms in older people. Policymakers and service providers should promote public education on the strategies of enhancing intergenerational solidarity and provide opportunities for younger generations to more actively engage in intergenerational interactions to strengthen affectual bonding and mutual understanding. Particularly, providing frequent emotional support and financial support to parents are effective ways to improve parents’ mental health, which are highly recommended in the context of declined filial ideology among younger generations [52].

This study also provided evidence pertaining to the mediating effects of loneliness in the association between overall and multiple domains of intergenerational relationships and depressive symptoms among older adults. This result supported the psychological pathway of the conceptual model proposed by Berkman et al. [31], who argued that mental health outcomes could be influenced by social relationships through multiple mechanisms. As a critical constitution of kinship in later life, the quality of intergenerational relationships was remarkably associated with the satisfaction of social and emotional needs among older people, and was further related to their mental health. In addition, this study enriched the understanding of intergenerational relationships by broadly examining the multifaceted components underlying the aforementioned mechanism. It was found that sense of loneliness completely mediated the effect of intergenerational conflict on depressive symptoms and partially mediated the effects of the remaining three dimensions. Undesirable parent–adult child relationships, such as those without children’s support, with reduced interactions, and with intensified conflicts, may initiate and reinforce older adults’ feelings of loneliness, thereby leading to mental problems or affective disorders [53, 54]. Therefore, in addition to improving intergenerational relationships, it would be advisable to encourage older adults, especially those with unsolvable intergenerational conflicts to participate in diverse social activities, such as community-based voluntary activities, physical sessions, and leisure or recreational activities. By doing so, older people can prevent depressive symptoms through broader networks and a reduced sense of loneliness.

Study limitations

This study has several limitations. Firstly, the characteristics of a cross-sectional study necessitate causal inferences of data analyses to be interpreted with caution; a longitudinal design should be employed to further verify our results. Secondly, the response rate (50.9%) did not reach a high level in this study, which may cause the potential of nonresponse bias [55]. Thirdly, as indicated by the partial mediating effects of loneliness on the associations between the three subdomains of intergenerational relationships (i.e., consensual-normative solidarity, structural-associational solidarity, affectual closeness) and depressive symptoms, alternative mediators may be present in the pathways of the associations identified in this study. In addition to investigating loneliness from a psychological perspective, paths from behavioral and physiological perspectives must be explored in the future studies. Finally, because improving affectual closeness and decreasing loneliness are key to preventing depression in old-age, other sources of intimate relationships, such as romantic partners and close friends, warrant attention in subsequent investigations.