Association between age and loneliness in different residential type and gender groups: evidence from China

Most people have experienced loneliness at some point in life. Loneliness is usually defined as a negative subjective feeling of deficit in social contact, which is often caused by the loss or lack of social relationships [1, 2]. Weiss (1975) specified two types of loneliness: “social loneliness” which refers to the absence of social networks one gets highly engaged in, and “emotional loneliness” which refers to the absence of intimate attachment figs [3]. Loneliness has a detrimental effect on individuals’ well-being, i.e., adverse physical and mental health outcomes [4, 5], and can be related to problem behaviors like suicide [6]. Various factors were found associated with loneliness, including demographic, socioeconomic, health-related, and personality factors [7, 8]. That association, however, needs to be further verified, especially in developing countries where a lack of attention has been paid to it.

Age has been identified as a prominent predictor of loneliness [9]. Considering the available research results, there were two types of relationship between age and loneliness in the available research. First, linear relationships imply “the older, the lonelier” [10, 11]. Second, other studies found non-linear relationships. For example, Luhman & Hawkley (2016) found that the highest level of loneliness was experienced by young adults and the oldest old [12], while other scholars found an inverted U-shaped relationship [13]. Other studies found no linkage between age and loneliness. Therefore, there is no conclusive finding about the relationship between age and loneliness.

The extant literature on that relationship was mostly established in Western countries, lacking a description of the associations between loneliness and age in less industrialized countries like China. China is the world’s most populous country, accounting for 18% of the global population [6]. In past decades, China has experienced great internal rural-to-urban migration [14], as a result of more flexible economic policies and a requirement for more labor in industrial factories than agricultural fields. This has contributed to rapid urbanization [15, 16]. In the past four decades, China’s urban population grew from 17.9 to 63.9% [14]. One study in 2015, which used a nationally representative sample of middle-aged and older Chinese, found that over one-third of respondents in urban areas were rural-to-urban migrants. Correspondingly, over one-third of respondents in rural areas in China were left-behind family members [17].

Along with the rapid social changes in China, there is a sharp rise in social problems and disparities in welfare and mental health care provision [18]. The residential type system has become a social boundary distinguishing rural residents from urban residents. Compared with urban residents, rural residents have relatively lower formal social security and other resources, such as medical and educational services [19, 20]. Numerous studies have found that rural residents have higher levels of loneliness. It has been widely reported by rural children who are separated from their migrant parents and suffer more psychological problems [21, 22]. It is also evident that older adults residing in rural areas have an elevated risk for depression [23]. However, the relationship between age and loneliness with different residential types in China is unknown in the extant research.

Two perspectives informed our exploration of the association between age and loneliness among people with different residential types. First, the distribution of loneliness may be inferred from the perspectives of collectivism and individualism [24, 25]. According to the WHO mental health surveys, mental health problems such as depression and anxiety had an overall lower prevalence among the populations in collectivist countries [26]. As for loneliness, it was believed that compared to residents in collectivistic countries, those residing in individualistic countries tended to have a lower level of loneliness [27, 28]. In the context of China, based on the general observation, it is well acknowledged that the components of individualism are more adaptive to the urban environment, while components of collectivism are more adaptive to the rural environment [29, 30]. This implies that urban and rural residents will have different associations of loneliness with age.

Second, the transition of individuals’ social and behavioral features can occur with the process of urbanization. Along with China’s unprecedented urbanization, the effect of the cultural mismatch between rural and urban areas is evident. The transformation in people’s living environments can lead to shifts in socialization and cultural values. For example, residents who are exposed to the urban and individualistic environment but still hold rural and collectivistic cultural values might find themselves maladaptive to the environment [31]. Accordingly, the rural-to-urban migrants would experience loneliness in the transition.

Recent studies demonstrated that loneliness was more common in urban areas [32, 33]. Thus, it is rational to assume that some urbanized Chinese of different ages are at greater risk of loneliness than their rural counterparts. For example, young adults who lived alone in cities for higher education or jobs experienced high living pressure, simple social connections and rare communication with others [34, 35]. It could take them more effort to rebuild their social networks, and they were found to have a higher level of loneliness than their rural counterparts [36, 37]. Residents living in rural areas experience more stable social networks to counter loneliness. This contrasts with urban residents, where the individualistic environment caused them to be more vulnerable to loneliness. Many urban residents also experience a cultural value mismatch. It is reasonable to assume that a difference can exist in the age distributions of loneliness between Chinese residing in urban and rural areas. It is meaningful to investigate the potential difference and target the specific groups of the population that deserve more attention and assistance in addressing loneliness.

The association between gender and loneliness is often assumed. Some scholars posit that women are more vulnerable to loneliness than men [38, 39]. This is because women live longer, and they are more likely to be spousal caregivers and experience widowhood. It is also known women might have higher expectations of social contact [40, 41]. Nevertheless, other scholars argue that men tend to spend more time alone from adolescence, and negative life events in late adulthood—such as losing loved ones—had a more detrimental impact, which increases loneliness [42]. The empirical results are inconclusive about whether women are lonelier than men [43, 44]. Therefore, despite the different development trajectories and characteristics of men and women, their gender roles in relationships are socialized to some extent [45], so the impact of gender could be moderated by other factors, such as age and social context [43]. Additionally, previous research mainly focused on gender comparison in particular developmental periods, for example, in young [12] or old age [40, 41], while there is a lack of lifespan perspective. Therefore, it is meaningful to test the potential gender difference in loneliness without a fixed a-priori hypothesis and explore the potential gender difference in the associations of loneliness with age.

To summarize, few studies have explored the age distribution of loneliness in China. The context in China warrants an exploration of the relationship between age and loneliness on the condition of residential type and gender. To address this gap, this study employs a nationally representative dataset to answer the following three questions: First, what is the relationship between age and loneliness among the Chinese population? Specifically, is there a linear or curved correlation? If a curved correlation exists, will it be a U-shape or an inverted U-shape? At what age does the highest level of loneliness appear? Second, could the age distribution of loneliness vary by residential type? Third, could the age distribution of loneliness vary by gender?

The findings of the association between age and loneliness supported some previous studies which suggested that there was an inverted U-shaped relationship between age and loneliness [8, 12]. The results showed that the loneliness of the entire sample followed an inverted U-shaped curve. The average score of loneliness of 3.72 and over 70% of the respondents with a score of 3 suggested that, for most of the respondents, loneliness was not a significant problem, which supports previous research [54‐56].

The finding of a higher level of loneliness among rural residents was consistent with the assertion that loneliness was prevalent among older adults in rural China [57]. This is because intergenerational emotional cohesion is lowered when the younger generation migrates to urban areas for work [58]. The majority of rural-to-urban migrants still hold rural hukou status as required by the hukou policy [59]. Epidemiological studies have found that many of them experience high levels of loneliness, resulting from social isolation, discrimination, and workplace injustice [60, 61]. Consequently, respondents who hold rural hukou can have an overall higher risk of loneliness compared to their urban counterparts, and that could derive from the change of living arrangements and weakened family cohesion, as well as maladaptation in urban areas [62].

This study also shows that urban residents’ age and loneliness followed an inverted U-shape curve with a peak at the age of 47, which differs from previous studies in industrialized countries, where loneliness is more prevalent in young adults or older adults than middle-age adults [11, 12]. There are two possible explanations for the findings. First, middle-aged people’s desire for social interaction may require higher engagement than younger adults. Middle-aged people may expect network interactions of higher quality or closeness, while the size of the network or romance of the relationship are not the key factors [63]. Therefore, although middle-aged people could master more social resources, their desire for interaction might not be met. The second explanation relates to the Chinese context. Most people experience loneliness at every life stage. Nevertheless, the drivers of loneliness in middle adulthood include shifts in family structure, career progression, and changes in health status. However, these challenges could be more difficult to manage in China than in industrialized countries [64]. Since the social welfare system for children and older adults is still underdeveloped in China, Chinese middle-aged urban residents may bear more responsibility for family care, and that might limit engagements that might meet their desire for social relationships.

In our study, there was no significant difference in the effect of gender on loneliness in the fully adjusted models. This result is similar to a previous meta-analysis [43], which found that men and women had similar levels of loneliness throughout life. However, our study found that men’s loneliness followed an inverted U-shape curve, with the peak value at the age of 55, which differs from another study in which loneliness peaked at the ages of 40 and 80 [8]. The finding provided sound evidence of the peak time of loneliness which was not specified in previous studies. This result could be explained by the persistent drop in Chinese female labor force participation in the past 30 years (from 79 to 43%), reflecting the return of the traditional Chinese gender division of labor—“men are breadwinners and women are homemakers” [65]. Accordingly, being male could evoke anxiety about the bureaucratic process, health management, and pre-retirement, which negatively impact men’s social interaction [66]. Moreover, one empirical research found that the division norm was negatively correlated with husbands’ marriage satisfaction [67], which is another risk factor for men’s loneliness. Therefore, middle-aged men were more vulnerable than men at other life stages.

Lastly, our study supports previous findings that being unmarried, having low education, low self-reported health level, co-residing with fewer families, and having no pension were associated with a greater level of loneliness for the entire sample. However, having a pension plan or not mattered only to the female and rural respondents.

Several limitations should be considered when interpreting the findings. First, this study employed a cross-sectional design which did not allow an investigation of individuals’ early life events and their adaptive process, making it impossible for this study to investigate the developmental trajectories of loneliness of each subgroup. As the exploration of prolonged loneliness—which is probably true for some subgroups of respondents—could help to better explain their current level of loneliness [68], it is necessary to examine the relationship between age and loneliness with a longitudinal design. Second, there should have been questions in the survey asking respondents for more information about their social economic status (e.g., their employment, the length of time they possessed urban hukou, and their social capital), social integration, and coping strategies they use to adapt to the urban environment [69]. Third, a more suitable measurement tool for loneliness is needed for future research. One recent study demonstrated that emotional loneliness among the Chinese could have a better indication of the process of urbanization than social loneliness [48]. Nevertheless, the measurement tools of this study had only three items and could not differentiate between emotional and social loneliness. In fact, CGSS2017 is part of the International Social Survey Program (ISSP). In previous studies [70, 71], this scale demonstrated good applicability and representativeness for measuring loneliness. Future studies are expected to provide further analysis of the relationship between age and dimensions of loneliness. Fourth, there are other well-studied factors associated with loneliness that should be included in future research, such as respondents’ personalities.

Despite the limitations, to the best of our knowledge, this study is the first one that used a nationally representative sample to investigate the loneliness of the Chinese population from young adulthood to late adulthood, instead of just focusing on groups at specific life stages. It treated age as a continuous variable instead of categorizing age into groups, allowing it to be among the first studies to explore the age distribution in loneliness. Moreover, this study considered gender and residential type and enriched the literature on loneliness in China and the social context of the nation.

This study examined the relationship between age and loneliness, in terms of gender and residential type by using a nationally representative dataset of 3899 adults in China. The respondents reported overall low-level loneliness. The female and rural subgroups had higher levels of loneliness than their counterparts. The difference between men and women was insignificant after controlling for all covariates. The regression results demonstrated that an inverted U-shaped tendency between age and loneliness existed for the entire group, and the male and urban subgroups. That tendency did not apply to the female and rural subgroups.