Loneliness as a risk factor for frailty transition among older Chinese people

Populations worldwide are rapidly aging, which presents a particularly severe challenge in China, a country where the population is aging at a significantly faster rate than other low- and middle-income countries [1]. According to data published by the National Bureau of Statistics of China, 17.3% of the total population were aged 60 years and older in 2017. However, due to the implementation of the one-child policy in the earlier decades and recent increasing population mobility, the family size in China has been declining substantially, leading to over 50% of the urban and 60% of rural elderly living in empty-nest households, and the proportion of older people living alone in China was 12.5% in 2010, an increase by 30% in the prior two decades [2, 3]. These findings all have important implications for the health and social care of the elderly.

Frailty is the most outstanding expression of population aging [4]. It is a syndrome that predicts vulnerability to adverse outcomes and is recognized as a dynamic state with the potential for reversibility [4‐7]. There are currently two major models of frailty: first, the frailty phenotype model that views frailty from the physiological systems and defines frailty as several biological syndromes [8]; second, the frailty deficit model measures frailty as problems resulting from a multidimensional system, including biological, physiological and psychological [9]. Despite a sizeable literature on the adverse outcomes of frailty, such as falls [10], disability [11, 12], hospitalization [13], institutional care [14, 15], and mortality [16‐18], relatively little is known about the transition of frailty in older people. Although frailty is inevitable with increasing age [19], it is not irreversible but agreeable to be a dynamic process involving improvement and natural procession [4].

Loneliness is a common and dissatisfaction feeling of one’s social relationship that is presently becoming a serious public health issue for older people [20, 21]. Loneliness has been observed to be associated with subsequent adverse outcomes, such as mortality [22, 23], comorbidity [24, 25], poor functional ability [26, 27], depression [28], and cognitive decline [29]. At the biological level, many studies have found that the feeling of loneliness is associated with increased blood pressure [30, 31], increased risk of cardio-cerebrovascular and inflammatory diseases [32‐34], impaired immune function [35], and increased likelihood of sarcopenia [36].

Frailty, defined by the phenotype model or frailty index, is associated with loneliness [37‐39]. A cross-sectional study of Mexican community-dwelling elderly found that loneliness was independently associated with frailty [37]. The cohort study discovered that the relationship between frailty and loneliness might be bidirectional: loneliness was related to the change in frailty status, and vice versa [38, 40]. However, studies on the association between loneliness and frailty were focused on loneliness and frailty risk, and none have specified the association between loneliness and frailty transition types, including remaining frail, worsening or improving in frailty status. Furthermore, the existing research was based mostly on western society; much less is known for older people in other studies.

A study of Chinese older adults had reported that 51.2 and 7.0% of older adults aged 60 years and older were prefrail and frail, respectively [41]. However, little research exists on frailty transitions in China. It was reported that about 30.4% of participants had transitioned between different frailty statuses in 2002–2005 [42]. Additionally, one study showed that loneliness was related to culture and social policies [43]. People from collectivist cultures are more likely to feel lonely [44]. It was reported that about 30% of Chinese older adults reported feelings of loneliness [45]. We speculated that loneliness could have more influence on frailty transition in China. Gender difference in frailty and loneliness is well known [46, 47]. Women tend to have a higher incidence of frailty than men [48, 49], and it was suggested that this might be attributable to both biological and socioeconomic factors [50]. Studies have shown that loneliness is strongly associated with adverse health conditions in men and women [51, 52]. Thus, it can be assumed that the association between loneliness and frailty transitions is gender related.

Our study aimed to examine the association between loneliness and frailty transition among older adults older than 60 years in China. We generated two hypotheses: 1) loneliness is related to frailty transition, and 2) the relationship between loneliness and frailty is different by gender. We believe that our study would help to close the gap in the existing literature by using a nationally representative longitudinal sample in China, and the results would also be useful in informing policy making in health and social care.

The present study investigated the association between loneliness and frailty transitions. We used the 2008/2009, 2011/2012 and 2014 surveys of CLHLS for the analysis, focusing on the difference in the relationships between the male and female participants.

Nearly half of the participants remained in the robust or prefrail state in the follow-up years, regardless of gender. The percentage was higher than that in a previous study in China reporting that 39.6% of participants remained in the robust or prefrail state in 2002–2005 [42], while the percentage was close to the pooled frailty transition rates among 16 cohorts from 2010 to 2018 [68]. This difference may be due to the baseline time, variations in the follow-up year and measurement of frailty. Obviously, they may indicate that a window exists during which early interventions may be taken for the elderly to maintain their health status as much as possible. Furthermore, we found that changes from worsening to greater physical frailty tended to be more common than recovering from greater physical frailty, and this pattern of transition was consistent with that in a previous study [42, 68]. Evidence on gender differences in frailty transition is rare. One study of older people in Hongkong between 2001 and 2003 found that women were less likely to decline in frailty status than men [69], whereas a longitudinal study in San Antonio did not find men to be at higher risk of declining frailty status [70]. Another cohort study in older Italian adults found that women were more likely to progress into worsened physical frailty than men with a mean follow-up of 4.4 years [71]. Our study showed that, compared with men, women were more likely to change frailty status, either improving or worsening, a finding that agrees with a recent systematic review [68] and requires further confirmation by more studies. Thus, frailty interventions may have different efficacies in men and women.

Our study designed four physical frailty transition types—remaining robust or prefrail, improvement, worsening and remaining frail—to ascertain the specific relationship between loneliness and frailty transition. Previous studies have identified loneliness to be related to frailty [37, 38]. In our study, we found that greater loneliness reduced the possibility of remaining robust or prefrail physical frailty after 3 years, a finding that is consistent with a study in England, in which greater loneliness was found to be associated with increased risk of physical frailty around 4 years later [40]. Additionally, our study used two follow-up periods to validate the relationship. We found that loneliness increased the risk of older people with worsening frailty as well as those remaining frail after 6 years. However but no significant relationship was shown in the 3-year period. These findings may indicate that loneliness not only increases the possibility of frail in older adults but also increases the likelihood of older adults becoming frailer and chronically frail.

Our study also used another model of frailty, the frailty index, to further verify the results on the relationship between loneliness and frailty transition. We found a clear difference in the association between the levels of loneliness and frailty index between the two survey periods. In the 3-year follow-up, we found that the loneliness at baseline was related to recovering the frailty status in the frailty index. This may be explained by the frailty status at baseline because a severe baseline frailty status is more likely to be improved during the follow-up period. Correspondingly, the relationship between frailty improvement and loneliness was no longer significant in the 6-year follow-up, suggesting that shorter follow-up periods provide more time for older people to change their frailty status. During the 6-year follow-up, loneliness was found to be positively related to remaining in the frail status, but no significant relationship was found in the 3-year follow-up. This finding was supported by a previous study showing that a longer follow-up period was associated with lower rates of remaining in the same frailty state [68]. Additionally, the relationship between frailty index transition and loneliness differed from that between loneliness and physical frailty. For example, loneliness was found not to be associated with worsening in the frailty index, but with worsening physical frailty over time. The broader definition of the frailty index may not have the same risk factors as those of physical frailty [40]. Our findings confirmed this point, and more studies are needed to consolidate the possibility.

Our study also showed that loneliness varied by gender [52]. Low resilience was associated significantly with loneliness and was more pronounced in males [51]. Conversely, high resilience can be a protective factor in facilitating older people to maintain their health status [72, 73]. Previous research also indicated the stressful impact of loneliness on men, as manifested by increased inflammatory responses [74]. The inflammatory response has a specific physiological basis in the geriatric syndrome of frailty [75], which may be a mechanism underlying gender differences in the relationship between loneliness and physical frailty. Moreover, women tended to have more informal networks, which may lead to more social support, whereas men maintain their social relationships more from the public sphere, which may not always be socially supportive [59, 76]. Our study indicated that the male participants were more sensitive to the relationship between loneliness and frailty transitions, defined by either physical frailty or the frailty index. Most of the significant relationships observed in the male participants were identical to the findings for all participants. Thus, loneliness in males warrants more attention.

Given that frailty in older adults may be modifiable, our findings have potential implications for both health and social care policy and practice. First, our study provides a picture of the frailty transition among older people in China, highlighting the importance of early interventions for older people to maintain or improve their health statuses, particularly for those in the robust or prefrail state. Second, our study used two frailty models, which demonstrated that the accumulated disadvantages can endanger the elderly in the long term. Frailty management is more than treating specific clinical syndromes and physiological risk factors. Third, previous studies have identified that health behavior and management can be useful to delay or reverse frailty, such as physical activity, nutrition and rehabilitation [77]. The relationship between loneliness and frailty transition indicated that psychological treatment is also worthwhile for frailty intervention in older adults. Effective loneliness interventions can delay the progression of frailty. Finally, the findings on gender differences in our study suggest that attention should be given to older men with loneliness and its adverse outcomes.

The study has several limitations. First, among the total participants, 68% had completed data on the frail scale and 63–67% on the frail index in the two survey periods. Those who did not complete the questionnaire tended to be frailer and lonelier. Our findings may underestimate the relationship between frailty transition and loneliness. Second, as mentioned earlier, only one question concerned loneliness in CLHLS, likely underrating the prevalence of loneliness. However, a single question concerning loneliness has been widely used in studies and is more feasible for older adults to understand the investigation of loneliness [78, 79]. Finally, mental health variables, such as depression, were not included in the CLHLS.

Our study examined the association between loneliness and frailty transition among older people in China and attempted to explore gender differences in the relationships. The results revealed that loneliness at baseline may lead to a reduced possibility of remaining in the robust or prefrail frailty state and that greater loneliness is associated with an increased risk of worsening frailty and remaining frail. The association between loneliness and frailty transition differs obviously between men and women. These findings should be considered when designing and implementing health and social care policies for older people.