Factors influencing loneliness among older people using homecare services during the COVID‐19 pandemic

Abstract

Background

Elderly homecare service users may reduce their level of social participation and interpersonal interactions due to physiological loss, which may lead to loneliness and depression over the years. However, there is a lack of research on loneliness among older people who use homecare services. The purpose of this study was to examine the factors influencing loneliness among older people using homecare services.

Methods

This is a longitudinal study conducted in communities in Central Taiwan, and data were collected using a structured questionnaire. The questionnaire was first administered as a pre‐test to obtain baseline information about the participants, and the same questionnaire was administered as a post‐test after 6 months to follow‐up. The pre‐ and post‐test questionnaires included five sections, that is, participant demographics, Brief Symptom Rating Scale, Interpersonal Interaction Scale (IIS), Frenchay Activities Index, and UCLA Loneliness Scale (UCLA).

Results

A total of 178 participants were recruited in this study. Results indicated that gender, whether participants eat alone or with others at dinner, social media use, perceived economic status, and IIS score were significantly correlated with the loneliness score on the UCLA. Furthermore, there was a significant increase in the loneliness score among male participants in the low loneliness group from baseline to 6 months follow‐up.

Conclusions

Gender, presence of others at dinner, social media use, perceived economic status, and interpersonal interaction skills are significant factors that influence loneliness among older people using homecare services. Men tend to experience higher levels of loneliness over time.

INTRODUCTION

As the average life expectancy continues to rise, the ageing of the post‐war baby boomers in Taiwan has led to the rapid ageing of the population structure. The implementation of long‐term care services has become one of the important development policies in Taiwan, and the demand for and use of homecare services is one of the core elements of long‐term care services. ¹ The majority of homecare service users are individuals with disabilities, including those who are unable to independently care for themselves due to functional limitations caused by accidents, frailty, congenital, or acquired illnesses. These users are prone to loneliness due to physiological limitations, loss of identity caused by changes in family and social roles, and lack of social support networks. ²

Loneliness is defined as a subjective and negative emotional state where an individual feels distressed about his or her expectations not being met for the quality and need of social (interpersonal) relationships. ³ Loneliness can have negative impacts on sleep, the immune system, physical activity, cognition, and may contribute to the development of Alzheimer's disease. ² Furthermore, loneliness is highly associated with depression, and higher levels of loneliness predict higher suicidal ideations. ⁴ Homecare service users may reduce their level of social participation and interpersonal interactions due to physiological loss, and the interaction generated over time may lead to negative emotions such as loneliness, depression, and bipolarity, which further increases the rate and risk of physiological deterioration and the mortality rate. ⁵ , ⁶ , ⁷ Therefore, reducing loneliness among these homecare service users may help slow down their health deterioration and improve their quality of life. Unfortunately, there is a scarcity of research on the factors influencing loneliness among individuals using homecare services.

An interpersonal interaction is a social interaction between two (or more) people, and it refers to one's perceptions, expectations, and reactions to others. One of the major causes of loneliness is interpersonal deficits. ⁸ Individuals need to interact with others to reduce loneliness and seek happiness. ⁹ Previous studies have associated loneliness with a lack of social relationships and interpersonal interactions. ¹⁰ During the ageing process, it is inevitable for individuals to experience loneliness when the level of social participation and interpersonal interactions decrease due to deteriorating physical functions and changes in social roles. ¹⁰ Therefore, building social relationships and interpersonal interactions is essential to maintain physical and mental well‐being. ¹¹ Compared to other age groups, older people are more vulnerable to social isolation, and their level of social participation (e.g. travel, leisure activities, visits) decreases when they have limited opportunities for interpersonal interactions, resulting in higher levels of loneliness. ¹²

In addition, older people may suffer from limitation in activities of daily living (ADLs) as their physical functions gradually deteriorate, thereby leading to functional impairments. ¹³ , ¹⁴ Many previous studies have identified ADLs as an important factor contributing to loneliness. ¹⁵ , ¹⁶ , ¹⁷ An individual's declining physical and health conditions can result in limitations in ADLs, dependence on others for assistance, and a significant increase in time spent at home, all of which are closely associated with loneliness. ¹⁸ On the other hand, ADLs such as driving, taking a walk, participating in leisure activities, and volunteering are all negatively associated with loneliness, and those who are able to independently complete these activities or with assistance are less prone to loneliness. ¹⁹ These findings suggest that ADLs and human interactions can help reduce loneliness, and receiving visits from family and friends or community care in everyday life can contribute to psychological satisfaction. ¹⁵ However, due to health limitations and lower daily living skills, homecare service users are prone to reduced human interactions resulting in social isolation, which indirectly creates loneliness. This is especially the case among older people. ²⁰

Recently, ageing has gained much attention as a result of the global trends in population ageing. A 16‐year follow‐up study ²¹ conducted among 635 older adults showed a gradual increase in the number of older people who experienced loneliness. Another study ²² in which 469 older people were followed for 28 years found that around one‐third of participants experienced loneliness during the 28 years, and the study found that women are more susceptible to loneliness. The topic of loneliness among older people has received increasing attention during the COVID‐19 outbreak. ²³ , ²⁴ However, most existing literature on loneliness in older adults are studies related to long‐term care facilities or studies comparing older people who live alone and older people in general. ²⁵ , ²⁶ , ²⁷ There has been little research that explores loneliness among older people using homecare services, which is not conducive to lowering the cost of long‐term care and improving the quality of life for homecare service users. Therefore, the purpose of this study was to examine the factors influencing loneliness among older people using homecare services and further explored changes among the elderly homecare service users with different levels of loneliness after 6 months. The results of this study will help increase the awareness for health and welfare units to give high priority to promoting mental health among older people who use homecare services.

METHODS

Study design and participants

This is a longitudinal observational study conducted in communities in Central Taiwan using a structured questionnaire. Data were collected between May 2021 and December 2021. Research assistants visited long‐term homecare service agencies in Central Taiwan and explained the study to their persons‐in‐charge. After obtaining written consents from the agencies, the research assistants visited potential participants or their family members to explain and recruit them into the study. Those interested in participating had to sign a written informed consent before completing the interview questionnaire. Participants were required to complete the questionnaire twice, one pre‐test to obtain baseline information and a post‐test administered at 6 months follow‐up. Both questionnaires were administered via one‐on‐one interviews. Inclusion criteria were as follows: (i) 60 years old or above; (ii) having received long‐term homecare services for over 3 months; (iii) able to communicate in Chinese and/or Taiwanese; and (iv) conscious and able to express personal opinions. Exclusion criteria were those who: (i) had significant cognitive impairments and thus were unable to communicate; (ii) could not understand the questionnaire and respond clearly; (iii) were unable to complete the questionnaire; and (iv) participated in the activities at the daycare centres or long‐term care facilities.

This study was reviewed and approved by the Human Research Ethics Committee of the Jen‐Ai Medical Foundation (No. 110–74). A total of 180 participants were recruited, two of which were excluded because they were unable to fully understand the questionnaire (interviewed by research assistant). Thus, a total of 178 participants signed the written consent form to participate in this study and completed the baseline questionnaire. At the 6 months follow‐up, 16 participants withdrew from the study for the following reasons: eight refused to complete the post‐test due to health reasons, six were unable to respond clearly to the questionnaire, and two were hospitalised. Therefore, only 162 participants completed the post‐test.

Questionnaire

The pre‐ and post‐test questionnaires included five sections. Section 1 consisted of questions related to participant demographic characteristics including gender, age, disability level, marital status, education level, family economic status, religion, number of children, number of individuals in the household, presence of others at dinner, phone or mobile phone use, social media use, perceived economic status and health status, number of days of homecare services received in a week, length of homecare services received, and frequency of interactions with primary caregivers.

Section 2 was the mental health scale. We adopted the Brief Symptom Rating Scale (BSRS‐5) developed by Ming‐Been Lee (2005), which has good reliability and validity in assessing depression and physical and mental symptoms among homecare service users. ²⁸ The BSRS‐5 contains five items on a five‐point Likert scale (0 = not at all, 1 = a little bit, 2 = moderately, 3 = quite a bit, and 4 = extremely), with a higher score indicating a higher degree of emotional distress. Cronbach's alpha of the BSRS‐5 was 0.91 for the pre‐test and 0.89 for the post‐test.

The Interpersonal Interaction Functional Scale developed by Wang et al. ²⁹ was used in Section 3 to evaluate the subjective perception of interpersonal interactions. This scale has good reliability and validity and consists of two subscales, that is, the Interpersonal Interaction Scale (IIS) and the Situational Anxiety Scale (SAS). In this study, only the subscale IIS was used. The IIS consists a total of eight items on a four‐point Likert scale (1 = completely disagree, 2 = slightly disagree, 3 = slightly agree, 4 = completely agree). Higher scores indicate lower interpersonal interaction skills, with a minimum score of eight and a maximum score of 32. Cronbach's alpha of the IIS was 0.90 for the pre‐test and 0.87 for the post‐test.

Section 4 was the Frenchay Activities Index (FAI) used to measure the frequency of the instrumental activities of daily living (IADLs) of homecare service users, and it has good reliability and validity scores. ³⁰ Although items in the FAI are not required for functional independence in daily life, they can reflect a high level of independence and social adaptability. ³¹ The scale consists of 15 items that are divided into three subscales, that is, Domestic chores, Leisure/work, and Outdoor activities. Among them, the items were further divided into everyday activities and seasonal activities. The frequency with which each activity is undertaken is assigned a score of zero to three, with a total score of 45. A lower score indicates less participation in the IADLs. This scoring focuses on the frequency of performing activities rather than the quality of these activities to avoid subjective judgements of the quality of activities. ³⁰ Cronbach's alpha of the FAI was 0.85 for the pre‐test and 0.87 for the post‐test.

In Section 5, the UCLA Loneliness Scale (UCLA, Version 3), developed by Russell, ³² was used to evaluate the level of loneliness among homecare service users. The UCLA was translated into Chinese by Chang and Yang ³³ with good reliability and validity. ³⁴ , ³⁵ The scale consists of 20 items on a four‐point Likert scale (1 = never, 2 = rarely, 3 = sometimes, 4 = always). The scores can range from 20 to 80, with higher scores indicating higher levels of loneliness. Cronbach's alpha of the UCLA was 0.91 for the pre‐test and 0.93 for the post‐test.

Statistical analysis

SPSS 22.0 for Mac (IBM Corp., Armonk, NY, USA) was used for data analysis. First, descriptive statistics were used to present the means and standard deviations of results from participant demographics, BSRS‐5, IIS, FAI, and UCLA. To determine the factors influencing loneliness in elderly homecare service users, a mixed‐effects model analysis was used to screen the demographic data that were significantly associated with that of the UCLA, where UCLA was used as the dependent variable, and each demographic characteristic was added as an independent variable. If P < 0.1, it was included in the next stage of the analysis. Then, the mixed‐effects model analysis was used again to identify factors associated with loneliness among the users (n = 178), and it included characteristics from the above demographic data. IIS and FAI were independent variables. UCLA was the dependent variable. The advantage of using the mixed‐effects model analysis is that participant data can be included despite missing data in repeated measurements. In addition, BSRS‐5 was included in model calibration because it might be a confounding variable to the UCLA. ³⁶

Finally, the median of the UCLA baseline score was used as the cut‐off point to divide the participants into the low loneliness group (scores ≤ the median) and high loneliness group (scores ≥ the median). To further examine the changes among participants in both high and low loneliness groups (n = 162) after 6 months, a paired t‐test was used to analyze if there is a significant difference between the UCLA baseline and post‐test scores.

RESULTS

Participants

The demographic characteristics of participants (baseline) are shown in Table 1. A total of 48 male and 130 female participants (mean age 77.8 years) were included. Among them, a majority were moderately disabled (53.9%), single/divorced/widowed (57.3%), had a high school diploma or lower (83.1%), from low‐income households (62.9%), religious (82.0%), accompanied by others at dinner (64.0%), used a mobile phone (60.7%), did not use social media (66.3%), had a fairly perceived economic status (67.4%), and had a fairly perceived health status (50.6%).

Table 1.

Demographic characteristics of the participants (baseline)

Demographic characteristics	N = 178
Gender
Male	48
Female	130
Age (mean ± SD)	77.79 ± 9.05
Disability level (care receiver)
Mild	62
Moderate	96
Severe	20
Marital status
Single/divorced/widowed	102
Married/cohabiting	76
Years of education
≤12 years	148
>12 years	30
Family economic status
Low‐income households	112
Low‐ and middle‐income households	34
General household	32
Religion
No	32
Yes	146
Number of children (mean ± SD)	2.92 ± 1.47
How many people live with (mean ± SD)	2.28 ± 1.91
Dinner
Eat alone	64
Eat with others	114
Phone or mobile phone use
No	70
Yes	108
Social media use
No	118
Yes	60
Economic status
Poor	36
Fair	120
Good	22
Health status
Poor	70
Fair	90
Good	18
Number of days of homecare service received in a week (mean ± SD)	3.95 ± 1.62
How long (months) to receive homecare services (mean ± SD)	14.34 ± 9.64
Frequency of interactions with primary caregivers
Almost no intersection	20
Infrequently	10
Ordinary	56
Frequently	46
Very frequently	46
BSRS‐5 (mean ± SD)	7.30 ± 4.73

SD, standard deviation; BSRS‐5, Brief Symptom Rating Scale.

In addition, the scores of BSRS‐5, IIS, FAI, and UCLA at baseline and 6 months follow‐up were 7.30 ± 4.73 and 6.93 ± 495; 16.04 ± 5.79 and 15.96 ± 5.64; 7.16 ± 6.9 and 7.62 ± 7.26; and 43.26 ± 11.00 and 44.15 ± 12.05, respectively.

Influencing factors of loneliness in homecare service users

Among the participant demographics, gender, religion, presence of others at dinner, social media use, and perceived economic status were included in the mixed‐effects model analysis. Results are shown in Table 2. The results indicated that gender, presence of others at dinner, social media use, perceived economic status, and IIS score were significantly correlated with the UCLA score (P < 0.05–0.01). In other words, men who ate alone at dinner, did not use social media, had a poorly perceived economic status, and lower interpersonal interaction skills experienced higher levels of loneliness.

Table 2.

Regression coefficients based on the mixed‐effects model analysis for demographic characteristics, IIS and FAI score

Dependent variable	UCLA
Independent variable	Estimate (standard error)	Odds ratio (95% CI)	P‐value
Time (ref: post‐test)	−1.27 (1.15)	−3.55, 1.01	0.27
Gender (ref: female)	4.44 (1.33)	1.81, 7.07	< 0.01
Religion (ref: yes)	2.29 (1.67)	−1.01, 5.58	0.17
Dinner (ref: eat with others)	5.20 (1.35)	2.52, 7.87	< 0.01
Social media use (ref: yes)	3.02 (1.53)	0.00, 6.04	0.05*
Economic status
Poor (ref: good)	7.38 (2.11)	3.22, 11.55	< 0.01
Fair (ref: good)	3.22 (1.77)	−0.28, 6.71	0.07
IIS	0.43 (0.14)	0.16, 0.71	< 0.01
FAI	−0.17 (0.11)	−0.38, 0.04	0.12

Abbreviations: UCLA, UCLA Loneliness Scale; BSRS‐5, Brief Symptom Rating Scale; IIS, Interpersonal Interaction Scale; FAI, Frenchay Activities Index; adjusted for BSRS‐5.

^*: P < 0.05.

Changes in loneliness after 6 months

The changes in UCLA scores between baseline (medium = 43) and 6 months follow‐up are shown in Table 3. The score in the low loneliness group increased and was close to statistical significance (P = 0.09). In the high loneliness group, there was a slight decrease of the score, but the decrease was not significant (P = 0.46). The impact of gender was further analyzed, and the results revealed a significant increase in the UCLA score among male participants in the low loneliness group (P = 0.01).

Table 3.

Comparison of post‐exercise changes in UCLA Loneliness Scale (UCLA) between low loneliness group and high loneliness group (N = 162)

	Low loneliness group (n = 84)			High loneliness group (n = 78)
Variable	Baseline	6 months follow‐up	P‐value	Baseline	6 months follow‐up	P‐value
UCLA (mean ± SD)
Total	35.26 ± 6.24	37.55 ± 10.07	0.09 †	52.44 ± 6.46	51.26 ± 9.81	0.46 †
Male	38.27 ± 6.36	44.27 ± 8.40	0.01* , ‡	51.85 ± 6.00	48.92 ± 11.21	0.23 ‡
Female	34.19 ± 5.93	35.16 ± 9.62	0.68 ‡	52.73 ± 6.78	52.42 ± 9.04	0.95 ‡

Abbreviation: SD, standard deviation.

^* P < 0.05.

^† Paired t‐test.

^‡ Wilcoxon sign rank.

DISCUSSION

During the COVID‐19 outbreak, there has been an increased focus on the issues related to loneliness among older adults. ³⁷ , ³⁸ The physical and psychological impacts due to loneliness may be more severe for disabled elderly homecare service users compared to the general elderly population. ³⁹ This study is one of the few studies that investigated the loneliness of older people using homecare services during the outbreak. Study results showed that the participants experienced higher levels of loneliness compared to the general elderly population ³³ but lower levels of loneliness compared to older people in long‐term care facilities, ⁴⁰ which may be related to social and interpersonal interactions with the outside world. These findings suggested that loneliness has an unignorable impact on older people with disabilities. Furthermore, the IIS scores of participants showed an average to low level of interpersonal interaction frequency, which is consistent with previous studies. ⁴¹ This finding suggested that the reduced frequency of interpersonal interactions may be due to functional limitations. Therefore, maintaining the social support networks of homecare service users so they can keep positive interactions and connections with others may help reduce loneliness. Further, the FAI scores indicated that participants have very poor IADLs, and we speculate that the frequency of activity engagement of the participants was reduced due to disabilities. In Asian cultures, individuals with disabilities tend to rely on family and friends for assistance, which may in turn increase the risk of further health deterioration. ⁴² Results of the mixed‐effects model analysis indicated that variables such as gender, presence of others at dinner, social media use, perceived economic status, and interpersonal interaction skills were significant factors contributing to loneliness among older people using homecare services. This result indicated that interpersonal interactions and social support are important influencing factors of loneliness.

Many studies have shown a significant difference in loneliness between genders, with women having higher levels of loneliness. ⁴¹ , ⁴³ , ⁴⁴ This may be due to the inequalities in socioeconomic status, power, and political rights between men and women; ⁴⁵ however, some studies have indicated a higher level of loneliness in men rather than women. ⁴⁶ , ⁴⁷ , ⁴⁸ Thus, there is still no consensus on the relationship between gender and loneliness. In this study, male participants experienced higher loneliness than female participants, which implied that older men may have higher levels of social isolation when they develop disabilities. This is possibly due to a higher reluctance to reveal their inner emotions, thereby limiting their access to social support. In contrast, women tend to reveal emotional distress and share their emotions with a wider range of people. ⁴⁹ Therefore, more attention should be paid to the mental health of male homecare service users.

In this study, the presence of others at dinner is significantly correlated with loneliness. We postulate that this may have resulted from the influence of the Asian culture where it is important to eat with family (and friends). Because dinner time is during the rest period, it is an appropriate time for frequent interactions with family, which is conducive to reducing loneliness. This is especially the case for home service users with functional limitations who are unable to go out or whose family members are at work during the day. They may even look forward to dinner time when they can interact with family (and friends). On the other hand, similar to previous studies, ⁵⁰ this study found a significant correlation between social media use and loneliness and reported that the use of social media can effectively reduce loneliness among older people. Social media platforms, such as Facebook, not only satisfy the social needs among older people through the exchange of information in the virtual world but also help bring older people closer to their communities, families, and friends. In addition, social media applications serve as a positive way for older people to achieve personal social life and learning as well as life stability, which leads to improved quality of life. ⁵⁰

The results of this study indicated that participants who possessed a poorly perceived economic status had higher levels of loneliness compared to those who were in good financial condition. This result supports the findings of previous studies. ⁵¹ One possible reason for this is that participants with poor economic status are less likely to participate in social activities, which may lead to higher levels of social isolation. Social isolation may contribute to higher loneliness. These results also demonstrated a significant correlation between interpersonal relations and loneliness (see Table 3). Previous studies have shown that homecare service users tend to have higher relationship needs due to their functional impairments. ⁵² Thus, it may be possible to expand the social support network and increase the frequency of interpersonal interactions of older people through ‘companion services’ offered by the homecare service workers. A Japanese study ⁵³ also proposed a solution to reduce loneliness among older people by establishing real interpersonal relationships, including social interactions, support, and companionship through online virtual communities.

Results in Table 3 also demonstrated that the loneliness of participants in the low loneliness group increased after 6 months. This finding is also similar to previous studies, ²² which showed that although elderly homecare service users do not necessarily experience high loneliness at the beginning; the level of loneliness may gradually increase over time. This finding also reminds us not to neglect loneliness among the elderly users (even if they do not experience significant loneliness at the beginning). Furthermore, male participants in the low loneliness group experienced a significant increase in their level of loneliness at the 6 months follow‐up. The feeling of loneliness was found to be one of the risk factors for increasing mortality in older men, ⁵⁴ which suggested that more social support should be provided to male homecare service users. On the other hand, the level of loneliness in the high loneliness group did not significantly change over time; in other words, high levels of loneliness may continue to exist and have a negative impact on physical and mental health over time. ¹¹ However, due to the limitation of the study design (i.e., participants were followed for only 6 months), future research should consider increasing the duration of the follow‐up period.

Limitations

Several limitations should be considered when interpreting the study results. First, participants were all from Central Taiwan, so special consideration should be given when referring to the results of this study. Second, self‐reported scales were used in this study. Although these scales have been widely implemented and have good psychometric properties, bias (e.g., recall bias, social desirability bias) may still occur. Third, the conduct of this study was impacted by COVID‐19 when the life and mood of participants were inevitably influenced by the external environment. In addition, the data collection period coincided with a serious COVID‐19 outbreak in Taiwan, during which the Taiwanese Ministry of Health and Welfare suspended community activities, and closed some public spaces. It increased the relative difficulty of collecting data during the lockdown situation. Despite the above limitations, the results of this study provide an important reference for understanding loneliness among older people using homecare services.

CONCLUSION

The study results indicated that gender, presence of others at dinner, social media use, perceived economic status, and interpersonal interaction skills were significant factors influencing loneliness among older people using homecare services during the COVID‐19 outbreak and that the level of loneliness may increase over time for men. Therefore, increasing adequate interpersonal interactions and social support and reducing social isolation may help reduce loneliness among homecare service users.

FUNDING INFORMATION

This work was funded by grants from the Ministry of Science and Technology, Taiwan (MOST 109‐2314‐B‐468 ‐009 ‐MY2; MOST‐111‐2314‐B‐468‐010‐MY2), and Asia University, Taiwan (ASIA‐109‐CMUH‐08; ASIA‐110‐CMUH‐23).

CONFLICTS OF INTEREST

The authors declare they have no conflict of interest.

ETHICS APPROVAL

This study was reviewed and approved by the Human Research Ethics Committee of the Jen‐Ai Medical Foundation (No. 110–74).

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.