Associations between frailty, sociodemographic characteristics and quality-of-life among community-dwelling older adults: A cross-sectional study
Abstract
Aim
To explore the quality-of-life among community-dwelling older adults in China and to examine the associations between frailty, sociodemographic characteristics and quality-of-life.
Design
A cross-sectional correlational study was adopted.
Methods
Questionnaire study of 311 community-dwelling older adults using the Life Satisfaction Questionnaire and FRAIL scale.
Results
Highest quality-of-life was found for the physical symptoms factor and the lowest for quality of everyday activities/fun. Frailty was associated with total quality-of-life and the physical symptoms and sickness impact factors. For total quality-of-life, the odds of being in the group with a median score or more decreased for frail older people (OR 0.30) versus non-frail and increased for those with medical insurance from employer versus basic (OR 2.30) and those doing exercise ≥30 min 3 days/week or more versus less (OR 2.12). Registered nurses caring for community-dwelling older adults should screen for and prevent frailty and encourage exercise to improve their quality-of-life.
1 INTRODUCTION
Population ageing is a growing phenomenon worldwide (WHO, 2021). In China, approximately 19% of the total population was aged 60 years or older in 2020 (Ministry of Civil Affairs of the People's Republic of China, 2021). Ageing populations also have an increased prevalence of frailty. In turn, frailty is associated with lower quality-of-life (QOL) (Kojima et al., 2016). High QOL is an important indicator of successful ageing (Choi et al., 2017) and an important outcome of nursing practice; thus, it is important to improve understanding of the associations between frailty, sociodemographic factors and QOL among community-dwelling older adults.
2 BACKGROUND
Frailty is characterized by decreased energy reserves and stress resistance and increased vulnerability to adverse outcomes such as falls (Cheng & Chang, 2017), disability (Kojima, 2017) and long-term care (Kojima, 2018), all of which are related to lower QOL among community-dwelling older adults (Schoene et al., 2019; Yeung & Breheny, 2021). QOL is a multidimensional construct that includes dimensions such as physical and psychological health, social relationships, and environmental situations (WHO, 2021b). Prefrail and frail older adults have reported lower QOL than those without frailty in the QOL dimensions of physical and social function, mental health, vitality, increased bodily pain, and role limitations (see e.g. Li et al., 2020; Zhang et al., 2019b). Furthermore, studies using the World Health Organization QOL (WHOQOL) instrument among community-dwelling older adults have identified associations between frailty and lower QOL values in the QOL dimensions of environmental situations (including financial resources, home environment, traffic situation and healthcare services) and social relationships (Gobbens & van Assen, 2017; Renne & Gobbens, 2018; Vanleerberghe et al., 2019). Thus far, studies have mostly explored the associations between frailty and overall QOL, or between frailty and the QOL dimensions of physical, social and mental health. Little research has been conducted about QOL dimensions, such as “quality of activities,” – that is, not just being able to take part in activities (physical ability) or participation but also its quality – and frailty. In addition, there is a need for studies that include frailty in combination with several sociodemographic factors when studying its association with QOL.
The associations between sociodemographic characteristics and QOL of community-dwelling older adults found in research include age (e.g. Karmakar et al., 2018; Li et al., 2020; Renne & Gobbens, 2018; Zhang et al., 2019b), sex (e.g. Gobbens & van Assen, 2017; Karmakar et al., 2018; Vanleerberghe et al., 2019; Zhang et al., 2019b), marital status (Gobbens & van Assen, 2017), economic situation (Karmakar et al., 2018; Gobbens & van Assen, 2017), educational level (Karmakar et al., 2018; Zhang et al., 2019b), occupation (Hammersley et al., 2021; Vinnikov et al., 2021) and multimorbidity (Renne & Gobbens, 2018; Zhang et al., 2019b). In addition, body mass index (BMI) was found to be negatively associated with health-related QOL among community-dwelling older adults (Li et al., 2020), whereas exercise improved QOL among the pre-frail (Chittrakul et al., 2020). However, earlier research has been inconsistent, and associations, whether positive or negative, have differed depending on which QOL dimensions were studied (e.g. age, sex, marital status and educational level). Sociodemographic variables, such as sex and age, are often included, whereas others are less studied, such as previous occupation, educational level and economic situation. When combined with frailty, influencing factors, such as multimorbidity, BMI and exercise, are of importance.
Not only are QOL dimensions, such as the “quality of activities,” less studied, but to some degree also the quality of different social relationships. Spending time doing fun and meaningful activities has been reported as an important QOL dimension for community-dwelling older adults, in addition to perceptions of physical and mental health, having close friendships, and living in a secure home and pleasant neighbourhood (Vanleerberghe et al., 2019). Thus, this study aims to determine the QOL status among community-dwelling older adults and to identify the associations between frailty, sociodemographic characteristics and QOL using an instrument including QOL dimensions, such as quality of activities, relationships with friends and family relations.
3 METHOD
3.1 Setting and sample
This study involved a convenience sample of 311 older adults from three communities in two districts of one city in Zhejiang province, China. The inclusion criteria were age ≧60 years, living in the community for ≥6 months and the ability to communicate and provide information. To estimate the sample sizes, Polit and Beck's (2017) recommendation was followed; for multiple regression analysis, the sample size should be 50 + 8 times the number of predictors. With 13 possible predictors and a response rate of 80%, there was a need for 193 participants. Questionnaires were administered during the yearly regular physical examinations of community residents aged ≥60 years. Of 350 questionnaires, 340 were returned (97% response rate), and 311 participants (89%) responded to the QOL instrument.
3.2 Data collection
Individuals who met the inclusion criteria were invited to participate in the study when visiting their primary care centre for their yearly regular physical examination from June to August 2019. Well-trained nursing students and healthcare staff worked together to screen for eligible participants. Questionnaires were placed into blank envelopes, distributed to eligible participants and completed alone or with the help of students and healthcare staff.
3.3 Instruments
The Life Satisfaction Questionnaire (LSQ) was used to assess QOL, which covers seven factors: physical symptoms (7 items), sickness impact (6 items), quality of daily activities/fun (3 items), quality of daily activities/meaningful (4 items), socioeconomic situation (3 items), quality of family relations (5 items) and quality of close friend relationships (5 items) (Carlsson & Hamrin, 2002). The responses were graded on a 7-point Likert scale. For factor scores, items are summarized, divided by the highest possible factor score and multiplied by 100. Possible scores range from 14 to 100, with higher scores indicating better QOL. The LSQ has been used with older adults in Sweden (Roos et al., 2016; Sjölund et al., 2021) and the LSQ-Chinese version showed acceptable validity and good internal consistency when used with older adults. Cronbach's alpha values were 0.77–0.86 and 0.93 for the factors and total scale, respectively (Lou et al., 2022). For frailty status, the FRAIL scale (Morley et al., 2012) was used, which includes five items: fatigue, resistance, ambulation, illnesses and weight loss. Participants with one or two symptoms/signs were identified as prefrail, and those with at least three were identified as frail. The FRAIL scale has been reported to have strong predictive validity for health, functioning and mortality.
The following sociodemographic variables were collected: age, sex, marital status, living situation and monthly income (Table 1). Age was categorized into three groups: 60–69 years, 70–79 years and ≥80 years old. Previous occupations (the years before retirement) were also collected and categorized into none (including if they do any handicraft and sell at home or in a small, nearby market), farmers and retired workers from the government or a company. Participants' medical insurance was also collected and categorized into basic medical insurance and medical insurance from employers. In China, basic medical insurance covers all people except for those who already have medical insurance from employers (National Healthcare Security Administration, 2021). Participants were also asked if they had any disease(s); 14 diseases were listed, and they could also add a disease if it was not included. Multimorbidity is defined as the presence of two or more concurrent diseases (Johnston et al., 2019). Exercise was assessed with one item regarding the days of doing moderate- or vigorous-intensity exercise for more than 30 min a week (described examples included walking, climbing, cycling, taiji or square dance). According to the guidelines of activities for national residents in China (General Administration of Sport of China, 2017), exercise was categorized into two groups: 0–2 days/week and 3–7 days/week. For BMI, weight and length were measured at the primary care center during the yearly regular physical examination.
| Variables | Mean (SD) |
|---|---|
| Age | 71.3 (6.9) |
| Variables | N (%) |
| Age (years) | |
| ≧80 | 50 (16.1) |
| 70–79 | 126 (40.5) |
| 60–69 | 135 (43.4) |
| Sex | |
| Male | 164 (52.7) |
| Female | 147 (47.3) |
| Marital status | |
| Others (without a partner) | 60 (19.3) |
| Married | 251 (80.7) |
| Living situation | |
| Living alone | 26 (8.4) |
| Living with families | 285 (91.6) |
| Monthly income (yuan, 7 missing data) | |
| <2000 | 112 (36.8) |
| 2,000 ~ 5,000 | 142 (46.7) |
| >5,000 | 50 (16.4) |
| Education | |
| Primary school or lower | 197 (63.3) |
| Middle school or higher | 114 (36.7) |
| Previous occupation types | |
| None | 97 (31.2) |
| Farmers | 106 (34.1) |
| Workers in department or company | 108 (34.7) |
| Medical insurance types | |
| Basic medical insurance for residents | 270 (86.8) |
| Employment medical insurance | 41 (13.2) |
| Multimorbidity | |
| No | 242 (77.8) |
| Yes | 69 (22.2) |
| Exercise more than 30 min | |
| 0–2 days/week | 121 (38.9) |
| 3–7 days/week | 190 (61.1) |
| Frailty status | |
| Nonfrail | 225 (72.3) |
| Prefrail | 64 (20.6) |
| Frail | 22 (7.1) |
- Abbreviations: BMI Body, Mass Index; SD, Standard Deviation.
3.4 Ethical considerations
This study was approved by the Medical Ethical Committee of NN University (No. blinded for review). Data collection was also approved by the health center leaders. All participants received oral and written study information and signed an informed consent form. Small gifts were offered to participants to stimulate participation and to thank them for their participation.
3.5 Data analysis
Data analysis was conducted using IBM SPSS (version 26). Missing data for LSQ items were substituted with the individual's median for that factor (a total of 12 missing data pieces). Due to the data distribution and variable types, the Mann–Whitney U-test, Kruskal–Wallis test and Spearman's correlation coefficient (rs) were used to determine associations between sociodemographic characteristics, frailty, total LSQ score and the factors. Possible predictors that met the significance level of p < .10 in the bivariate analyses were entered into binary logistic regression models based on the 25th, 50th and 75th percentiles of the total LSQ score and the factor scores. Thus, three models were tested for each QOL outcome (DeCoster et al., 2011).
4 RESULTS
4.1 Participant characteristics and QOL
The mean age of the participants was 71 years (range, 60–94 years), and 52.7% were male. Most participants were married (80.7%) and lived with their families (91.6%), which, in Chinese households, mostly represents living together with their partner and/or one of their children's families. The participants were 7.1% frail, 20.6% prefrail and 72.3% nonfrail (Table 1).
The highest (best) QOL levels were found for the factors of physical symptoms (mean 81.4, SD 13.4) and sickness impact (mean 71.8, SD 12.0), whereas the lowest scores were found for the factors of quality of everyday activities/fun (mean 49.8, SD 20.0) and quality of friend relationships (mean 58.4, SD 14.8) (Table 2). Related-sample Friedman two-way analyses indicated that there were statistically significant differences between all factors except for pairwise comparisons between the quality of everyday activities/meaningful and quality of friend relationships, quality of family relations and socioeconomic situation.
| Variables | Mean (SD) | α | (1) | (2) | (3) | (4) | (5) | (6) | (7) | (8) | (9) |
|---|---|---|---|---|---|---|---|---|---|---|---|
| BMI (1) | 24.2 (4.0) | – | |||||||||
| Physical symptoms (2) | 81.4 (13.4) | 0.87 | −.12* | – | |||||||
| Sickness impact (3) | 71.8 (12.0) | 0.73 | −.15* | .65** | – | ||||||
| Quality of everyday activities fun (4) | 49.8 (20.0) | 0.93 | .04 | −.01 | .08 | – | |||||
| Quality of everyday activities meaningful (5) | 62.0 (14.6) | 0.88 | −.02 | .40** | .33** | .42** | – | ||||
| Socioeconomic situation (6) | 66.5 (14.5) | 0.94 | −.04 | .32** | .26** | .28** | .58** | – | |||
| Quality of family relation (7) | 65.2 (12.9) | 0.63 | .01 | .01 | .06 | .16** | .06 | .10 | – | ||
| Quality of close friend relationship (8) | 58.4 (14.8) | 0.75 | −.02 | −.10 | .01 | .03 | −.28** | −.25** | .36** | – | |
| Total QOL (9) | 67.2 (7.9) | 0.85 | −.10a | .75** | .74** | .39** | .58** | .50** | .33** | .17* | – |
- Note: rs values from Spearman correlation analysis, α Cronbach's alpha.
- Abbreviations: BMI Body Mass Index, QOL Quality-Of-Life, SD Standard Deviation. *p < .05, **p < .001.
- a p < 0.10.
The results indicated statistically significant differences in QOL related to participant age (two QOL factors and total QOL score), monthly income (three QOL factors and total QOL score), educational level (two QOL factors), previous occupational status (four QOL factors), medical insurance type (three QOL factors and total QOL score), multimorbidity (three QOL factors), performance of physical exercise (two QOL factors and total QOL score) and frailty status (two QOL factors and total QOL score) (Table 3). In addition, BMI was significantly associated with two QOL factors (Table 2).
| Variables | Physical symptoms Mean (SD) | Sickness impact Mean (SD) | Quality of everyday activities fun Mean (SD) | Quality of everyday activities meaningful Mean (SD) | Socio-economic situation Mean (SD) | Quality of family relation Mean (SD) | Quality of close friend relationship Mean (SD) | Total quality of life Mean (SD) |
|---|---|---|---|---|---|---|---|---|
| Age | ||||||||
| ≧80① | 75.7 (14.5) | 67.0 (12.6) | 47.0 (17.8) | 57.8 (15.0) | 63.7 (13.9) | 66.0 (10.1) | 60.3 (13.8) | 64.4 (8.7) |
| 70–79② | 83.6 (12.7) | 72.4 (12.1) | 48.2 (21.6) | 62.6 (15.1) | 67.7 (15.1) | 64.8 (13.5) | 56.1 (15.4) | 67.3 (7.3) |
| 60–69③ | 81.6 (13.1) | 73.0 (11.4) | 52.3 (19.0) | 63.1 (13.8) | 66.4 (14.1) | 65.3 (13.4) | 59.7 (14.3) | 68.0 (8.0) |
| Kruskal-Wallis (p-value) | .004 | .013 | .055 | .059 | .335 | .755 | .096 | .030 |
| Pairwise comparisons (p-value adjusted by Bonferroni correction) | ② > ① (.003) ③ > ① (.030) | ② > ① (.015) ③ > ① (.025) | ② > ① (.044) ③ > ① (.041) | |||||
| Sex | ||||||||
| Male | 81.8 (13.5) | 72.5 (12.2) | 48.7 (19.1) | 60.8 (14.2) | 66.0 (14.2) | 64.7 (13.0) | 58.4 (14.5) | 67.0 (7.5) |
| Female | 81.0 (13.3) | 71.1 (11.8) | 51.1 (20.9) | 63.4 (15.0) | 67.0 (14.9) | 65.8 (12.9) | 58.4 (15.1) | 67.3 (8.4) |
| Mann–Whitney U (p-value) | .550 | .113 | .305 | .230 | .802 | .714 | .793 | .556 |
| Marital status | ||||||||
| Others (without partner) | 82.7 (14.1) | 71.23 (13.7) | 50.6 (25.6) | 63.0 (19.3) | 66.4 (18.2) | 67.2 (16.9) | 56.3 (20.0) | 67.5 (9.6) |
| Married | 81.2 (13.2) | 71.9 (11.7) | 49.4 (18.4) | 61.8 (13.3) | 66.5 (13.6) | 64.8 (11.8) | 58.6 (13.2) | 67.1 (7.5) |
| Mann–Whitney U (p-value) | .314 | .911 | .673 | .776 | .365 | .310 | .724 | .570 |
| Living situation | ||||||||
| Living alone | 83.1 (15.3) | 68.0 (16.2) | 48.2 (26.2) | 64.0 (20.4) | 66.7 (18.4) | 64.0 (18.8) | 54.4 (17.9) | 66.1 (10.2) |
| Living with families | 81.3 (13.2) | 72.2 (11.6) | 50.0 (19.4) | 61.9 (14.0) | 66.5 (14.2) | 65.4 (12.3) | 58.7 (14.4) | 67.3 (7.7) |
| Mann–Whitney U (p-value) | .369 | .211 | .489 | .499 | .983 | .673 | .402 | .993 |
| Monthly income | ||||||||
| <2,000 yuan① | 80.7 (13.9) | 70.6 (11.9) | 53.9 (20.5) | 63.0 (16.2) | 65.9 (15.5) | 67.3 (13.2) | 59.2 (16.6) | 67.7 (8.5) |
| 2,000 ~ 5,000 yuan② | 80.0 (12.6) | 71.8 (11.7) | 46.4 (18.8) | 59.9 (13.5) | 65.1 (13.9) | 64.1 (12.6) | 57.2 (13.2) | 65.8 (7.1) |
| >5,000 yuan③ | 86.8 (13.2) | 74.1 (13.3) | 51.9 (20.9) | 65.1 (14.1) | 70.7 (13.6) | 64.2 (13.5) | 60.1 (15.5) | 69.8 (8.6) |
| Kruskal–Wallis (p-value) | .003 | .309 | .002 | .021 | .042 | .168 | .135 | .008 |
| Pairwise comparisons (p-value adjusted by Bonferroni correction) | ③ > ② (.002) ③ > ① (.018) | ① > ② (.002) | ③ > ② (.021) | ③ > ② (.009) | ||||
| Educational level | ||||||||
| Primary or lower | 81.2 (13.8) | 71.3 (12.1) | 49.2 (20.4) | 60.8 (15.7) | 65.1 (15.0) | 65.4 (13.2) | 57.1 (15.2) | 66.5 (7.9) |
| Middle school or higher | 81.8 (12.7) | 72.7 (12.0) | 50.9 (19.2) | 64.2 (12.5) | 68.8 (13.4) | 64.9 (12.6) | 60.6 (13.7) | 68.3 (7.9) |
| Mann–Whitney U (p-value) | .789 | .386 | .233 | .003 | .015 | .632 | .491 | .081 |
| Previous occupational status | ||||||||
| No① | 81.6 (12.2) | 72.2 (10.2) | 49.7 (12.0) | 57.2 (11.0) | 60.8 (11.1) | 65.1 (9.2) | 62.2 (9.9) | 66.7 (6.4) |
| Farmer② | 79.7 (13.5) | 70.4 (12.6) | 51.3 (23.7) | 63.4 (17.6) | 67.8 (16.3) | 67.8 (14.8) | 56.3 (18.0) | 67.0 (9.0) |
| Workers in department or company③ | 83.0 (14.2) | 72.9 (12.9) | 48.5 (21.7) | 65.2 (13.3) | 70.3 (14.0) | 62.9 (13.5) | 57.0 (14.4) | 67.7 (8.1) |
| Kruskal–Wallis (p-value) | .129 | .345 | .791 | <.001 | <.001 | .024 | <.001 | .609 |
| Pairwise comparisons (p-value adjusted by Bonferroni correction | ② > ① (.017) ③ > ① (<.001) | ② > ① (.001) ③ > ① (<.001) | ② > ③ (.034) | ① > ② (.025) ① > ③ (<.001) | ||||
| Medical insurance (MI) | ||||||||
| Basic MI | 80.7 (13.6) | 71.1 (11.9) | 49.7 (19.3) | 61.5 (14.7) | 65.8 (14.4) | 65.5 (12.7) | 58.4 (14.8) | 66.8 (7.9) |
| Employment MI | 86.4 (11.2) | 76.4 (12.3) | 50.6 (24.0) | 65.6 (13.9) | 70.7 (15.2) | 63.7 (14.7) | 58.5 (15.0) | 70.0 (8.0) |
| Mann–Whitney U (p-value) | .010 | .012 | .922 | .075 | .041 | .275 | .355 | .028 |
| Multimorbidity | ||||||||
| No | 80.6 (13.5) | 72.1 (12.0) | 50.2 (18.6) | 61.1 (14.2) | 65.6 (13.9) | 65.3 (11.7) | 59.1 (13.6) | 67.0 (7.6) |
| Yes | 84.6 (12.7) | 71.0 (12.2) | 48.5 (24.3) | 65.3 (16.0) | 69.6 (16.4) | 65.0 (16.6) | 55.8 (18.2) | 67.8 (9.1) |
| Mann–Whitney U (p-value) | .015 | .403 | .317 | .048 | .073 | .402 | .005 | .923 |
| Exercise ≥ 30 min | ||||||||
| 0–2 days/week | 79.6 (14.1) | 70.9 (11.5) | 48.7 (20.6) | 59.2 (15.9) | 63.9 (15.0) | 66.1 (12.4) | 59.1 (15.4) | 66.2 (8.4) |
| 3–7 days/week | 82.6 (12.9) | 72.4 (12.4) | 50.5 (19.6) | 63.8 (13.6) | 68.2 (14.0) | 64.7 (13.3) | 57.9 (14.4) | 67.8 (7.5) |
| Mann–Whitney U (p-value) | .071 | .229 | .300 | <.001 | .002 | .240 | .140 | .011 |
| Frailty status | ||||||||
| Nonfrailty① | 82.4 (12.3) | 74.1 (10.7) | 49.5 (19.8) | 62.2 (14.2) | 66.6 (14.4) | 65.3 (13.3) | 59.2 (14.2) | 67.9 (7.7) |
| Prefrailty② | 81.9 (13.7) | 68.4 (11.7) | 50.2 (20.5) | 61.4 (15.9) | 65.6 (14.5) | 65.5 (12.0) | 56.6 (17.0) | 66.3 (7.2) |
| Frailty③ | 70.1 (17.7) | 57.9 (14.7) | 52.4 (21.2) | 62.2 (15.8) | 68.4 (15.9) | 63.9 (11.7) | 55.5 (13.9) | 62.0 (9.7) |
| Kruskal–Wallis (p-value) | .004 | <.001 | .311 | .654 | .890 | .713 | .283 | .002 |
| Pairwise comparisons (p-value adjusted by Bonferroni correction) | ① > ③ (.003) ② > ③ (.011) | ① > ② (.004) ① > ③ (<.001) ② > ③ (.012) | ①>③ (.002) ② > ③ (.042) | |||||
- Note: When using Kruskal–-Wallis test, pairwise post hoc comparisons were adjusted by Bonferroni correction. Bold text indicates statistically significant values.
- Abbreviations: QOL, Quality Of Life; SD, Standard Deviation.
4.2 Logistic regression analyses of frailty, sociodemographic variables and QOL
The results were statistically significant for all models on the omnibus test (i.e. the outcome variables were split at the 25th, 50th and 75th percentiles of QOL) for all QOL factors except two. Results for the physical symptoms factor and the total LSQ were significant in Models 1 and 2 (M1, M2), and the quality of family relations was significant in M1. The Nagelkerke R2 of the significant models ranged from 0.058 (quality of close friend relationships) to 0.222 (quality of everyday activities/meaningful) (Table 4).
| Physical symptoms as dependent variable (higher scores better QOL) | ||||||
|---|---|---|---|---|---|---|
| Included independent variables in the models | Model 1 | Model 2 | Model 3 | |||
| OR (95%CI) | p-values | OR (95%CI) | p-values | OR (95%CI) | p-values | |
| Age (60–69) | 2.91 (1.35–6.24) | 0.006 | 1.63 (0.79–3.39) | 0.188 | 4.78 (1.57–14.59) | 0.006 |
| Age (70–79) | 2.71 (1.27–5.77) | 0.010 | 1.47 (0.71–3.05) | 0.295 | 3.34 (1.08–10.34) | 0.036 |
| BMI | 0.95 (0.88–1.01) | 0.113 | 0.94 (0.88–1.00) | 0.034 | 0.97 (0.90–1.04) | 0.410 |
| Medical insurance (MI) (Employment MI) | 1.77 (0.66–4.70) | 0.253 | 2.25 (1.02–4.95) | 0.044 | 1.12 (0.47–2.66) | 0.793 |
| Multimorbidity (yes) | 2.64 (1.19–5.84) | 0.017 | 1.89 (1.03–3.45) | 0.039 | 1.18 (0.60–2.29) | 0.633 |
| Exercise ≥30 min (3–7 days/week) | 1.48 (0.86–2.57) | 0.158 | 1.51 (0.92–2.47) | 0.103 | 1.10 (0.61–1.97) | 0.751 |
| Prefrail | 1.47 (0.70–3.07) | 0.306 | 0.72 (0.40–1.33) | 0.300 | 1.44 (0.72–2.90) | 0.299 |
| Frail | 0.34 (0.13–0.87) | 0.025 | 0.31 (0.11–0.90) | 0.031 | 0.93 (0.29–2.97) | 0.898 |
| Omnibus test (p-values) | <.001 | <.001 | .141 | |||
| Nagelkerke R square | 0.143 | 0.125 | 0.061 | |||
| Hosmer and Lemeshow Test (p-values) | .310 | .435 | .965 | |||
| Sickness impact as dependent variable (higher scores better QOL) | ||||||
| Age (60–69) | 2.45 (1.15–5.19) | 0.020 | 2.41 (1.15–5.03) | 0.019 | 2.71 (1.11–6.62) | 0.029 |
| Age (70–79) | 2.54 (1.18–5.44) | 0.017 | 2.17 (1.04–4.53) | 0.039 | 2.60 (1.07–6.34) | 0.036 |
| BMI | 0.96 (0.90–1.03) | 0.258 | 0.93 (0.87–0.99) | 0.017 | 0.94 (0.88–1.00) | 0.054 |
| Medical insurance (MI) (Employment MI) | 1.28 (0.53–3.07) | 0.586 | 2.02 (0.95–4.29) | 0.068 | 2.84 (1.37–5.86) | 0.005 |
| Prefrail | 0.57 (0.30–1.09) | 0.091 | 0.56 (0.30–1.02) | 0.059 | 0.79 (0.40–1.55) | 0.490 |
| Frail | 0.08 (0.03–0.24) | <0.001 | 0.22 (0.07–0.67) | 0.008 | 0.26 (0.06–1.18) | 0.082 |
| Omnibus test (p-values) | <.001 | <.001 | <.001 | |||
| Nagelkerke R square | 0.179 | 0.132 | 0.113 | |||
| Hosmer and Lemeshow Test (p-values) | .897 | .916 | .675 | |||
| Quality of everyday activities fun as dependent variable | ||||||
| Age (60–69) | 0.74 (0.37–1.46) | 0.384 | 1.10 (0.55–2.20) | 0.793 | 1.70 (0.76–3.83) | 0.197 |
| Age (70–79) | 1.47 (0.74–2.94) | 0.277 | 2.26 (1.13–4.51) | 0.021 | 2.83 (1.27–6.30) | 0.011 |
| Monthly income (2,000–5,000 yuan) | 0.35 (0.20–0.60) | <0.001 | 0.38 (0.22–0.64) | <0.001 | 0.36 (0.21–0.64) | <0.001 |
| Monthly income (>5,000 yuan) | 0.58 (0.28–1.17) | 0.129 | 0.99 (0.50–1.96) | 0.972 | 1.15 (0.58–2.28) | 0.688 |
| Omnibus test (p-values) | .001 | <.001 | <.001 | |||
| Nagelkerke R square | 0.086 | 0.098 | 0.101 | |||
| Hosmer and Lemeshow Test (p-values) | .378 | .908 | .495 | |||
| Quality of everyday activities meaningful as dependent variable | ||||||
| Age (60–69) | 2.95 (1.32–6.61) | 0.009 | 1.60 (0.77–3.30) | 0.207 | 1.74 (0.74–4.09) | 0.200 |
| Age (70–79) | 4.64 (2.03–10.61) | <0.001 | 2.70 (1.30–5.65) | 0.008 | 1.51 (0.65–3.54) | 0.340 |
| Education (Middle school or higher) | 1.92 (0.87–4.25) | 0.107 | 1.94 (1.02–3.69) | 0.045 | 0.97 (0.52–1.82) | 0.917 |
| Previous occupational status (Farmer) | 1.61 (0.79–3.26) | 0.189 | 1.73 (0.93–3.20) | 0.083 | 3.14 (1.47–6.72) | 0.003 |
| Previous occupational status (Workers in department or company) | 2.40 (1.01–5.67) | 0.046 | 2.31 (1.13–4.74) | 0.022 | 3.51 (1.56–7.89) | 0.002 |
| Medical insurance (MI) (Employment MI) | 1.20 (0.36–4.03) | 0.765 | 0.63 (0.25–1.55) | 0.313 | 1.03 (0.45–2.35) | 0.952 |
| Multimorbidity (yes) | 1.55 (0.66–3.68) | 0.317 | 1.48 (0.75–2.91) | 0.262 | 1.35 (0.72–2.52) | 0.348 |
| Exercise ≥30 min (3–7 days/week) | 3.35 (1.78–6.31) | <0.001 | 2.72 (1.59–4.62) | <0.001 | 1.43 (0.80–2.55) | 0.231 |
| Omnibus test (p-values) | <.001 | <.001 | .010 | |||
| Nagelkerke R square | 0.222 | 0.177 | 0.093 | |||
| Hosmer and Lemeshow Test (p-values) | .669 | .413 | .958 | |||
| Socio-economic situation as dependent variable | ||||||
| Education (Middle school or higher) | 2.55 (1.05–6.20) | 0.039 | 1.24 (0.72–2.16) | 0.441 | 0.94 (0.47–1.88) | 0.859 |
| Previous occupational status (Farmer) | 1.63 (0.78–3.40) | 0.194 | 2.78 (1.50–5.18) | 0.001 | 2.94 (1.18–7.36) | 0.021 |
| Previous occupational status (Workers in department or company) | 1.68 (0.70–4.01) | 0.246 | 3.77 (1.93–7.36) | <0.001 | 5.14 (2.02–13.08) | 0.001 |
| Medical insurance (MI) (Employment MI) | 0.84 (0.25–2.79) | 0.776 | 1.01 (0.48–2.16) | 0.971 | 0.94 (0.40–2.26) | 0.899 |
| Multimorbidity (yes) | 1.32 (0.56–3.08) | 0.528 | 1.21 (0.68–2.15) | 0.507 | 1.90 (0.99–3.65) | 0.053 |
| Exercise ≥30 min (3–7 days/week) | 2.69 (1.40–5.19) | 0.003 | 1.30 (0.79–2.16) | 0.301 | 1.48 (0.78–2.83) | 0.234 |
| Omnibus test (p-values) | .001 | <.001 | <.001 | |||
| Nagelkerke R square | 0.127 | 0.113 | 0.120 | |||
| Hosmer and Lemeshow Test (p-values) | .423 | .867 | .430 | |||
| Quality of family relation as dependent variable | ||||||
| Previous occupational status (Farmer) | 1.11 (0.49–2.52) | 0.809 | 1.23 (0.70–2.13) | 0.470 | 1.60 (0.78–3.27) | 0.197 |
| Previous occupational status (Workers in department or company) | 0.37 (0.18–0.75) | 0.006 | 0.84 (0.49–1.46) | 0.537 | 0.81 (0.37–1.79) | 0.608 |
| Omnibus test (p-values) | .002z | .388 | .156 | |||
| Nagelkerke R square | 0.065 | 0.008 | 0.020 | |||
| Hosmer and Lemeshow Test (p-values) | 1.000 | 1.000 | 1.000 | |||
| Quality of close friend relationship as dependent variable | ||||||
| Age (60–69) | 0.52 (0.23–1.19) | 0.122 | 0.43 (0.21–0.86) | 0.018 | 0.78 (0.37–1.65) | 0.518 |
| Age (70–79) | 0.53 (0.24–1.20) | 0.129 | 0.61 (0.31–1.21) | 0.155 | 0.98 (0.48–2.03) | 0.967 |
| Previous occupational status (Farmer) | 0.65 (0.33–1.29) | 0.218 | 0.40 (0.22–0.71) | 0.002 | 0.47 (0.26–0.86) | 0.014 |
| Previous occupational status (Workers in department or company) | 0.36 (0.18–0.68) | 0.002 | 0.27 (0.15–0.49) | <0.001 | 0.43 (0.23–0.80) | 0.008 |
| Multimorbidity (yes) | 0.50 (0.28–0.90) | 0.020 | 0.54 (0.29–1.01) | 0.053 | 0.66 (0.34–1.28) | 0.221 |
| Omnibus test (p-values) | .001 | <.001 | .026 | |||
| Nagelkerke R square | .095 | .137 | .058 | |||
| Hosmer and Lemeshow Test (p-values) | .905 | .627 | .828 | |||
| Total of QOL as dependent variable | ||||||
| Age (60–69) | 1.86 (0.88–3.94) | 0.106 | 2.00 (0.97–4.15) | 0.061 | 2.44 (0.98–6.07) | 0.055 |
| Age (70–79) | 2.02 (0.95–4.30) | 0.069 | 1.60 (0.77–3.30) | 0.207 | 1.95 (0.78–4.89) | 0.154 |
| BMI | 0.96 (0.90–1.03) | 0.252 | 0.95 (0.89–1.01) | 0.093 | 0.99 (0.92–1.06) | 0.683 |
| Education (Middle school or higher) | 1.59 (0.86–2.97) | 0.142 | 0.98 (0.57–1.67) | 0.933 | 0.88 (0.48–1.63) | 0.690 |
| Medical insurance (MI) (Employment MI) | 1.53 (0.57–4.13) | 0.401 | 2.30 (1.03–5.13) | 0.041 | 1.67 (0.73–3.80) | 0.225 |
| Exercise ≥30 min (3–7 days/week) | 1.61 (0.93–2.80) | 0.091 | 2.12 (1.27–3.52) | 0.004 | 1.10 (0.62–1.97) | 0.741 |
| Prefrail | 1.17 (0.59–2.32) | 0.662 | 0.87 (0.48–1.59) | 0.648 | 1.12 (0.57–2.23) | 0.736 |
| Frail | 0.30 (0.12–0.76) | 0.011 | 0.30 (0.10–0.88) | 0.028 | 0.79 (0.25–2.50) | 0.691 |
| Omnibus test (p-values) | 0.004 | <0.001 | 0.635 | |||
| Nagelkerke R square | 0.103 | 0.125 | 0.030 | |||
| Hosmer and Lemeshow Test (p-values) | 0.996 | 0.614 | 0.739 | |||
- Note: Variables that in the comparative analyses had p-values less than .10 (Tables 2, 3) were included in the logistic regression. Monthly income was not included as it measures a bit the same as we capture with educational status, occupational status, and medical insurance type when it was significant together with theses variables. Bold text indicates statistically significant values.
- Abbreviations: BMI, Body Mass Index; CI, Confidence Interval;. OR, Odds Ratio; QOL, Quality-Of-Life.
For the QOL factor of physical symptoms in M1, frailty (versus nonfrail) decreased the odds of being in the group with better QOL, whereas multimorbidity and younger age (60–69 years; 70–79 years versus ≥80 years) increased the odds for being in the group with better values for physical symptoms (split at the 25th percentile). For M2, the outcome split at the median, significant variables were frailty, BMI, multimorbidity and medical insurance from employer. The results of Model 3 (M3) were non-significant. The significant variables for sickness impact were age and frailty in M1. In M2, age and frailty remained significant, and BMI was added. In M3, age and medical insurance from employer were related to sickness impact (Table 4 for Odds Ratio).
Regarding the factor of quality of everyday activities/fun, monthly income (2,000–5,000 versus <2,000 yuan) decreased the odds of being in the higher QOL group in M1, M2 and M3, whereas age (70–79 versus ≥80 years) increased the odds in M2 and M3. The significant variables for quality of everyday activities/meaningful in M1 were age, occupation – department or company and exercise. In M2, age, occupation department or company, and exercise remained significant, while educational level was added. In M3, both occupation – department or company and farmers were significant. The significant variables for socioeconomic situation were educational level and exercise in M1, and occupation – farmer and department or company in both M2 and M3. For quality of family relations in M1, occupation – department or company decreased the odds of being in the higher QOL group. The results for M2 and M3 were non-significant. For quality of close friend relations in M1, the significant variables were occupation - department or company and multimorbidity. In M2, occupation – department or company remained significant, and age (60–69 years) and occupation - farmer were added. In M3, occupation–farmer and department or company remained significant. Regarding total LSQ score, the significant variables were frailty in M1 and medical insurance from employers, exercise, and frailty in M2, while the M3 results were non-significant.
5 DISCUSSION
The scores were highest (i.e. better QOL) for the QOL factors of physical symptoms and sickness impact, and lowest for the factors of quality of everyday activities/fun and quality of close friend relationships. Logistic regression analyses indicated that age, BMI, educational level, previous occupation status, medical insurance type, multimorbidity, exercise and frailty were associated with different QOL factors in different directions. Nagelkerke R2 values for the models ranged from .058 to .222. Regarding total QOL and QOL factor scores, the models were mostly non-significant or had lower Nagelkerke R2 when trying to predict the participants at the 75th percentile or higher, except for socioeconomic situation. However, for nurses, it is most important to support those with worse QOL, that is, to be aware of changeable factors that could improve their QOL, such as preventing high BMI, frailty and promoting exercise activities.
Similar to our results, Abolhassani et al. (2019) reported the highest scores for the factors of health, mobility and feeling of safety and the lowest scores for social and cultural life. In contrast, a study from India found the highest scores for the social relations factor (Karmakar et al., 2018). However, the social functioning and relationship factors in these studies did not account for aspects such as joy and meaning. Takashima et al. (2020) reported that, for older people, social activities meant a fulfilling social life, maintaining stable family relationships and maintaining safety and peace in the community. In China, the rate of participation in social activities among community-dwelling older adults was reported as very low, and the activity types were limited to caring for other older adults or grandchildren, community cleaning or neighbourhood patrols (Xun, 2021). This presents a possible reason for the low levels indicated for the factors of quality of everyday activities/fun and quality of close friend relations reported by participants in our study, while values for quality of everyday activities/meaningful and quality of family relations were somewhat higher but still quite low. Therefore, community nurses should pay more attention to asking older people about their interests and offering a variety of everyday community activities, including activities where older people meet and interact with others to improve their overall QOL.
The logistic regression analysis indicated that frailty had a significant negative association with total QOL score and with the QOL factors of physical symptoms and sickness impact when other sociodemographic data were held constant. The association was significant in two of the three models when the outcome variables were split between the 25th percentile and the median. QOL is a multidimensional concept, and previous studies (Li et al., 2020; Zhang et al., 2019b) have found that frailty status is associated with QOL dimensions such as physical health, mental health and social functioning. Li et al. (2020) reported that fatigue, low resistance, slow movement, illness and loss of weight all contributed to physical frailty, which was independently associated with the physical domain of QOL, meaning that it was similar to our sickness impact factor. Poor knowledge and understanding of frailty among Chinese primary care professionals and community-dwelling older adults have been reported (Coker et al., 2019), and Obbia et al. (2020) indicated that competencies required by primary care professionals, such as registered nurses and physicians, to assess and prevent frailty were not included in their training. Therefore, older adults with frailty in China may not have contact with or help from primary care professionals. Our findings indicate that there is a need to screen for and prevent frailty among community-dwelling older adults and provide nurses working in the community and older adults themselves with knowledge about frailty to promote successful and healthy ageing. Our results also indicate that having two or more diseases (i.e. multimorbidity) increases the odds of having better QOL in the physical symptoms dimension when other variables were held constant, and in the bivariate analysis. This suggests that disease diagnoses indicate regular contact with primary care, whereas frailty among older adults does not indicate regular contact. Sand et al. (2021) found that patients with multimorbidity received support from their families and general practitioners to treat and manage their chronic illnesses, which in turn helped them maintain their everyday activities. However, other studies have indicated that multimorbidity is negatively associated with sensory abilities (WHOQOL-OLD) (Renne & Gobbens, 2018) and physical function (Gu et al., 2018).
Similar to our results on multimorbidity and the quality of close friend relationships, other studies (Gobbens & van Assen, 2017; Renne & Gobbens, 2018) have found that older people with multimorbidity reported lower levels of social participation. This may be explained by the findings of Duguay et al. (2014) who found that participants with multimorbidity preferred to do things alone rather than joining social activities due to their health status. Our findings about BMI and the QOL factors of physical symptoms and sickness impact were consistent with the results of Öztürk et al. (2018), suggesting that nurses should work to prevent high BMI. Our findings also indicate that exercise was positively associated with the total QOL score and the quality of daily activities/meaningfulness, which is also consistent with prior results (Chittrakul et al., 2020). Therefore, improvements in community exercise activities (e.g. easier access and social aspects) and their active promotion by nurses appear important. Our study indicated that participants with medical insurance from their employer had higher odds of having a better total QOL and better scores for physical symptoms and sickness impact than those with basic medical insurance. The main types of medical insurance in China are medical insurance from employers and basic medical insurance (National Healthcare Security Administration, 2021); the latter has some restrictions about the amount or types of medical services and purchasing medicines (Zhang et al., 2019a). Participants with medical insurance from their employer can therefore utilize better healthcare services, which also seems to be important to a better total QOL score, and the factors of physical symptoms and sickness impact. Thus, to ensure equal care, nurses should pay extra attention to those with basic medical insurance.
Regarding occupational status, employed older adults reported higher QOL than those who were unemployed (Kwak & Kim, 2017). However, we compared occupational status in the past and found that participants previously working in government departments or companies reported a higher score in quality of everyday activities/meaningful, socioeconomic situation and quality of close friend relations than those without a job. For the quality of family relations and close friends, we found that participants who had worked in a department or company reported lower QOL levels. These results may be explained by participants who had worked in a department or company with a retirement pension, were economically independent, and had been away from their families more.
5.1 Clinical implications
Our findings indicate that registered nurses should encourage older adults to participate in more meaningful and fun social activities to improve their QOL, as scores for this were very low. For example, exercise activities >30 min 3–7 days/week, increased the odds for a better total QOL and for the everyday activities/meaningful factor. This could be discussed in the regular annual physical examinations of those who can exercise. Both frailty and sickness impact should be screened for and discussed, especially with frail patients. Nurses should pay extra attention to those older adults with frailty, higher BMI, lower educational level and basic medical insurance who are more likely to have a lower QOL.
5.2 Limitations
The cross-sectional design of this study limits the exploration of causal relationships, and the convenience sampling method restricts the generalizability of the findings. However, the response rate was high and participant characteristics were similar to those reported in other studies (Kosilov et al., 2020; Li et al., 2020). We believe that our results can be generalized to Chinese community-dwelling older adults who can visit health centres, as it is a very important group where nurses could work to promote healthy ageing. Other strengths of the study were the low rate of internal missing data and the use of validated instruments. Cronbach's alpha values were >0.70 for the total score and all LSQ factors, except for the quality of family relations (0.63).
6 CONCLUSION
QOL domains such as the quality of everyday activities and the quality of friend relations were poor among Chinese community-dwelling older adults. Associated factors differed between the QOL domains and depended on whether associations were explored with the QOL outcome split at the 25th, 50th or 75th percentile. It was generally more difficult to predict being in the highest QOL group (75th percentile or higher). Frailty is associated with the total QOL score and the factors of physical symptoms and sickness impact. Further measures in nursing practice are required to screen for and prevent frailty to improve QOL and healthy ageing among community-dwelling older adults. Future interventional nursing studies are required to prevent frailty and to stimulate fun and meaningful everyday activities among community-dwelling older adults, which might in turn improve their QOL.
AUTHOR CONTRIBUTIONS
Conception and design: Lijuan Xu, Xuefen Lan, Yan Lou, Maria Engström. Acquisition of data: Lijuan Xu. Analysis of data: Lijuan Xu, Maria Engström. Interpretation of data: Lijuan Xu, Maria Engström. Drafting of the manuscript: Lijuan Xu, Maria Engström. Revision of Manuscript: Lijuan Xu, Xuefen Lan, Yan Lou, Maria Engström. All authors read and approved the final manuscript.
- Substantial contributions to conception and design, acquisition of data or analysis and interpretation of data;
- Drafting the article or revising it critically for important intellectual content.
ACKNOWLEDGEMENTS
The authors thank all the older people and managers of primary care centres of supporting data collection.
FUNDING INFORMATION
This study was supported by research project of National Natural Science Foundation of China (71904073), and research project of Natural Science Foundation of Zhejiang Province (LY19G030001).
CONFLICT OF INTREST
No conflict of interest was declared by the authors.
ETHICAL STATEMENTS
The survey has been approved by Medicine Ethical Committee of Lishui University (No. 2019–0081).
CONSENT TO PARTICIAPTE
Oral and written informed consent was obtained from all participants included in the study.
Open Research
DATA AVAILABILITY STATEMENT
The dataset generated during the current study is available from the corresponding author on reasonable request.
