If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Department of Nursing, Chinese Academy of Medical Sciences Peking Union Medical College, Peking Union Medical College Hospital (Dongdan Campus), Beijing, China
School of Nursing, Guangzhou Medical University, Guangzhou, ChinaDepartment of Nursing, The First Affiliated Hospital of Shenzhen University / Shenzhen Second People's Hospital, Shenzhen, China
Department of Nursing, The First Affiliated Hospital of Shenzhen University / Shenzhen Second People's Hospital, Shenzhen, ChinaSchool of Nursing, Anhui Medical University, Hefei, China
Department of Nursing, The First Affiliated Hospital of Shenzhen University / Shenzhen Second People's Hospital, Shenzhen, ChinaSchool of Nursing, University of South China, Hengyang, China
Address correspondence to Xiaohua Xie, MPH, Department of Nursing, The First Affiliated Hospital of Shenzhen University / Shenzhen Second People's Hospital, Shenzhen, China; or Xinjuan Wu, MSN, Department of Nursing, Chinese Academy of Medical Sciences Peking Union Medical College, Peking Union Medical College Hospital (Dongdan Campus), Beijing, China.
Address correspondence to Xiaohua Xie, MPH, Department of Nursing, The First Affiliated Hospital of Shenzhen University / Shenzhen Second People's Hospital, Shenzhen, China; or Xinjuan Wu, MSN, Department of Nursing, Chinese Academy of Medical Sciences Peking Union Medical College, Peking Union Medical College Hospital (Dongdan Campus), Beijing, China.
Department of Nursing, Chinese Academy of Medical Sciences Peking Union Medical College, Peking Union Medical College Hospital (Dongdan Campus), Beijing, China
To report the overall prevalence of social frailty among older people and provide information for policymakers and authorities to use in developing policies and social care.
Design
A systematic review and meta-analysis.
Setting and participants
We searched 4 databases (PubMed, Embase, Web of Science, and Google Scholar) to find articles from inception to July 30, 2022. We included cross-sectional and cohort studies that provided the prevalence of social frailty among adults aged 60 years or older, in any setting.
Methods
Three researchers independently reviewed the literature and retrieved the data. A risk of bias tool was used to assess each study’s quality. A random-effect meta-analysis was performed to pool the data, followed by subgroup analysis, sensitivity analysis, and meta-regression.
Results
From 761 records, we extracted 43 studies with 83,907 participants for meta-analysis. The pooled prevalence of social frailty in hospital settings was 47.3% (95% CI: 32.2%–62.4%); among studies in community settings, the pooled prevalence was 18.8% (95% CI: 14.9%–22.7%; P < .001). The prevalence of social frailty was higher when assessed using the Tilburg Frailty Indicator (32.3%; 95% CI: 23.1%–41.5%) than the Makizako Social Frailty Index (27.7%; 95% CI: 21.6%–33.8%) or Social Frailty Screening Index (13.4%; 95% CI: 8.4%–18.4%). Based on limited community studies in individual countries using various instruments, social frailty was lowest in China (4.9%; 95% CI: 4.2%–5.7%), followed by Spain (11.6%; 95% CI: 9.9%–13.3%), Japan (16.2%; 95% CI: 12.2%–20.3%), Korea (26.6%; 95% CI: 7.1%–46.1%), European urban centers (29.2%; 95% CI: 27.9%–30.5%), and the Netherlands (27.2%; 95% CI: 16.9%–37.5%). No other subgroup analyses showed any statistically significant prevalence difference between groups.
Conclusion and Implications
The prevalence of social frailty among older adults is high. Settings, country, and method for assessing social frailty affected the prevalence. More valid comparisons will await consensus on measurement tools and more research on geographically representative populations. Nevertheless, these results suggest that public health professionals and policymakers should seriously consider social frailty in research and program planning involving older adults.
all of which can lead to frailty, a clinical geriatric syndrome. Frailty was defined as a state with abnormal physiology and reduced reserves, resulting in increasing vulnerability, individual dependence, and susceptibility to death.
The original definition of frailty was predominantly physical. Nowadays, the concept of frailty has extended to include cognitive, psychological, physical, and social aspects.
Physical frailty and cognitive frailty have been well described, and a systematic review showed that the pooled prevalence rates of physical frailty and cognitive frailty were 12% (95% CI: 11%–13%) and 9% (95% CI: 8%–11%).
In 2017, based on the concept of social needs in the social production functions theory, social frailty was defined as a continuum of being at risk of losing, or having lost social resources, social behaviors, social activities, and self-management abilities to fulfill basic social needs.
Social frailty is independently associated with geriatric depression among older adults living in northern Japan: a cross-sectional study of ORANGE registry.
and is widely used among older adults. Social frailty is a serious concern among older adults because it has a negative impact on physical and mental health, resulting in declining self-care ability,
conducted a study among hospitalized older adults and found the prevalence of social frailty was 53.13%. The prevalence of social frailty therefore varies across countries and settings and can also be different when assessed using different tools. Recently, several new studies have explored the prevalence of social frailty among older adults.
However, social frailty is a relatively new concept, and a critical systematic look at the literature is merited. There are also no data on the prevalence of social frailty worldwide. This systematic review, therefore, aimed to conduct a meta-analysis synthesizing the pooled prevalence of social frailty among older adults, and to identify which factors could influence the prevalence of social frailty among older adults.
Materials and Methods
Protocol
This review complied with the PRISMA guidelines for meta-analysis.
The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration.
All the articles were published on the Internet, and institutional review board approval was therefore not sought. We have registered our protocol in PROSPERO (CRD42022321635).
Search Strategy
We searched in PubMed, EMBASE, Web of Science, and Google Scholar for studies from the date of the database’s establishment to July 30, 2022. The search terms included MESH terms and keywords. The keywords were: (elder∗ OR “older people” OR “older adult∗” OR aged OR senior∗) AND (“social frailty” OR “social vulnerability” OR “social frail∗”) AND (Prevalence OR Epidemiology∗). The detailed search strategy is shown in Supplementary Material 1. To avoid missing any relevant studies, we also examined all references from the included literature and other relevant articles.
Inclusion and Exclusion Criteria
The inclusion criteria were as follows: (1) participants aged 60 or older; (2) studies that reported the prevalence of social frailty, using a clear definition of social frailty; and (3) cross-sectional studies and cohort studies, regardless of language or country. The exclusion criteria were as follows: (1) abstracts, reviews, comments, or conferences; and (2) data on the prevalence of social frailty were not available or were insufficient to calculate the prevalence of frailty.
Study Selection and Data Extraction
Two of us (X.M.Z. and S.M.C.) independently screened records based on the title, abstract, and full texts, after deleting duplicated articles using EndnoteX9. A third reviewer with extensive experience in evidence-based training settled any disagreements.
The data were extracted by 3 authors (S.Y.X., M.F.G., and S.M.C.), who checked each other’s results. Any disagreements were discussed until reaching a consensus. The following data were extracted from individual articles: author, year of publication, country, setting, study design, age (mean or median), sample size, case of female/male participants, prevalence of social frailty, social frailty definition, sample frame, and selection method. For cohort studies, we also extracted the baseline data measuring social frailty, when appropriate. When studies used a multidimensional frailty assessment, such as the Tilburg Frailty Indicator, we extracted only the data on social frailty.
Assessment of Study Quality
Eligible studies were assessed by 3 investigators (S.M.C., M.F.G., S.Y.X.) using a risk of bias tool
to evaluate the study’s quality. It consists of 10 items and a summary assessment. Items 1 to 4 assess the external validity of the study in the areas of selection and nonresponse bias, and items 5 to 10 assess the internal validity. Items 5 to 9 evaluate measurement deviations, and item 10 evaluates deviations associated with the analysis. When the study received a score of 1, this indicated that it was a low risk of that particular type of bias. When checklist items were not reported or unclear, they were given a score of zero, and considered to be high risk. The overall risk of bias was classified into 3 categories: low (score > 8), moderate (score 6–8), and high (score ≤ 5).
Statistical Analysis
The prevalence of social frailty was either obtained directly from the article or calculated from available data. Cochran’s Q and the I2 statistic were used to determine whether there was significant heterogeneity among the studies. I2 values of 25%, 50%, and 75% represent low, moderate, and high heterogeneity. There was a significant degree of heterogeneity across the studies (I2 = 99.66%) because of the various assessments of social frailty used, as well as the location, sample size, population, and study design. We therefore used the random-effects model to pool the prevalence of social frailty. Visual Funnel plots and Begg’s test were used to analyze publication bias. We also performed a broad subgroup analysis based on country, age (<75 years vs ≥75 years), setting, sample size (≤500 and >500), study design, the tool used to assess social frailty (Makizako social frailty index, Social Frailty Screening Index, Tilburg Frailty Indicator, and others) and gender. Finally, we also used sensitivity and meta-regression analysis. We used Stata 16.0 (StataCorp) for all the data analyses.
Results
Search Results
We obtained 761 records from 4 databases, with 339 from PubMed, 166 from Embase, 116 from Web of Science, and 140 from Google Scholar. A total of 492 duplicated records were removed using EndnoteX9 software. After selecting based on titles and abstracts, we examined the full text of 93 papers. Of these, 14 records were review or conference abstracts, 29 records did not provide the prevalence of social frailty, and 7 studies focused on people younger than 60 years. This gave a total of 43 studies for review and analysis.
Social frailty is independently associated with geriatric depression among older adults living in northern Japan: a cross-sectional study of ORANGE registry.
Association of social frailty with physical health, cognitive function, psychological health, and life satisfaction in community-dwelling older Koreans.
Screening value of social frailty and its association with physical frailty and disability in community-dwelling older koreans: aging study of pyeongchang rural area.
Social frailty has a stronger impact on the onset of depressive symptoms than physical frailty or cognitive impairment: a 4-year follow-up longitudinal cohort study.
Factors associated with physical, psychological and social frailty among community-dwelling older persons in Europe: a cross-sectional study of Urban Health Centres Europe (UHCE).
The characteristics of the 43 included studies are summarized in Table 1. They were published between 2013 and 2022, contained 83,907 participants from 10 countries, and had sample sizes ranging from 48 to 11,241. The mean or median age ranged from 69.5 to 81.7 years, and 6 studies did not provide the detailed mean or median age for the total sample.
Social frailty is independently associated with geriatric depression among older adults living in northern Japan: a cross-sectional study of ORANGE registry.
Social frailty has a stronger impact on the onset of depressive symptoms than physical frailty or cognitive impairment: a 4-year follow-up longitudinal cohort study.
Association of social frailty with physical health, cognitive function, psychological health, and life satisfaction in community-dwelling older Koreans.
Screening value of social frailty and its association with physical frailty and disability in community-dwelling older koreans: aging study of pyeongchang rural area.
Factors associated with physical, psychological and social frailty among community-dwelling older persons in Europe: a cross-sectional study of Urban Health Centres Europe (UHCE).
Social frailty is independently associated with geriatric depression among older adults living in northern Japan: a cross-sectional study of ORANGE registry.
Association of social frailty with physical health, cognitive function, psychological health, and life satisfaction in community-dwelling older Koreans.
Screening value of social frailty and its association with physical frailty and disability in community-dwelling older koreans: aging study of pyeongchang rural area.
Social frailty has a stronger impact on the onset of depressive symptoms than physical frailty or cognitive impairment: a 4-year follow-up longitudinal cohort study.
Factors associated with physical, psychological and social frailty among community-dwelling older persons in Europe: a cross-sectional study of Urban Health Centres Europe (UHCE).
Detailed information on how social frailty was measured and defined in each study is shown in Supplementary Table 1. Overall, 35 studies involved community-dwelling older people,
Social frailty is independently associated with geriatric depression among older adults living in northern Japan: a cross-sectional study of ORANGE registry.
Association of social frailty with physical health, cognitive function, psychological health, and life satisfaction in community-dwelling older Koreans.
Screening value of social frailty and its association with physical frailty and disability in community-dwelling older koreans: aging study of pyeongchang rural area.
Social frailty has a stronger impact on the onset of depressive symptoms than physical frailty or cognitive impairment: a 4-year follow-up longitudinal cohort study.
Factors associated with physical, psychological and social frailty among community-dwelling older persons in Europe: a cross-sectional study of Urban Health Centres Europe (UHCE).
Table 1Summary of the Included Studies on the Prevalence of Social Frailty (n = 43)
Author, year
Population Source
Sampling Methods
Setting
Country
Design
Age, y, Mean (SD)/Median
Sample Size
Female
Definition of Social Frailty
Prevalence of Frailty, %
Sugie 2022
Older people living in Nammoku village
Population survey
Community
Japan
Cross-sectional
81.5 (4.5)
268
147
LSNS-6
8.20
Miyata 2022
Residents aged 65 years from Tarumizu city
Population survey
Community
Japan
Cross-sectional
74.2 (6.4)
596
379
MSFI
18.60
Kume 2021
General population of Akita prefecture
Population survey
Community
Japan
Cross-sectional
75.0 (5.8)
313
213
MSFI
21.70
Kodama 2022
Older adults from 6 areas of Yokote city in Akita prefecture
Convenience sample
Community
Japan
Cohort study
≥60
161
NA
MSFI
13.04
Hayashi 2021
Older people living in Minokamo city
Population survey
Community
Japan
Cross-sectional
81.1 (5.0)
988
520
SFSI
15.30
Doi 2021
Part study of NCGG-SGS; older adults living in Obu city
Population survey
Community
Japan
Cohort study
71.7 (5.3)
4642
2330
MSFI
44.66
Ye 2021
Baseline data of Urban Health Centres Europe project
Population survey
Community
Europe
Cross-sectional
79.7 (5.7)
2289
1379
TFI
29.40
Usui 2021
Eligible older people from Kisen Hospital and Kisen Clinic
Convenience sample
Hospital
Japan
Cross-sectional
74.1 (6.8)
158
45
MSFI
59.50
Ono 2021
Older adults from agricultural community located in Aizu region of Fukushima prefecture
Population survey
Community
Japan
Cohort study
79.4 (3.5)
748
398
MSFI
23.26
Ono 2021
Patients with gastrointestinal cancer
Convenience sample
Hospital
Japan
Cohort study
>60
181
42
MSFI
46.80
Ko 2021
Data from the 2017 National Survey of Older Koreans from 17 metropolitan cites and provinces
National sample
Community
Korea
Cross-sectional
74.5 (6.2)
10,081
6035
MSFI
44.70
Jujo 2021
Older hospitalized people with heart failure from 15 hospitals
National sample
Hospital
Japan
Cohort study
≥65
1240
527
MSFI
66.50
Huang 2021
Older people from a community center in Nagoya
Convenience sample
Community
Japan
Cohort study
69.5 (4.5)
663
376
Bunt’s social frailty concept
6.30
Chen 2021
Population-based study in Rugao, Jiangsu province; older adults from 31 communities of Jiang’an township
Population survey
Community
China
Cross-sectional
75.3 (3.9)
1764
820
Six questions
3.60
Okumura 2020
Patients with gastrointestinal cancer form a single university hospital
Convenience sample
Hospital
Japan
Cohort study
71
48
11
MSFI
27.00
Nagai 2020
Older people living in Tamba-Sasayama city
Population survey
Community
Japan
Cross-sectional
NA
625
NA
SFSI
7.70
Zhang 2019
Baseline data of Urban Health Centres Europe project
Population survey
Community
Europe
Cross-sectional
79.7 (5.6)
2167
1313
TFI
29.00
Yoo 2019
Older people from a national multicenter cohort study, living in urban and rural regions
National sample
Community
Korea
Cross-sectional
76.2 (3.93)
1539
815
MSFI
11.70
Tsutsumimoto 2019
Older people from Obu city, Aichi prefecture
Population survey
Community
Japan
Cohort study
71.7 (5.3)
3720
1918
MSFI
10.80
Park 2019
Older adults living in the rural area of Pyeongchang
Population survey
Community
Korea
Cross-sectional
74.9 (6.0)
408
236
MSFI
20.50
Nakakubo 2019
Older people recruited from Obu city
Population survey
Community
Japan
Cross-sectional
71.91 (5.49)
4427
2274
MSFI
11.20
Makizako 2018
Older people living in Tarumizu city
Population survey
Community
Japan
Cross-sectional
75.4 (6.5)
353
264
MSFI
14.70
Hirase 2019
Older people from Unzen city attending community-based exercise classes
Convenience sample
Community
Japan
Cross-sectional
79.1 (6.0)
248
206
MSFI
22.20
Yamada 2018
Older people living in a city in Shiga prefecture
Population survey
Community
Japan
Cohort study
75.2 (6.6)
6603
3692
SFSI
18.00
Tsutsumimoto 2018
Older people from Obu city
Population survey
Community
Japan
Cohort study
71.3 (0.08)
3538
1820
MSFI
29.90
Makizako 2018
Longitudinal data of older adults derived from first and second waves of OSHPE
Population survey
Community
Japan
Cohort study
70.4 (4.1)
1226
633
MSFI
3.60
Ma 2018
Older adults from Beijing community
Population survey
Community
China
Cohort study
≥60
1697
NA
HALFT scale
7.70
Bae 2018
Older people living in Obu city
Population survey
Community
Japan
Cross-sectional
72.5 (5.2)
4251
2291
MSFI
10.37
Tsutsumimoto 2017
Data from Obu study of health promotion for elderly people; older people from Obu city
Population survey
Community
Japan
Cross-sectional
71.9 (5.5)
4425
2274
MSFI
11.10
Yamashita 2021
Older hospitalized people with heart failure from 15 hospitals
National sample
Hospital
Japan
Cohort study
81
1332
574
MSFI
41.60
Alonso 2020
Noninstitutionalized adults aged 65 years or more from Asturias
Population survey
Community
Spain
Cross-sectional
76.2 (7.2)
445
255
Accumulated functional deficits
23.10
Noguchi 2021
Older people recruited at Togo public town hall
Population survey
Community
Japan
Cross-sectional
73.0 (5.8)
300
153
MSFI
17.33
Makizako 2015
Older people from Obu study of health promotion for elderly people and living in Obu
Population survey
Community
Japan
Cohort study
71.7 (5.3)
4304
2207
MSFI
10.20
Inoue 2022
Participants in a frailty clinic of a general geriatric hospital
Convenience sample
Hospital
Japan
Cross-sectional
76.5 (7.2)
495
340
SFSI
12.30
Gobbens 2021
Participants at Aalborg University Hospital
Convenience sample
Hospital
Netherlands
Cross-sectional
76.8 (7.5)
1267
633
TFI
53.13
Hironaka 2020
Older adults residing in 9 towns in Itabashi ward in Tokyo
Population survey
Community
Japan
Cross-sectional
73.3 (6.6)
682
415
MSFI
23.20
Renne 2018
Older adults from small villages close to Amsterdam
Population survey
Community
Netherlands
Cross-sectional
76.5 (5.1)
241
116
TFI
32.78
Gobbens 2017
Dutch older people recruited via Web-based questionnaire
Convenience sample
Community
Netherlands
Cross-sectional
76.6 (4.7)
671
205
TFI
31.10
Garre-Olmo 2013
Older people living in 8 rural villages
Population survey
Community
Spain
Cohort study
81.7 (4.8)
875
509
Accumulated functional deficits
8.90
Lee 2020
Nationwide survey of older adults in Korea
National sample
Community
Korea
Cohort study
72.9 (6.7)
11,241
6711
Bunt’s social frailty concept
7.70
Verver 2019
Older adults in different regions of North-Holland province
Population survey
Community
Netherlands
Cross-sectional
78.8 (6.4)
1768
546
TFI
18.40
Adachi 2022
Hospitalized older adults in cardiac rehabilitation
Convenience sample
Hospital
Japan
Cohort study
75
184
62
MSFI
65.2
Ko 2022
Older people living in the community
NA
Community
Korea
Cross-sectional
≥65
735
234
MSFI
48.4
CFAI, Comprehensive Frailty Assessment Instrument; HALFT, acronym for the 5 components: Help, participation, Loneliness, Financial, and Talk; LSNS-6, Lubben Social Network Scale.
MSFI, Makizako Social Frailty Index; NA, not applicable; NCGG-SGS, National Center for Geriatric and Gerontology-Study of Geriatric Syndromes; OSHPE, Obu Health Promotion for the Elderly Study; SFSI, Social Frailty Screening Index; TFI, Tilburg Frailty Indicator.
The studies included in our systematic review had a low to moderate risk of bias. Overall, 11 showed a low risk of bias, and 32 had a moderate risk. Their grades varied from 6 to 10 on a scale of 1 to 10 (Supplementary Table 2).
Prevalence of Social Frailty
The prevalence of social frailty in individual studies ranged from 3.6% to 66.5%. The pooled prevalence of social frailty was 23.9% (95% CI: 19.9%−27.9%, I2 = 99.66%, P < .001) (Figure 2).
Fig. 2Meta-analysis for the prevalence of social frailty among older adults.
Subgroup analyses based on country by excluding those studies from hospital settings
The results showed that the pooled prevalence of social frailty in Japan and Korea was 16.2% (95% CI: 12.2%–20.3%) and 26.6% (95% CI: 7.1%–46.1%), respectively. The pooled prevalence of social frailty was lowest in China, at 4.9% (95% CI: 4.2%–5.7%). The pooled prevalence of social frailty was 11.6% (95% CI: 9.9%–13.3%) in Spain and 27.2% (95% CI: 16.9%–37.5%) in the Netherlands. Two studies were from the Urban Health Centres Europe project, covering 5 European countries, and the pooled prevalence of social frailty was 29.2% (95% CI: 27.9%–30.5%) (Table 2).
Table 2Subgroup Analyses by Country, Age, Setting, Sample Size, Design, Gender, Definition for the Meta-analysis of Prevalence of Social Frailty
We classified age into 2 groups (at least 75 years vs younger than 75 years). The pooled prevalence of social frailty was higher in those aged 75 years or older than those younger than 75 years (24.3%; 95% CI: 18.7%–29.9% vs 20.0%, 95% CI: 13.9%–26.1%). However, the difference was not statistically significant (P = .40) (Table 2).
Subgroup analyses by study design and sample size
The pooled prevalence of social frailty was 22.2% (95% CI: 17.4%–27.1 %) for studies with sample sizes greater than 500 and 27.6% (95% CI: 19.9%–35.2%) in studies with sample size of less than 500. In addition, the pooled prevalence of social frailty was 22.9% (95% CI: 17.4%–28.5%) in cross-sectional studies and 25.4% (95% CI: 19.1%–31.7%) in cohort studies. These differences were not statistically significant (P = .65 for study design and P = .35 for sample size) (Table 2).
Subgroup analyses by setting
Overall, 43 studies reported the setting, including 79,002 participants in the community and 4905 participants in hospital, with a higher prevalence of social frailty in hospital (47.3%, 95% CI: 32.2%–62.4% vs 18.8%, 95% CI: 14.9%–22.7%). The difference was statistically significant (P < .001) (Table 2). Regarding community settings, 27 studies were based on population surveys, 3 studies used national samples, and 4 studies used convenience samples. The subgroup analysis based on sampling methods showed no statistically significant difference in the pooled prevalence of social frailty among community-dwelling older adults (P = .94) (Supplementary Figure 1).
Subgroup analyses by social frailty assessment tool
Studies used several different social frailty assessment tools. Some used single-dimensional social frailty assessment tools but others used a multidimensional assessment. Overall, 26 studies used the Makizako Social Frailty Index and had a pooled prevalence of 27.7% (95% CI: 21.6%–33.8%). Four studies used the Social Frailty Screening Index, and had a prevalence of 13.4% (95% CI: 8.4%–18.4%), 6 studies used the Tilburg Frailty Indicator, with a prevalence of 32.3% (95% CI: 23.1%–41.5%), and 7 studies used other tools and had a pooled prevalence of 8.8% (95% CI: 6.5%–11.1%). Differences in the prevalence of social frailty between studies using different definitions were statistically significant (P < .001) (Table 2).
Subgroup analysis by gender
Only 19 studies reported the prevalence of social frailty by gender. The pooled prevalence of social frailty was higher in women than men (26.0%, 95% CI: 17.6%–34.3% vs 23.5%, 95% CI: 16.7%–30.2%). However, the difference was not statistically significant (P = .55) (Table 2).
Publication bias
The publication bias test found asymmetry across these studies (Supplementary Figure 2A), and the results of Begg’s test suggested a possible publication bias (P = .011) (Supplementary Figure 2B).
Meta-regression
The meta-regression on age showed that age did not affect the prevalence of social frailty, with a regression coefficient of 0.058 (95% CI: −0.0098 to 0.021), P = .45, 38 studies (Supplementary Figure 3).
Sensitivity
After omitting each study and pooling the remaining studies using sensitivity analysis, there were no statistically significant differences between the combined effect value and the total combined value, indicating that the results of this study were stable (Supplementary Figure 4).
Discussion
Our meta-analysis showed that the pooled prevalence of social frailty was 23.9% (95% CI: 19.9%−27.9%). Our subgroup analysis showed that the country, setting, and definition of social frailty affected the prevalence of social frailty among older adults. However, age, gender, and study design had no effect. To the best of our knowledge, this is the first meta-analysis to summarize the evidence of the prevalence of social frailty among older adults.
This study found statistically significant differences in social frailty prevalence by country. Social frailty was lowest in China (4.9%), followed by Spain (11.6%), Japan (16.2%), Korea (26.6%), 5 European countries (29.2%), and the Netherlands (27.2%) among community-dwelling older adults. These differences may reflect differences in sampling, measurement tools, and/or differences in social structures, cultures,
More studies from different countries, such as China, Spain, and the Netherlands, are needed to fully explore this topic.
The subgroup analysis based on study design showed that the prevalence of social frailty was 22.9% (95% CI: 17.4%–28.5%) for cross-sectional studies and 25.4% (95% CI: 19.1%–31.7%) for cohort studies. The most likely explanation for this difference is methodology: cross-sectional and cohort studies have different inclusion and exclusion criteria. Generally, most of the cross-sectional studies of older community-dwelling people recruited participants who could walk independently and could communicate with the investigator.
Screening value of social frailty and its association with physical frailty and disability in community-dwelling older koreans: aging study of pyeongchang rural area.
Factors associated with physical, psychological and social frailty among community-dwelling older persons in Europe: a cross-sectional study of Urban Health Centres Europe (UHCE).
at the baseline survey. In addition, 7 cohort studies provided baseline data on social frailty (which were extracted for our data analysis) at the beginning of the survey, and therefore were not true cohort studies. In addition, 6 studies excluded from their final analysis data for some older people who did not complete the follow-up. Therefore, missing data owing to noncompletion of follow-up may have led to differences in the prevalence of social frailty.
The common tools used to assess social frailty are the Makizako Social Frailty Index, Social Frailty Screening Index, and Tilburg Frailty Indicator. The pooled prevalence of social frailty measured by the 3 instruments respectively was 27.7%, 13.4%, and 32.3% (P < .001). The most common assessment tool was the Makizako Social Frailty Index, which was used in 26 studies. The highest prevalence was with the use of Tilburg Frailty Indicator. Bessa et al.
also found different results for alternative measures of social frailty. The higher social frailty prevalence when using the Tilburg Frailty Indicator may be because the studies using this scale had a higher mean age, all ≥75 years. In addition, there are also other differences between the Makizako Social Frailty Index and Tilburg Frailty Indicator. Tilburg Frailty Indicator has been used to assess social frailty,
The Tilburg Frailty Indicator may also be more sensitive because it contains only 3 items on social frailty. The Makizako Social Frailty Index, however, was specifically designed to assess social frailty, and it contains 5 items. Participants are classified into socially frail, pre-frail, and robust.
No consensus has been reached on the most appropriate tools for assessing social frailty. The Tilburg Frailty Indicator is usually chosen in Europe, and the Makizako Social Frailty Index and Social Frailty Screening Index were more likely to be used in Japan and South Korea. This suggests that there are some challenges in measuring social frailty across cultures, especially in validating the efficacy, applying assessment tools in other countries, and establishing a standardized method for social frailty screening.
The subgroup analysis by setting showed that the prevalence of social frailty was lower among people in the community than in hospital (18.8% vs 47.3%; P < .001). This is similar to the result of another meta-analysis, in which the overall prevalence of frailty in older adult individuals in hospitals was 41.4%.
This may be because older adult patients in hospital tended to do fewer social activities because of their illness or condition, and were in poorer physical condition than people living in the community. Hospitalized older adults are also more likely to have negative emotions.
In addition, hospitalized patients may be taking more drugs and medication, and the side effects of these could have a negative impact on social frailty. Some studies have suggested that depression,
Screening value of social frailty and its association with physical frailty and disability in community-dwelling older koreans: aging study of pyeongchang rural area.
Hospitalized patients are therefore more likely to experience social frailty. It may be helpful for hospital staff to assess the social frailty of hospitalized older adults in a timely manner, conduct individualized assessments, and develop intervention measures to delay or prevent the occurrence of social frailty.
Our results showed that the pooled prevalence of social frailty was higher in adults aged 75 years or older than in those younger than 75 years (24.3% vs 20.0%). Surprisingly, however, this result was not statistically significant (P = .40), and the meta-regression did not suggest that age influenced the prevalence of social frailty. However, we still think this is a possibility. A previous study found that age was a risk factor for social frailty.
Older age is associated with living alone, which is a component of social frailty. In addition, aging causes natural deterioration of bodily functions. A study showed that after controlling for confounding factors, hearing loss was associated with social frailty (odds ratio 2.17; 95% CI: 1.43–3.30).
In the age subgroup analysis, our results were similar to another meta-analysis of frailty, which also found that the prevalence of frailty increased with age, although the correlation was weak.
This indicates that the prevalence of social frailty may vary with age, and therefore may be age-related.
Our meta-analysis had several strengths. First, to our knowledge, this meta-analysis is the first to have analyzed the prevalence of social frailty, and therefore provides an overall picture of social frailty among older adults. It highlights the importance of early screening and intervention to reduce the impact of social frailty on older people. Second, the analysis included a large number of studies from different countries around the world, and the sample size was large, with two-thirds of studies including more than 500 people. Third, we used comprehensive analyses, including various subgroup analyses, sensitivity analyses, meta-regression, and tests of publication bias, to identify whether the study results were stable and reliable.
Our meta-analysis also had some limitations. First, the review did not cover any particular population groups, such as older nursing home residents. Second, the international sample was uneven, with significantly more studies from Japan than other countries. Third, not all the articles were of high quality and there might be some publication bias. Fourth, we did not search any gray literature, which may have biased our findings to the degree of uncertainty. Fifth, our results should be taken with caution because there is considerable heterogeneity. However, this is often unavoidable in meta-analyses of observational studies and does not necessarily negate the findings.
We performed sufficient analyses to investigate potential sources of heterogeneity, including subgroup, random-effects meta-regression and sensitivity analyses. The results show that country, setting, and definition of social frailty make significant contributions to heterogeneity. Sixth, one subgroup analysis showed very large CIs. The 95% CI was between 7.1% and 46.1% for people in Korea. Finally, the different assessment tools arrive at very different estimates, and therefore our meta-analysis results should be used with caution and especially across different measurement tools.
Conclusion and Implications
The prevalence of social frailty is high in the older adult population. Our study found that it was influenced by country, settings, and the assessment tools for social frailty. However, it is clear that public health professionals and policymakers need to take social frailty among older adults seriously. Future clinical trials or cohort studies are needed to evaluate the best intervention and strategies for reducing the prevalence of social frailty.
Acknowledgments
We thank Melissa Leffler, MBA, from Liwen Bianji (Edanz) (www.liwenbianji.cn) for editing the English text of a draft of this manuscript.
Author contributions: X.M.Z. and S.M.C. were responsible for drafting, revising, and finalizing the manuscript. X.H.X. and X.J.W. initiated the study concept and design. X.M.Z. and S.M.C. were responsible for determining the research strategy and screening. S.Y.X., M.F.G., and S.M.C. were responsible for data extraction, data analysis, and quality bias assessment.
Supplementary File 1. Search Strategy
PubMed:
#1: Search: "social frailty" Sort by: Most Recent
#2: Search: "social vulnerability" Sort by: Most Recent
#3: Search: "social frail∗" Sort by: Most Recent
#4: Search: elder∗ OR "older people∗" OR "older adult∗" OR aged OR senior∗ Sort by: Most Recent
#5: Search: (Prevalence OR Epidemiology∗). Sort by: Most Recent
#6: Search: (("social frailty") OR ("social vulnerability")) OR ("social frail∗") Sort by: Most Recent
#7: Search: (((("social frailty") OR ("social vulnerability")) OR ("social frail∗")) AND ((Prevalence OR Epidemiology∗).)) AND (elder∗ OR "older people∗" OR "older adult∗" OR aged OR senior∗) Sort by: Most Recent
Embase:
#1('social frailty' OR 'social vulnerability'/exp OR 'social frail∗')
#2:('older adults' OR 'older people' OR 'aged'/exp OR 'aged' OR 'aged patient' OR 'aged people' OR 'aged person' OR 'aged subject' OR 'elderly' OR 'elderly patient' OR 'elderly people' OR 'elderly person' OR 'elderly subject' OR 'senior citizen' OR 'senium')
#3: ('prevalence'/exp OR 'epidemiology'/exp OR 'clinical epidemiology' OR 'confounding factors (epidemiology)' OR 'confounding factors, epidemiologic' OR 'controlled before after studies' OR 'controlled before and after studies' OR 'controlled before and after study' OR 'controlled before-after studies' OR 'effect modifier, epidemiologic' OR 'effect modifiers (epidemiology)' OR 'effect modifiers (psychology)' OR 'environmental epidemiology' OR 'epidemiologic confounding factors' OR 'epidemiologic effect modifier' OR 'epidemiologic factors' OR 'epidemiologic methods' OR 'epidemiologic research' OR 'epidemiologic research design' OR 'epidemiologic studies' OR 'epidemiologic study characteristics' OR 'epidemiologic study characteristics as topic' OR 'epidemiologic survey' OR 'epidemiological research' OR 'epidemiology' OR 'epidemiometry' OR 'historically controlled study' OR 'interrupted time series analysis' OR 'precipitating factors' OR 'sampling studies')
#4: #1 AND #2 AND #3
Supplementary File 2. PRISMA Checklist
Tabled
1
Section/topic
#
Checklist item
Reported on page #
Title
Title
1
Identify the report as a systematic review, meta-analysis, or both.
Page 1
Abstract
Structured summary
2
Provide a structured summary including, as applicable: background; objectives; data sources; study eligibility criteria, participants, and interventions; study appraisal and synthesis methods; results; limitations; conclusions and implications of key findings; systematic review registration number.
Page 1–2
Introduction
Rationale
3
Describe the rationale for the review in the context of what is already known.
Pages 3–4
Objectives
4
Provide an explicit statement of questions being addressed with reference to participants, interventions, comparisons, outcomes, and study design (PICOS).
Pages 4–5
Methods
Protocol and registration
5
Indicate if a review protocol exists, if and where it can be accessed (eg, Web address), and, if available, provide registration information including registration number.
Page 5 line 92
Eligibility criteria
6
Specify study characteristics (eg, PICOS, length of follow-up) and report characteristics (eg, years considered, language, publication status) used as criteria for eligibility, giving rationale.
Page 5 lines 101–107
Information sources
7
Describe all information sources (eg, databases with dates of coverage, contact with study authors to identify additional studies) in the search and date last searched.
Page 4 lines 98–99
Search
8
Present full electronic search strategy for at least 1 database, including any limits used, such that it could be repeated.
Page 5 lines 96–98 Supplementary File 1
Study selection
9
State the process for selecting studies (ie, screening, eligibility, included in systematic review, and, if applicable, included in the meta-analysis).
Describe method of data extraction from reports (eg, piloted forms, independently, in duplicate) and any processes for obtaining and confirming data from investigators.
Page 6 lines 111–120
Data items
11
List and define all variables for which data were sought (eg, PICOS, funding sources) and any assumptions and simplifications made.
Page 6 lines 114–120
Risk of bias in individual studies
12
Describe methods used for assessing risk of bias of individual studies (including specification of whether this was done at the study or outcome level), and how this information is to be used in any data synthesis.
Page 6 lines 121–131
Summary measures
13
State the principal summary measures (eg, risk ratio, difference in means).
Page 7 lines 133–134
Synthesis of results
14
Describe the methods of handling data and combining results of studies, if done, including measures of consistency (eg, I2) for each meta-analysis.
Page 7 lines 134–140
Section/topic
#
Checklist item
Reported on page #
Risk of bias across studies
15
Specify any assessment of risk of bias that may affect the cumulative evidence (eg, publication bias, selective reporting within studies).
Page 7 line 140
Additional analyses
16
Describe methods of additional analyses (eg, sensitivity or subgroup analyses, meta-regression), if done, indicating which were prespecified.
Page 7 lines 140–145
Results
Study selection
17
Give numbers of studies screened, assessed for eligibility, and included in the review, with reasons for exclusions at each stage, ideally with a flow diagram.
For all outcomes considered (benefits or harms), present, for each study: (1) simple summary data for each intervention group, (2) effect estimates and confidence intervals, ideally with a forest plot.
Pages 8–9 lines 176–179
Synthesis of results
21
Present results of each meta-analysis done, including confidence intervals and measures of consistency.
Summarize the main findings including the strength of evidence for each main outcome; consider their relevance to key groups (eg, healthcare providers, users, and policy makers).
Pages 11–15 lines 241–328
Limitations
25
Discuss limitations at study and outcome level (eg, risk of bias), and at review-level (eg, incomplete retrieval of identified research, reporting bias).
Pages 15–16 lines 329–344
Conclusions
26
Provide a general interpretation of the results in the context of other evidence, and implications for future research.
Page 16 lines 346–352
Funding
Funding
27
Describe sources of funding for the systematic review and other support (eg, supply of data); role of funders for the systematic review.
Tabled
1Social Frailty Screening Index: The measure includes 4 questions.
Items
0 point
1 point
Are you satisfied with your economic condition?
Very satisfied, satisfied
Unsatisfied, very unsatisfied
Do you live alone?
No
Yes
How often do you participate in the following groups: volunteer, sports, hobby, learning or cultural, nursing care prevention, senior citizens, or residents’ associations?
“Four or more times a week,” “Two or 3 times a week,” “Once a week,” “One to 3 times a month,” “A few times in a year”
Never
How do you get along with your neighbors?
“I have a neighbor who comes and goes to each house,” “I have a neighbor to chat with on the street,” “I have a neighbor to say hello to,” and “I do not communicate with neighbors”
“I have a neighbor to say hello to” or “I do not communicate with neighbors”
Note: Scores of 2 or more as social frailty, 1 as pre-social frailty, and 0 as non-social frailty.
Tabled
1Tilburg Frailty Indicator (TFI): The TFI is a questionnaire based on a multidimensional approach to frailty and was made and validated for use in primary care. Part B consists of 15 self-reported questions covering 3 domains: physical (8 items, score range 0–8), psychological (4; 0–4) and social frailty (3; 0–3). Items have answer categories 0 (no) and 1 (sometimes or yes). Participants with total score of at least 5 were diagnosed as being frail. The cut points for physical, psychological, and social frailty were 3, 2, and 2, respectively.
Tabled
1Lubben Social Network Scale (LSNS-6): The LSNS-6 is an extremely rapid measure of social isolation in a clinical setting. The total scale score is an equally weighted sum of scores on the six 5-point items.
Items
0–5 points
How many relatives do you see or hear from at least once a month?
How many relatives do you feel close to such that you could call on them for help?
How many relatives do you feel at ease with that you can talk about private matters?
How many friends do you see or hear from at least once a month?
How many friends do you feel close to such that you could call on them for help?
How many friends do you feel at ease with that you can talk about private matters?
Note: Total scores range from 0 to 30. Social frailty (n = 22) as LSNS-6 scores <12.
Tabled
1Bunt’s social frailty concept: A validated 4-item questionnaire derived from Bunt’s social frailty concept (general resources, social resources, social behavior, and fulfillment of basic social needs) was used to assess social frailty.
Items
1 point
0 point
Financial difficulty
Need financial support
No need for financial support
Household status
Living alone
Not living alone
Social activity
Nonparticipation in social activities
Regular participation in social activities
Regular contact with others
Total scores of the Lubben Social Network Scale: <12 points
Total scores of the Lubben Social Network Scale: ≥12 points
Note: Social frailty (2–4 points), social prefrailty (1 point), and social robustness (0 points).
When faced with troubles, do you have anyone to talk to?
∗
In the past week, have you visited your friends, or have your friends come to visit you?
∗
Do you feel lonely?
∗
Financial difficulty
∗
Note: Participants showing none or 1 of these components were considered non-socially frail. those showing 2 or 3 components were considered pre-socially frail; and those showing 4 or more components were considered socially frail.
Tabled
1HALFT scale: The HALFT scale comprises 5 items in Chinese.
Items
Yes (1 point)
No (0 point)
Inability to help others within the past 12 months
Limited social participation in the previous 12 months
Loneliness in the past week
Financial difficultly over the past 12 months
Not having anyone to talk to every day
Note: The HALFT scale score ranges from 0–5 points: a score of 0 was considered non-socially frail; 1–2 was considered pre-socially frail, and a score of ≥3 indicated social frailty.
Supplementary Table 2Risk of Bias of Included Studies (n = 42)
Author, year
External Validity Items
Internal Validity Items
Overall Score
Overall Risk of Bias
1
2
3
4
5
6
7
8
9
10
Sugie 2022
0
0
0
1
1
1
1
1
1
1
7
Moderate
Miyata 2022
0
0
1
0
1
1
1
1
1
1
7
Moderate
Kume 2021
0
0
0
1
1
1
1
1
1
1
7
Moderate
Kodama 2022
0
0
0
1
1
1
1
1
1
1
7
Moderate
Hayashi 2021
0
0
1
0
1
1
1
1
1
1
7
Moderate
Doi 2021
0
1
1
1
1
1
1
1
1
0
8
Moderate
Ye 2021
1
0
1
1
1
1
1
1
1
0
8
Moderate
Usui 2021
0
1
1
0
1
1
1
1
1
0
7
Moderate
Ono 2021
0
1
1
0
1
1
1
1
1
0
7
Moderate
Ono 2021
0
0
0
1
1
1
1
1
1
1
7
Moderate
Ko 2021
1
1
1
1
1
1
1
1
1
1
10
Low
Jujo 2021
1
1
1
1
1
1
1
1
1
1
10
Low
Huang 2021
0
1
1
1
1
1
1
1
1
1
9
Low
Chen 2021
0
1
1
1
1
0
0
1
1
1
7
Moderate
Okumura 2020
0
1
1
1
1
1
1
1
1
1
9
Low
Nagai 2020
0
0
0
0
1
1
1
1
1
1
6
Moderate
Zhang 2019
1
1
0
1
1
0
1
1
1
1
8
Low
Yoo 2019
1
1
1
1
1
0
1
1
1
1
9
Low
Tsutsumimoto 2019
1
0
1
0
1
1
1
1
1
1
8
Moderate
Park 2019
1
1
0
1
1
1
1
1
1
1
9
Low
Nakakubo 2019
1
0
1
1
1
0
0
1
1
1
7
Moderate
Makizako 2018
1
1
1
1
1
0
0
1
1
1
8
Moderate
Hirase 2019
0
1
0
1
1
0
1
1
1
1
7
Moderate
Yamada 2018
0
1
1
1
1
0
1
1
1
1
8
Moderate
Tsutsumimoto 2018
1
1
1
0
1
0
0
1
1
1
7
Moderate
Makizako 2018
1
1
1
0
1
1
1
1
1
1
9
Low
Ma 2018
0
1
1
1
1
1
1
1
1
1
9
Low
Bae 2018
0
1
1
0
1
1
1
1
1
1
8
Moderate
Tsutsumimoto 2017
1
1
1
1
1
1
1
1
1
1
10
Low
Yamashita 2021
0
1
1
1
1
0
1
1
1
0
7
Moderate
Alonso 2020
0
1
1
1
1
0
0
1
1
1
7
Moderate
Noguchi 2021
0
1
1
1
1
0
0
1
1
1
7
Moderate
Makizako 2015
1
1
1
1
1
0
0
1
1
1
8
Moderate
Inoue 2022
0
1
0
1
0
1
1
1
1
1
7
Moderate
Gobbens 2021
0
1
1
1
1
0
1
1
1
1
8
Moderate
Hironaka 2020
0
1
1
0
1
1
1
1
1
1
8
Moderate
Renne 2018
0
1
1
0
1
0
1
1
1
1
7
Moderate
Gobbens 2017
1
1
1
0
1
0
1
1
1
1
8
Moderate
Garre-Olmo 2013
0
1
1
0
1
0
1
1
1
1
7
Moderate
Lee 2020
1
1
1
1
1
0
0
1
0
1
7
Moderate
Adachi 2022
0
1
0
1
1
1
1
1
1
1
8
Moderate
Ko 2022
1
1
0
0
0
1
1
1
1
1
7
Moderate
Verver 2019
1
1
1
0
1
1
1
1
1
1
9
Low
Note: 0: no, 1: yes, overall risk of bias: low (score >8), moderate (score 6–8), or high (score ≤5). Items scored: (1)Was the study's target population a close representation of the national population in relation to relevant variables (eg, age, sex)? (2) Was the sampling frame a true or close representation of the target population? (3) Was some form of random selection used to select the sample, OR, was a census undertaken? (4) Was the likelihood of nonresponse bias minimal? (5) Were data collected directly from the subjects (as opposed to a proxy)? (6) Was an acceptable case definition used in the study? (7) Had the study instrument that measured the parameter of interest (eg, prevalence of comorbidity) been tested for reliability and validity (if necessary)? (8)Was the same mode of data collection used for all subjects? (9) Was the length of the shortest prevalence period for the parameter of interest appropriate? (10) Were the numerator(s) and denominator(s) for the parameter of interest appropriate?
Social frailty is independently associated with geriatric depression among older adults living in northern Japan: a cross-sectional study of ORANGE registry.
The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration.
Association of social frailty with physical health, cognitive function, psychological health, and life satisfaction in community-dwelling older Koreans.
Screening value of social frailty and its association with physical frailty and disability in community-dwelling older koreans: aging study of pyeongchang rural area.
Social frailty has a stronger impact on the onset of depressive symptoms than physical frailty or cognitive impairment: a 4-year follow-up longitudinal cohort study.
Factors associated with physical, psychological and social frailty among community-dwelling older persons in Europe: a cross-sectional study of Urban Health Centres Europe (UHCE).
This work was supported by the Peking Union Medical College Hospital Research Fund (Grant number: ZC201900516) and the First Affiliated Hospital of Shenzhen University Research Fund (Grant number: 20223357018).