Global evidence on falls and subsequent social isolation in older adults: a scoping review.

BACKGROUND: Falls are a leading cause of injury-related hospitalizations among adults aged 65 years and older and may result in social isolation.
OBJECTIVE: To summarise evidence on falls and subsequent social isolation and/or loneliness in older adults through a scoping review. ELIGIBILITY CRITERIA: Studies were eligible for inclusion if the population had a mean age of 60 years or older, they examined falls and subsequent social isolation, loneliness, fear of falling or risk factors and were primary studies (eg, experimental, quasi-experimental, observational and qualitative). SOURCES OF EVIDENCE: MEDLINE, CINAHL, Embase, Ageline and grey literature from inception until 11 January 2021. CHARTING
METHODS: A screening and charting form was developed and pilot-tested. Subsequently, two reviewers screened citations and full-text articles, and charted the evidence.
RESULTS: After screening 4993 citations and 304 full-text articles, 39 studies were included in this review. Participants had a history of falling (range: 11% to 100%). Most studies were conducted in Europe (44%) and North America (33%) and were of the cross-sectional study design (66.7%), in the community (79%). Studies utilised 15 different scales. Six studies examined risk factors for social isolation and activity restriction associated with fear of falling. Six studies reported mental health outcomes related to falls and subsequent social isolation.
CONCLUSIONS: Consistency in outcome measurement is recommended, as multiple outcomes were used across the included studies. Further research is warranted in this area, given the ageing population and the importance of falls and social isolation to the health of older adults. SCOPING REVIEW REGISTRATION NUMBER: 10.17605/OSF.IO/2R8HM. © Author(s) (or their employer(s)) 2022. Re-use permitted under CC BY-NC. No commercial re-use. See rights and permissions. Published by BMJ.

A robust methodology including a thorough and extensive literature search was used to review the literature in the area.

There were no date or language limits for studies eligible for inclusion in this scoping review.

Scoping reviews do not assess the quality of included studies and we cannot confirm the directional causality between falls and social isolation.

Introduction

Addressing social isolation in older adults is a growing priority in Canada, as over 30% of older adults are at risk of social isolation.1 Social isolation among older adults is associated with adverse health outcomes including cognitive decline, depression, anxiety and dementia.2

Globally, falls are the second leading cause of unintentional injury death, making falls a major public health concern.3 In Canada, falls are the leading cause of injury-related hospitalisations among adults aged 65 years and older, and 20%–30% of older adults experience at least one fall each year.4 Falls may result in serious health-related consequences including physical (eg, fractures), physiological (eg, cognitive decline) and psychological (eg, anxiety, depressive symptoms, fear of falling and social isolation) outcomes.5

Given the detrimental outcomes associated with both falls and social isolation, there is a need to understand the relationship between falls and subsequent social isolation in older adults. The current scoping review is focused on falling and the subsequent experience of social isolation and/or loneliness and to ascertain whether the COVID-19 context affected the relationship between falls and subsequent social isolation.

Methods

Protocol and registration

The protocol for this scoping review was developed in accordance with the JBI (formerly Joanna Briggs Institute) guidance for scoping reviews and registered with Open Science Framework.6 An integrated knowledge translation approach was used,7 whereby colleagues from the Public Health Agency of Canada (YJ, KA, MD and AG-B) co-developed the review. The results are reported using the Preferred Reporting Items for Systematic reviews and Meta-analysis (PRISMA) extension to scoping reviews8 supplemented by PRISMA 2020.9

Patient and public involvement

Through the Strategy for Patient-Oriented Research (SPOR) Evidence Alliance, we collaborated closely with a patient partner who provided feedback on our protocol, participated in full-text screening piloting and provided input on the manuscript (JB).

Search strategy

An experienced librarian developed our comprehensive literature search strategy, which was peer-reviewed by a second information specialist using the Peer Review of Electronic Search Strategies (PRESS) checklist.10 MEDLINE, CINAHL, Embase and Ageline were searched from inception until 11 January 2021 (online supplemental appendix 1). References of included studies and relevant reviews were scanned. Grey literature (ie, unpublished or difficult to locate studies) was searched using the Canadian Agency for Drugs and Technologies in Health’s Grey Matters checklist.11

Eligibility criteria

The population of interest were older adults with a mean age of 60 years or older. The concept was the relationship between falls and subsequent social isolation or loneliness. As mentioned in our related systematic review on interventions for social isolation after falling, social isolation and loneliness are distinct concepts.12 Social isolation included a decrease in any of the following: number of social contacts, feeling of belonging, fulfilling relationships, engagement with others and quality of their personal network.12 We defined loneliness as ‘the unpleasant experience that occurs when a person’s network of social relations is deficient in some way, either quantitatively or qualitatively’.13 For our primary objective, the context included any community or institutional setting. For our secondary objective, we limited the context to include studies that specified their consideration of the COVID-19 pandemic. Studies including participants reporting a history of falling (ie, regardless of the proportion of the sample who fell), the role of fear of falling in this relationship, as well as any risk (eg, medication use, frailty) or protective (eg, exercise, gait or balance training) factors were considered eligible for inclusion.

Eligible study designs included primary research studies of experimental (e.g., randomised controlled trials), quasi-experimental (e.g., non-randomised controlled trials, controlled before and after studies and interrupted time series), observational (e.g., cohort studies, case–control studies and cross-sectional studies), qualitative (phenomenological, ethnography, qualitative interview, etc.) and mixed-method (e.g., convergent parallel, embedded and explanatory sequential) design. No restrictions based on study year, language of publication or study duration were applied.

Study selection

A screening form was developed and a pilot-test using 50 citations was completed with 80% agreement, and subsequently, all remaining titles and abstracts were screened independently by pairs of reviewers (SMT, AP, JF, GM and AH). Discrepancies were resolved by a third reviewer.

Similarly, two pilot-tests were completed for full-text article screening (achieving 27% and 40% agreement, respectively), screening criteria were revised, and subsequently, full-text articles were assigned to independent pairs of reviewers. Discrepancies were resolved by a third reviewer.

Data charting

A charting form was developed to capture data on study characteristics, population characteristics and outcomes of interest. Relevant outcomes included any data illustrating the relationship between falls and subsequent social isolation, including the role of fear of falling, and other risk factors or protective factors. A pilot-test was conducted using five studies, sufficient agreement was achieved, and subsequently, full data charting was completed by independent pairs of reviewers. Discrepancies were resolved by a third reviewer.

Analysis and presentation of results

The review findings were summarised descriptively using summary tables.

Results

After screening 4993 citations and 304 full-text articles against our eligibility criteria, 39 studies were identified as eligible for inclusion based on our primary objective for this review (figure 1). No studies were identified when limiting to the COVID-19 context for our secondary objective. Study and patient characteristics have been summarised in table 1 and detailed data are reported in online supplemental appendices 2 and 3.

Figure 1

PRISMA 2020 study flow diagram. PRISMA, Preferred Reporting Items for Systematic reviews and Meta-analysis.

Table 1

Summary of study and patient characteristics

Characteristics	Number (%)
Study characteristics (n=39)
Geographical region
Asia	5 (12.8)
Australia	1 (2.5)
Europe	17 (43.6)
North America	13 (33.3)
South America	3 (7.7)
Study design
Cohort	6 (13.8)
Cross-sectional	26 (66.7)
Qualitative	7 (19.4)
Study duration
Not applicable	29 (74.3)
≤ 1 year	5 (12.8)
≥ 1 year	5 (12.8)
Patient characteristics
Mean age	74.9 (range, 65.0 to 95.0)
Not reported	11 (28.2)
65.0–69.9 years	4 (10.2)
70.0–74.9 years	8 (20.5)
75.0–79.9 years	14 (35.9)
≥80.0 years	2 (5.1)
Proportion of female participants	Mean: 65.3% (range, 42.5 to 88.9)
Sample size	Mean: 3043.6 (9 to 43487)
<100	11 (28.2)
100–499	11 (28.2)
500–999	3 (7.7)
1000–1999	4 (10.2)
2000–5000	4 (10.2)
>5000	6 (15.4)
Study setting
Community	31 (79.4)
Medical	6 (15.4)
Nursing home	1 (2.5)
Multisite	1 (2.5)
Participants living alone	44.1% (range, 0 to 100)
Participants with a history of falling	Mean: 50.8% (range, 11.2 to 100)
Not reported*	11 (28.2)
≤25%	6 (15.4)
25–40%	10 (25.6)
40–85%	5 (12.8)
>85%	7 (17.9)

*Not reported for the overall sample.

Study characteristics

The publication year for included studies ranged from 1987 to 2020, with more than half published since 2010. Most studies were conducted in Europe (17/39, 44%) and North America (13/39, 33%). More than half of the studies were cross-sectional study design (66.7%) and seven qualitative studies were included. Most were conducted in the community (79%). Studies utilised 15 different scales and a variety of self reported responses to assess variables such as social isolation and loneliness (eg, 18-item Lubben Social Network Scale, 6-item de Jong-Gierveld Loneliness Scale). Six studies identified risk factors for social isolation and for activity restriction due to fear of falling (table 2). Six studies reported mental health outcomes (online supplemental appendix 4).

Table 2

Potential risk factors for social isolation and activity restriction associated with fear of falling

Author, year	Risk factor	Associated evidence
Social isolation after injurious fall
Nicholson, 200525	Sex (female)	The authors noted a strong positive correlation between injurious falls and social isolation for women (ρ=−0.5; p=0.01), but this was not significant for men.
Activity restriction due to fear of falling
Zijlstra, 2007	Aged 80 years or older	OR: 1.56 (95% CI, 1.24 to 1.95)
	Fair perceived general health	OR: 2.92 (95% CI, 2.43 to 3.52)
	Poor perceived general health	OR: 5.7 (95% CI, 3.57 to 9.12)
Curcio et al, 200932	Poor perceived health	OR: 1.38 (95% CI, 1.06 to 1.79)
	Depression	OR: 1.76 (95% CI, 1.38 to 2.24)
	Low social participation	OR: 1.52 (95% CI, 1.20 to 1.92)
	Difficulties in activities of daily living	OR: 1.65 (95% CI, 1.16 to 2.32)
	Decreased physical activity	OR: 1.35 (95% CI, 1.06 to 1.70)
	Polypharmacy	OR: 1.56 (95% CI, 1.14 to 2.14)
	Below poverty level	OR: 1.32 (95% CI, 1.05 to 1.65)
Dias, 2011	Depression	χ2=15.2, p=0.004
	Exhaustion (frailty)	χ2=9.2, p=0.01
	Participation in social activities	χ2=10.4, p=0.016
Murphy et al, 200240	Two or more chronic conditions	ARR: 1.34 (95% CI, 1.08 to 1.65)
	Slow-timed physical performance	ARR: 1.44 (95% CI, 1.18 to 1.75)
Merchant et al, 202034	Sarcopenia	OR, 8.13 (95% CI, 1.52 to 43.41)

ARR, adjusted risk ratio; OR, odds ratio.

Patient characteristics

Across all studies, the number of included patients was 118 702, with an average of 3043 patients per study. Their mean age ranged from 65 to 95 years. Approximately 65% of patients were female. Most studies included participants with a history of falling, ranging from 11% to 100% of the study population.

Cohort studies

Among the 39 included studies, six were cohort studies (online supplemental appendix 5). Tinetti and Williams14 demonstrated a significant relationship between multiple non-injurious falls and a decline in social functioning (regression coefficient=−0.538 (p<0.05)), measured using the Social Activity Scale, in a sample of 770 older adults after 3 years of follow-up.14 Similarly, Pin and Spini15 found that in their cohort of 16 583 participants, those who fell showed decreased social participation after falling (p<0.001), which was no longer statistically significant when frailty was added in the model.15

Vellas et al16 compared people who fall versus those who did not in two populations: a retirement home (n=118) and older adults living at home (n=60).16 Among the older adults who lived at home, they noted that fewer fallers were able to maintain the same level of activity after 6 months of follow-up when compared with non-fallers (p<0.02).

van der Meulen et al17 assessed social participation (using the Frenchay Activities Index) in 260 older adults with low and high levels of concern about falling over 14 months.17 They reported significant differences (specific results not reported) between the groups, with lower social participation scores in those who had a higher level of concern about falling.

In 4680 older adults, Yu et al18 reported a significant relationship between the number of falls and loneliness scores (measured using the three-item University of California, Los Angeles (UCLA) Loneliness Scale) across three time points over 4 years (B=0.008, p<0.05).18 A cohort study by Hajek and König19 looked at loneliness (as measured using the Bude and Lantermann scale) and social isolation (measured using the De Jong Gierveld Loneliness Scale) and their link to fear of falling 669 older adults.19 They compared older adults with an onset of fear of falling, to those who had no fear. Their findings revealed that the end of fear of falling was associated with lower loneliness scores (β = −0.06, p<0.05) and other negative psychosocial outcomes (eg, increased depressive symptoms).

Cross-sectional studies related to falls and social isolation

Of the 26 cross-sectional studies included in this review, 11 reported on the relationship between falls and social isolation or loneliness (online supplemental appendix 6).

Quach20 examined the relationship between falls and scores on the Social Relationship Index including 8464 participants. He noted that participants who reported experiencing a fall or multiple falls had a lower social relationship index score (mean, 3.24 and 3.08 respectively) compared with those who had not fallen (mean, 3.34; p<0.0001).

Hajek and König21 examined variables associated with a history of falling in 7808 participants. They found those reporting a fall in the previous 12 months had higher loneliness scores (De Jong Gierveld Loneliness Scale; β=0.08, p<0.001) and social exclusion scores (Bude and Lantermann scale; β=0.08, p<0.001) compared with those who had not fallen.

Schnittger et al22 conducted a study in 579 older adults identifying risk factors for different pathways of loneliness—emotional loneliness, social loneliness (both measured using the De Jong Gierveld Loneliness Scale) and social support (measured using the Lubben Social Network Scale). A history of falls was the only biological variable that was identified as a statistically significant risk factor for inclusion in the model for social support (correlation coefficient=−0.247; p<0.003).

Stel et al23 reported a statistically significant decline in social activities in 204 older adults who experienced a fall inside their home (OR: 2.6 (95% CI: 1.1 to 6.5); p<0.05), and Vanden Wyngaert et al24 reported an association between risk of falls and participation in social roles and activities in 154 older adult haemodialysis patients (PROMIS questionnaire; R2=0.11; p=0.01). Finally, Nicholson25 reported a strong positive relationship between experiencing an injurious fall and increasing social isolation in a sample of 68 older adults (Lubben Social Network Scale; ρ=−0.4; p<0.05), and highlighted that this relationship was stronger in women (ρ=−0.5; p=0.01). Additionally, they assessed this relationship using both the Family and Friends subscales of the Lubben Social Network Scale and found that the correlation was specific to the Friends subscale (ρ=−0.43; p<0.05).

Iliffe et al26 and Robins et al27 found no statistically significant associations between falls and social isolation using the Lubben Social Network Scale in a sample of 3139 older adults and the Friendship Scale for social isolation in a sample of 245 older adults, respectively. Similarly, van Lankveld et al28 and Faria et al29 found no correlation between falls and loneliness, using the De Jong Gierveld Loneliness Scale in a sample of 579 older adults, and the UCLA Scale in a sample of 48 older adults, respectively. Additionally, Finn 30 noted no difference in scores for the Older Americans Resources and Services (OARS) social support scale when comparing fallers to non-fallers in a nursing home setting (n=49).

Cross-sectional studies related to fear of falling and social isolation

Seven studies examined fear of falling linked to falls and social isolation (online supplemental appendix 7). Gagnon et al31 reported a statistically significant positive relationship between fear of falling and social support in a sample of 105 older adults (measured using the confiding-relationships component of the Bedford Life Events and Difficulties Schedule modified for elderly subjects; Wald χ2=3.77; p=0.05). Curcio et al32 reported a strong relationship between fear of falling and low social participation in 1668 older adults (OR, 1.52; 95% CI, 1.20 to 1.92; p<0.01). Petrinec et al33 (2020) identified fear of falling as an independent predictor of social functioning (as measured by the Medical Outcomes Study 36-item Short-Form General Health Survey; β=−0.29) in 108 older adults.

Merchant et al34 and Iliffe et al26 showed no statistically significant relationship between fear of falling and social isolation in 493 older adults and 3139 older adults, respectively. Ferreira et al35 and Kara et al36 showed no association between fear of falling and social participation (n=7935) or fear of falling and loneliness (n=47), respectively.35 36

Cross-sectional studies related to falls and activity restriction due to fear of falling

Eight studies examined the relationship between falls and activity restriction due to fear of falling (online supplemental appendix 7). Tinetti et al37 and Apikomonkon38 both reported a statistically significant decrease in activity due to fear of falling in individuals who experienced a fall compared with those who had not (n=1103, χ2=13.1, p<0.001; and n=546, χ2=5.49, p<0.05, respectively).37 38 Similarly, in 1668 older adults, Curcio et al32 demonstrated that those who restricted activity due to fear of falling were more likely to have experienced a fall in the year prior (OR: 1.48 (95% CI, 1.18 to 1.86); p=0.001), and Mendes da Costa et al39 demonstrated that activity restriction increased in those with multiple falls over the past year (OR, 3.04; 95% CI, 1.70 to 5.42). Murphy et al40 and Choi and Ko41 showed that a history of injurious falls was independently associated with activity restriction due to fear of falling (n=1064, adjusted risk ratio: 1.36; 95% CI, 1.11 to 1.66; p=0.003; and n=4247, OR, 3.03; 95% CI, 1.21 to 7.54, p=0.008, respectively).

Howland et al42 reported no relationship between the experience of a fall and activity restriction in a sample of 266 older adults (OR: 1.094; 95% CI, 0.376 to 3.177; p=0.869), as did Choi and Ko41 (OR, 2.12; 95% CI, 0.96 to 4.67; p=0.062) among 4247 older adults. Similarly, Merchant et al34 (2020) also reported no significant relationship between the number of falls and fear-based activity restriction in 493 older adults (OR, 1.4; 95% CI, 0.94 to 2.20).

Qualitative studies

Seven qualitative studies were included (online supplemental appendix 8). All participants interviewed were older adults (n=124), including 51 stroke survivors43 44 and 10 experiencing frailty.45 Common categories identified across these studies were activity restriction to manage fear of falling, changing behaviours to avoid falling,43 45–47 feeling restricted due to reduced mobility after falling,43 44 48 increasing dependence on caregivers,43 45 developing fear of falling,43 45 feelings of loneliness or isolation43 48 and a negative impact on identity or autonomy.47

Discussion

We conducted a comprehensive scoping review including 39 studies examining the relationship between falls and subsequent social isolation. We limited the scoping review to studies that identified social isolation after a fall, this was due to the request of the commissioning knowledge user. More than half of the studies were published since 2010, suggesting increased interest in the relationship between falls and social isolation in older adults. Social isolation and loneliness were measured using a variety of outcome measures across studies, such as degree of activity, varying scales for loneliness, social isolation, social participation and social support. This highlights the growing need for consistency in the measurement of social isolation and loneliness to allow for meaningful comparison across studies. Cornwall and Waite49 highlight previous efforts to consolidate different measures of social isolation and build off this work. They combined multiple measures of social isolation to develop two scales that measure distinct dimensions of social isolation—social disconnectedness and perceived isolation.

Only a few studies examined risk factors and mental health outcomes related to falls and subsequent social isolation. Risk factors linked to social isolation and activity restriction included age, sex/gender, poor perceived health, poverty, frailty and comorbidity. Few studies also documented an association between activity restriction due to fear of falling and depression. Our findings suggest the presence of gaps in the literature for these important outcomes, highlighting the need for further research. No randomised trials exploring interventions for social isolation after a fall were identified in our scoping review, highlighting another gap in the literature and an area for future research to explore.

We did not identify any studies on falls and subsequent social isolation that were specific to the COVID-19 context, highlighting another gap in the evidence base. A scoping review by Kasar and Karaman50 suggest that older adults face increased social isolation as a result of pandemic-related restrictions, which can result in increased loneliness and reduced quality of life. They also highlighted how technology can be used to deliver virtual or tele-health support services, and to allow older adults stay connected with their social networks.50 A systematic review by Larson et al assessed the impact of COVID-19 lockdowns on physical activity in older adults and reported that most studies demonstrated a decline in physical activity or an increase in sedentary behaviours in this population. The effectiveness of physical activity and exercise in preventing falls and fractures in older adults is well established in the literature.51–53 A decline in physical activity in older adults could lead to sarcopenia, and an increased risk of falls or fractures.53

There are several strengths to our scoping review, such as the use of the JBI guide, and the PRISMA-extension for scoping reviews. A comprehensive literature search was conduced and several different types of study designs were included. However, limitations include that all studies were conducted in middle–high-income or high-income economy countries. This suggests that our results may not be generalisable to low-income and middle-income countries, highlighting a gap in the literature. Many of the included studies were cross-sectional and we cannot confirm the directional causality between falls and social isolation without more robust research. Furthermore, none of the included studies specifically focused on culturally and linguistically diverse backgrounds, who might be at greater risk of social isolation after experiencing a fall. Additional research is warranted in this area.54 In addition, we were unable to update the literature search due to lack of capacity and funding.

In summary, we found a dearth of research, particularly examining risk factors and mental health outcomes related to social isolation and falling in older adults. Further research is warranted in this area, given the importance of falls and social isolation to the health of older adults.