Abstract
We investigated the association between perceived neighbourhood characteristics and falls in community-dwelling adults, using data from Wave 5 and 6 of the Survey of Health, Ageing and Retirement in Europe (SHARE). We included 25,467 participants aged 50 to 103 years (mean age 66.2 ± 9.6, 58.5% women), from fourteen European countries (Austria, Belgium, Czech Republic, Denmark, Estonia, France, Germany, Israel, Italy, Luxembourg, Slovenia, Spain, Sweden, Switzerland). At baseline, we recorded individual-level factors (socio-demographic, socio-economic and clinical factors), contextual-level factors (country, urban versus rural area, European region) and perceived neighbourhood characteristics (vandalism or crime, cleanliness, feeling part of neighbourhood, helpful neighbours, accessibility to services) for each participant. We recorded falls in the six months prior to the baseline and 2-year follow-up interviews. The associations between neighbourhood characteristics and falls were analysed by binary logistic regression models; odds ratios (95% confidence intervals) were calculated. Participants reporting-versus not reporting-vandalism or crime had an increased falls risk of 1.16 (1.02–1.31) at follow-up, after full adjustment; lack of cleanliness, feeling part of the neighbourhood, perceiving neighbours as helpful and difficult accessibility to services were not associated with falls. Vandalism or crime was consistently associated with increased falls risks in women, adults without functional impairment and urban areas residents. In conclusion, adverse neighbourhood environments may account for inequality in falls risk among middle-aged and older adults and could be added to fall risk stratification tools.
Supplementary Information
The online version contains supplementary material available at 10.1007/s10433-022-00685-3.
Keywords: Population-based cohort study, Falls, Physical disorder, Neighbourhood social cohesiveness, Safety
Introduction
Falls are a major public health issue (Tinetti et al. 1988; Rubenstein 2006; Kelsey et al. 2012; Li et al. 2006; Li et al. 2014; Nevitt et al. 1989), affecting one in three community-dwelling older adults yearly and frequently causing major injuries (Rubenstein 2006; Kelsey et al. 2012). Falls result in increased morbidity, mortality, institutionalization and costs (Nevitt et al. 1989; Stevens et al. 2006; Tinetti and Williams 1997). Preventing falls by addressing their environmental risk factors may benefit many adults in the long term (Diez Roux 2002; Diez Roux and Mair 2010).
Neighbourhood physical, social and service environments influence health (Diez Roux 2002; Diez Roux and Mair 2010; Pickett and Pearl 2001). Neighbourhood deprivation is associated with increased mortality (Pickett and Pearl 2001), while better neighbourhoods are linked to lower incidence of chronic diseases, including diabetes (Christine et al. 2015), obesity (Auchincloss et al. 2013) and hypertension (Kaiser et al. 2016). Perceived neighbourhood safety is associated with better health (Lovasi et al. 2014) and more physical exercise (Evenson et al. 2012; Shenassa et al. 2006), while perceived disorder associates with poorer health and depression (Diez Roux and Mair 2010; Vaz et al. 2020). Fear of crime leads to psychological stress, limitations on physical activity and social isolation (Stafford et al. 2007; Piro et al. 2006; Lorenc et al. 2013). Of note, fear of crime is weakly associated with objective measures of crime, while triggered by signs of neglect in the neighbourhood and buffered by social cohesiveness (Lorenc et al. 2013). Furthermore, difficult accessibility to services may affect health by limiting check-ups, access to medication and healthy food.
Few studies explored the relationship between neighbourhood and falls (Nicklett et al. 2017; Lee et al. 2019; Lo et al. 2016; do Nascimento et al. 2017). Yet, the neighbourhood may affect falls risk in various ways. First, physical neighbourhood characteristics may directly affect it. About half of the falls in community-dwelling adults occur outdoors (Kelsey et al. 2012; Li et al. 2006, 2014) and may be precipitated by environmental factors, such as tripping or slipping on objects (Li et al. 2006). Indeed, older adults perceive uneven walking surfaces, inadequate street maintenance, poor lighting and unsafe traffic patterns as falls risk factors (Chippendale and Boltz 2015). Second, fear of crime and lack of social cohesiveness may discourage outdoor physical activity, leading to deconditioning (Lorenc et al. 2013). Indeed, many outdoor falls occur in adults spending limited time outdoors (Li et al. 2014). Finally, adults with negative perceptions of their neighbourhood may have chronic stress response activation, poorer strength and performance, compared to those with positive perceptions (Balfour and Kaplan 2002).
Furthermore, sex, age and functional status may modulate the relationship between neighbourhood and falls (Diez Roux 2002; Stafford et al. 2005). Women express fear of crime more frequently than men and may be more affected by it (Lovasi et al. 2014; Stafford et al. 2007; Piro et al. 2006; Lorenc et al. 2013). Older, frailer adults may spend more time in their neighbourhood and rely more on its resources, compared to middle-aged and fit adults; as a result, older frailer adults may be more vulnerable to adverse neighbourhood environments than middle-aged fit adults (Diez Roux 2002).
This study aims to explore the associations between perceived neighbourhood characteristics—vandalism or crime, cleanliness, social cohesiveness, service accessibility—and falls in community-dwelling adults in the Survey of Health, Ageing and Retirement in Europe (SHARE), at baseline and follow-up.
Methods
Study design
We used data from Wave 5 and 6 of SHARE, a multidisciplinary, cross-national, longitudinal survey of ageing processes in adults (Börsch-Supan et al. 2013; Börsch-Supan 2020a, b; Börsch-Supan 2020b). Started in 2004, SHARE comprises biennial waves; it obtained ethical approval by the University of Mannheim and the Max Planck Society (Börsch-Supan 2020a, b; Börsch-Supan 2020b). The Netherlands did not participate in Wave 6.
Figure 1 shows the study flow chart. At baseline (Wave 5), 66,188 participants had a main interview. We excluded those younger than 50 years or of unknown age (n = 1,185), those in the Netherlands (n = 4,118) or in a nursing home or unknown (n = 655), those who were not direct respondents on neighbourhood or co-variates (n = 22,825) and those with missing data on either falls (n = 2), neighbourhood (n = 1,076) or co-variates (n = 4,346). At follow-up (Wave 6), we further excluded 6,514 participants without a main interview or data on falls.
Fig. 1.
Flow-chart of study inclusion criteria; abbreviations: n: number; SHARE: Survey of Health, Ageing and Retirement in Europe; the 14 countries were: Austria, Belgium, Czech Republic, Denmark, Estonia, France, Germany, Israel, Italy, Luxembourg, Slovenia, Spain, Sweden and Switzerland
Therefore, this study includes 25,467 participants, aged ≥ 50 years at baseline, with complete data. They were residents from 14 countries, which we classified into three regions: Northern and Western Europe: Austria, Belgium, Denmark, France, Germany, Luxembourg, Sweden and Switzerland; Southern Europe: Israel, Italy and Spain; Eastern Europe: Czech Republic, Estonia and Slovenia. This classification reflects both the North–South and East–West gradients in health and socio-economic conditions across SHARE countries (Börsch-Supan et al. 2013).
Those participants lost to follow-up were more likely to be men and have previous falls compared to those included (supplementary Table S1). Those participants who were lost to follow-up were more likely to report difficult accessibility to services in their neighbourhood, not feeling part of their neighbourhood and not feeling that their neighbours would be helpful, compared to those included (supplementary Table S1). No differences in perceived vandalism or crime or cleanliness of the neighbourhood were found between participants lost to follow-up and those included (supplementary Table S1).
Individual-level co-variates
At baseline, age, education, self-rated health, co-morbidities, medications, falls, dizziness, physical inactivity, living alone, “difficulty in making ends meet”, vision, hearing, grip strength and functional limitations were recorded.
Educational level was classified based on the International Standard Classification of Education (ISCED)-97 (OECD, 1999), as follows: no or pre-primary; primary; lower secondary; upper secondary; post-secondary non-tertiary; first stage of tertiary and second stage of tertiary education. We dichotomised self-rated health as good (“excellent”, “very good” or “good”) versus poor (“fair” or “poor”). Based on previous literature (Deandrea et al. 2010), we selected these self-reported co-morbidities: heart attack, hypertension, high cholesterol, stroke, diabetes, chronic lung disease, cancer, Parkinson’s, cataracts, hip fracture, other fractures, cognitive impairment, affective/emotional disorder and any arthritis (rheumatoid arthritis and osteoarthritis/other rheumatism). Similarly, we selected these self-reported medications: anti-hypertensives, drugs for pain (drugs for joint and other pain) and psychotropic drugs (drugs for sleep, anxiety or depression).
At baseline, participants were asked about falls and dizziness in the previous six months, with the question: “For the past six months at least, have you been bothered by any of the health conditions on this card?” options included: “falling down” and “dizziness, faints or blackouts”. Physical inactivity was “never vigorous nor moderate physical activity”. We classified participants as those reporting difficulty in making ends meet (“some” or “great difficulty”) versus no difficulty (“easily” or “fairly easily”).
We classified self-reported vision as good (distant and close vision were “excellent”, “very good” or “good”) versus poor (distant or close vision were “fair” or “poor”). Similarly, we classified self-reported hearing as good versus poor.
As previously described (Mohd Hairi et al. 2010), hand-grip strength was measured using a handheld dynamometer (Smedley, S Dynamometer, TTM, Tokyo, 100 kg). Participants were then instructed to squeeze the dynamometer as hard as possible. Two values were recorded for each hand. Measurements were considered valid if the two measurements of one hand differed by less than 20 kg. Values of zero or those above 100 kg were considered invalid. We retrieved maximum grip strength for each participant.
Participants self-reported functional limitations in activities of daily living (ADL) (Katz et al. 1963)-dressing, walking across a room, bathing, eating, getting in or out of bed, using the toilet-and instrumental activities of daily living (IADL) (Lawton and Brody 1969)-using a map, preparing meals, shopping, phoning, taking medications, doing housework, managing money, using transport, doing laundry. We categorized the participants as those with functional impairment (one or more ADL or IADL limitation) versus those without functional impairment.
Contextual-level co-variates
We created a dummy variable for each of the 14 countries of residence. The interviewers classified areas of residence as urban (“big city”, “the suburbs or outskirts of a big city”, “a large town” or “a small town”) versus rural (“a rural area or village”). As mentioned before, we defined three European regions.
Neighbourhood
At baseline, the participants were asked whether they agreed with these statements: “vandalism or crime is a big problem in this area”, “this area is kept very clean”, “I really feel part of this area” and “If I were in trouble, there are people in this area who would help me”. We dichotomised the answers as “agree” (“strongly agree” or “agree”) versus “disagree” (“disagree” or “strongly disagree”).”This area” was “everywhere within a 20 min walk or a kilometre” of the participant’s home.
Participants were asked how easy it was to get to four services (bank, grocery store, general practitioner, pharmacy). We dichotomised service accessibility as easy (“very easy” or “easy” to all services) versus difficult (“difficult” or “very difficult” to one or more service).
We identified five negative neighbourhood characteristics: vandalism or crime; lack of cleanliness; not feeling part; not perceiving neighbours as helpful and difficult service accessibility. We refer to “feeling part” and “perceiving neighbours as helpful” as neighbourhood social cohesiveness.
Falls
At follow-up, participants were again asked about falls in the six months preceding the interview. We dichotomised falls as no falls versus one or more falls.
Statistical analyses
We tested for differences in baseline characteristics between sexes and between participants with negative and positive perceptions of their neighbourhood (for each characteristic), using Pearson’s chi-square test for categorical variables and t test for age (continuous variable).
We used binary logistic regression models to assess the associations between each neighbourhood characteristic at baseline (determinants) and falls at baseline and follow-up, respectively (outcomes). We calculated odds ratios (95% confidence intervals). We performed our analyses in four steps. First, we adjusted our analyses for age and sex (Model 1); then, we further adjusted for education, self-rated health, co-morbidities, medications, living alone, difficulty in making ends meet, vision, hearing, country, area of residence (urban versus rural) and functional status (Model 2); then, we further adjusted for physical inactivity and maximum grip strength (Model 3); finally, we further adjusted for dizziness, faints or blackouts and, in analyses at follow-up, falls at baseline (Model 4). We entered each co-morbidity and medication variable in Model 2, 3 and 4, separately. We chose to explore the associations between neighbourhood characteristics and falls not only at baseline, but also at follow-up, so that perceptions of neighbourhood preceded the falls (to exclude reverse causation, i.e. that the neighbourhood perception may worsen among the fallers because of falling).
We performed sex-stratified analyses, as previous literature suggested that women may be more likely to report vandalism or crime and more negatively impacted by fear of crime, than men (Lorenc et al. 2013).We performed sensitivity analyses by stratifying the participants according to: (1) age groups (50 to 64 years; ≥ 65 years); (2) functional status; (3) area (urban; rural) and (4) European regions. These sensitivity analyses were based on previous literature suggesting that: (1) older adults and adults with functional impairment may be more vulnerable to adverse neighbourhood effects than middle-aged and fitter adults (Diez Roux 2002) and (2) the prevalence and impact of neighbourhood characteristics may vary by area and European region (World Health Organization 2016). We tested for interaction between each neighbourhood characteristics and (1) sex, (2) age (used as a continuous variable), (3) functional status and (4) area, respectively, in the association between each neighbourhood characteristic and falls risk at follow-up.
Furthermore, we performed sensitivity analyses at follow-up in the sample of participants: (1) not changing residence—and thus living in the same neighbourhood—during follow-up; and (2) not reporting falls at baseline.
Our Models of adjustment included all neighbourhood characteristics simultaneously. Furthermore, we repeated all analyses using alternative Models of adjustment by including each neighbourhood characteristic at a time, to facilitate comparisons with previous studies that may include a few but not all these neighbourhood characteristics. All analyses were performed using SPSS 25.
Results
Baseline interviews were in 2013. Our study population included 25,467 participants (14,902 women), aged 50 to 103 years at baseline (mean age 66.2 ± SD 9.6). About two thirds of participants lived in urban areas (n = 17,164) and one third in rural areas (n = 8,303). Vandalism or crime was reported by 4,540 (17.8%) participants, lack of cleanliness by 3,469 (13.6%), not feeling part of neighbourhood by 1,647 (6.5%), not perceiving neighbours as helpful by 3,013 (11.8%) and difficult accessibility to services by 5,954 (23.4%). Women were more likely to report vandalism or crime, lack of cleanliness, not feeling part of neighbourhood and difficult accessibility to services, than men (Table 1).
Table 1.
Characteristics of total study population at baseline and stratified by sex
| All (n = 25,467) | Men (n = 10,565) | Women (n = 14,902) | P value | |
|---|---|---|---|---|
| Age (years), mean (SD) | 66.2 (9.6) | 66.2 (9.4) | 66.3 (9.7) | 0.540 |
| Neighbourhood characteristics, n (%) | ||||
| Vandalism or crime | 4,540 (17.8) | 1,782 (16.9) | 2,758 (18.5) | 0.001 |
| Clean area | 21,998 (86.4) | 9,281 (87.8) | 12,717 (85.3) | < 0.001 |
| Feeling part | 23,820 (93.5) | 9,939 (94.1) | 13,881 (93.1) | 0.003 |
| Helpful neighbours | 22,454 (88.2) | 9,276 (87.8) | 13,178 (88.4) | 0.124 |
| Easy accessibility to services | 19,513 (76.6) | 8,450 (80.0) | 11,063 (74.2) | < 0.001 |
| Educational level, n (%) | ||||
| No or pre-primary | 933 (3.7) | 374 (3.5) | 559 (3.8) | < 0.001 |
| Primary | 4,046 (15.9) | 1,465 (13.9) | 2,581 (17.3) | |
| Lower secondary | 4,299 (16.9) | 1,569 (14.9) | 2,730 (18.3) | |
| Upper secondary | 8,686 (34.1) | 3,760 (35.6) | 4,926 (33.1) | |
| Post-secondary non-tertiary | 1,261 (5.0) | 520 (4.9) | 741 (5.0) | |
| First stage of tertiary | 6,019 (23.6) | 2,744 (26.0) | 3,275 (22.0) | |
| Second stage of tertiary | 223 (0.9) | 133 (1.3) | 90 (0.6) | |
| Self-rated health, n (%) | ||||
| Good | 16,684 (65.5) | 7,269 (68.8) | 9,415 (63.2) | < 0.001 |
| Poor | 8,783 (34.5) | 3,296 (31.2) | 5,487 (36.8) | |
| Co-morbidities, n (%) | ||||
| Heart attack | 2,697 (10.6) | 1,337 (12.7) | 1,360 (9.1) | < 0.001 |
| Hypertension | 10,256 (40.3) | 4,126 (39.1) | 6,130 (41.1) | 0.001 |
| High cholesterol | 6,000 (23.6) | 2,407 (22.8) | 3,593 (24.1) | 0.014 |
| Stroke | 823 (3.2) | 368 (3.5) | 455 (3.1) | 0.056 |
| Diabetes | 3,126 (12.3) | 1,469 (13.9) | 1,657 (11.1) | < 0.001 |
| Chronic lung disease | 1,498 (5.9) | 658 (6.2) | 840 (5.6) | 0.048 |
| Cancer | 1,285 (5.0) | 529 (5.0) | 756 (5.1) | 0.812 |
| Parkinson’s disease | 150 (0.6) | 70 (0.7) | 80 (0.5) | 0.196 |
| Cataracts | 2,182 (8.6) | 782 (7.4) | 1,400 (9.4) | < 0.001 |
| Hip fracture | 397 (1.6) | 150 (1.4) | 247 (1.7) | 0.131 |
| Other fractures | 1,525 (6.0) | 603 (5.7) | 922 (6.2) | 0.112 |
| Cognitive impairment | 128 (0.5) | 50 (0.5) | 78 (0.5) | 0.577 |
| Affective/emotional | 1,495 (5.9) | 398 (3.8) | 1,097 (7.4) | < 0.001 |
| Any arthritis | 6,340 (24.9) | 1,828 (17.3) | 4,512 (30.3) | < 0.001 |
| Drugs, n (%) | ||||
| Anti-hypertensives | 10,859 (42.6) | 4,458 (42.2) | 6,401 (43.0) | 0.228 |
| Drugs for pain | 5,920 (23.2) | 1,689 (16.0) | 4,231 (28.4) | < 0.001 |
| Psychotropic drugs | 3,062 (12.0) | 830 (7.9) | 2,232 (15.0) | < 0.001 |
| Falls at baseline, n (%) | 1,719 (6.7) | 452 (4.3) | 1,267 (8.5) | < 0.001 |
| Dizziness, faints or blackouts, n (%) | 3,409 (13.4) | 1,054 (10.0) | 2,355 (15.8) | < 0.001 |
| Physical inactivity, n (%) | 1,961 (7.7) | 677 (6.4) | 1,284 (8.6) | < 0.001 |
| Lives alone, n (%) | 7,592 (29.8) | 2,287 (21.6) | 5,305 (35.6) | < 0.001 |
| Difficulty in making ends meet, n (%) | 8,271 (32.5) | 2,976 (28.2) | 5,295 (35.5) | < 0.001 |
| Poor vision, n (%) | 6,393 (25.1) | 2,466 (23.3) | 3,927 (26.4) | < 0.001 |
| Poor hearing, n (%) | 4,734 (18.6) | 2,296 (21.7) | 2,438 (16.4) | < 0.001 |
| Maximum grip strength (Kg/m2), mean (SD) | 33.3 (11.7) | 43.2 (9.9) | 26.3 (6.8) | < 0.001 |
| Area, n (%) | ||||
| Urban | 17,164 (67.4) | 7,008 (66.3) | 10,156 (68.2) | 0.002 |
| Rural | 8,303 (32.6) | 3,557 (33.7) | 4,746 (31.8) | |
| Functional impairment, n (%) | 4,195 (16.5) | 1,347 (12.7) | 2,848 (19.1) | < 0.001 |
P values are calculated by Pearson’s chi-square test for categorical variables or student T test for continuous variables
Educational level was categorised based on ISCED-97
n number, SD standard deviation
Participants perceiving vandalism or crime, versus those not perceiving it, were more likely to be women, have lower grip strength and report poor self-rated health, most co-morbidities, use of medications, physical inactivity, living alone, living in urban areas, difficulty in making ends meet, sensory and functional impairment (Table 2). Likewise, the other negative neighbourhood characteristics were associated with poor self-rated health, many co-morbidities, medications, difficulty in making ends meet, lower grip strength, sensory and functional impairment (supplementary Tables S2, S3, S4 and S5).
Table 2.
Characteristics of study population at baseline stratified by perceived vandalism or crime in the neighbourhood
| Vandalism or crime | P value | ||
|---|---|---|---|
| No | Yes | ||
| (n = 20,927) | (n = 4,540) | ||
| Age (years), mean (SD) | 66.2 (9.6) | 66.4 (9.2) | 0.295 |
| Women, n (%) | 12,144 (58.0) | 2,758 (60.7) | 0.001 |
| Educational level, n (%) | |||
| No or pre-primary | 762 (3.6) | 171 (3.8) | < 0.001 |
| Primary | 3,222 (15.4) | 824 (18.1) | |
| Lower secondary | 3,420 (16.3) | 879 (19.4) | |
| Upper secondary | 7,180 (34.3) | 1,506 (33.2) | |
| Post-secondary non-tertiary | 1,016 (4.9) | 245 (5.4) | |
| First stage of tertiary | 5,150 (24.6) | 869 (19.1) | |
| Second stage of tertiary | 177 (0.8) | 46 (1.0) | |
| Self-rated health, n (%) | |||
| Good | 14,031 (67.0) | 2,653 (58.4) | < 0.001 |
| Poor | 6,896 (33.0) | 1,887 (41.6) | |
| Co-morbidities, n (%) | |||
| Heart attack | 2,149 (10.3) | 548 (12.1) | < 0.001 |
| Hypertension | 8,288 (39.6) | 1,968 (43.3) | < 0.001 |
| High cholesterol | 4,847 (23.2) | 1,153 (25.4) | 0.001 |
| Stroke | 639 (3.1) | 184 (4.1) | 0.001 |
| Diabetes | 2,456 (11.7) | 670 (14.8) | < 0.001 |
| Chronic lung disease | 1,147 (5.5) | 351 (7.7) | < 0.001 |
| Cancer | 1,027 (4.9) | 258 (5.7) | 0.031 |
| Parkinson’s disease | 120 (0.6) | 30 (0.7) | 0.486 |
| Cataracts | 1,717 (8.2) | 465 (10.2) | < 0.001 |
| Hip fracture | 324 (1.5) | 73 (1.6) | 0.769 |
| Other fractures | 1,196 (5.7) | 329 (7.2) | < 0.001 |
| Cognitive impairment | 104 (0.5) | 24 (0.5) | 0.784 |
| Affective/emotional | 1,187 (5.7) | 308 (6.8) | 0.004 |
| Any arthritis | 5,000 (23.9) | 1,340 (29.5) | < 0.001 |
| Drugs, n (%) | |||
| Anti-hypertensives | 8,734 (41.7) | 2,125 (46.8) | < 0.001 |
| Drugs for pain | 4,646 (22.2) | 1,274 (28.1) | < 0.001 |
| Psychotropic drugs | 2,396 (11.4) | 666 (14.7) | < 0.001 |
| Falls at baseline, n (%) | 1,280 (6.1) | 439 (9.7) | < 0.001 |
| Dizziness, faints or blackouts, n (%) | 2,582 (12.3) | 827 (18.2) | < 0.001 |
| Physical inactivity, n (%) | 1,520 (7.3) | 441 (9.7) | < 0.001 |
| Lives alone, n (%) | 6,052 (28.9) | 1,540 (33.9) | < 0.001 |
| Difficulty in making ends meet, n (%) | 6,363 (30.4) | 1,908 (42.0) | < 0.001 |
| Area, n (%) | |||
| Urban | 13,708 (65.5) | 3,456 (76.1) | < 0.001 |
| Rural | 7,219 (34.5) | 1,084 (23.9) | |
| Poor vision, n (%) | 5,098 (24.4) | 1,295 (28.5) | < 0.001 |
| Poor hearing, n (%) | 3,833 (18.3) | 901 (19.8) | 0.016 |
| Maximum grip strength, mean (SD) | 33.5 (11.8) | 32.2 (11.4) | < 0.001 |
| Functional impairment, n (%) | 3,326 (15.9) | 869 (19.1) | < 0.001 |
P values are calculated by Pearson’s chi-square test for categorical variables or student T test for continuous variables
n number, SD standard deviation
Participants in urban areas were more likely to report vandalism or crime, lack of cleanliness and lack of social cohesiveness, but less likely to report difficult accessibility to services, compared to those in rural areas (supplementary Table S6). Variations in neighbourhood characteristics, especially service accessibility, were observed across European regions (supplementary Table S7) and countries (supplementary Table S8).
Cross-sectional analyses
At baseline, 1,719 (6.7%) participants—452 (4.3%) men and 1,267 (8.5%) women—reported falls.
In age- and sex- and neighbourhood-adjusted analyses, each negative neighbourhood characteristic, versus the positive, was associated with an increased falls risk (supplementary Table S9). After full adjustment, vandalism or crime remained associated with an increased falls risk in the total population, in men (supplementary Table S10) and in women (supplementary Table S11); lack of cleanliness was associated with an increased falls risk among men (supplementary table S10); difficult accessibility to services was associated with an increased falls risk in the total population and among men (supplementary Tables S9 and S10).
Analyses at follow-up
Follow-up interviews were in 2015. Mean follow-up was 23.6 (± SD 3.4) months. At follow-up, 1,865 (7.3%) participants—535 (5.1%) men and 1,330 (8.9%) women—reported falls.
In age- and sex- and neighbourhood-adjusted analyses, vandalism or crime, lack of cleanliness, not feeling part of the neighbourhood and difficult accessibility to services were all associated with an increased falls risk at follow-up (Table 3).
Table 3.
Association between neighbourhood characteristics and falls risk at 2-year follow-up (all participants, n = 25,467)
| n of falls | Model 1 All (n = 25,467) | Model 2 All (n = 25,467) | Model 3 All (n = 25,467) | Model 4 All (n = 25,467) | |||||
|---|---|---|---|---|---|---|---|---|---|
| OR [95% CI] | P value | OR [95% CI] | P value | OR [95% CI] | P value | OR [95% CI] | P value | ||
| Vandalism or crime | |||||||||
| No | 1,422 | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | ||||
| Yes | 443 | 1.38 [1.22; 1.55] | < 0.001 | 1.19 [1.05; 1.35] | 0.006 | 1.19 [1.05; 1.35] | 0.006 | 1.16 [1.02; 1.31] | 0.027 |
| Clean area | |||||||||
| Yes | 1,549 | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | ||||
| No | 316 | 1.22 [1.07; 1.40] | 0.004 | 1.10 [0.96; 1.27] | 0.174 | 1.11 [0.96; 1.28] | 0.158 | 1.08 [0.94; 1.25] | 0.275 |
| Feeling part | |||||||||
| Yes | 1,706 | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | ||||
| No | 159 | 1.28 [1.07; 1.54] | 0.007 | 1.06 [0.88; 1.29] | 0.524 | 1.05 [0.87; 1.27] | 0.595 | 1.06 [0.87; 1.28] | 0.588 |
| Helpful neighbours | |||||||||
| Yes | 1,597 | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | ||||
| No | 268 | 1.15 [0.99; 1.33] | 0.061 | 1.04 [0.90; 1.21] | 0.578 | 1.04 [0.89; 1.20] | 0.654 | 1.02 [0.87; 1.18] | 0.839 |
| Easy accessibility | |||||||||
| Yes | 1,204 | 1 (ref) | 1 (ref) | 1 (ref) | 1 (ref) | ||||
| No | 661 | 1.49 [1.35; 1.66] | < 0.001 | 1.18 [1.05; 1.32] | 0.005 | 1.15 [1.02; 1.28] | 0.021 | 1.12 [1.00; 1.26] | 0.060 |
Odds ratios and 95% confidence intervals are calculated by binary logistic regression. Analyses were adjusted for: Model 1: age, sex, vandalism or crime, clean area, feeling part, helpful neighbours and easy accessibility to services; Model 2: all co-variates of Model 1 and education, self-rated health, heart attack, hypertension, high cholesterol, stroke, diabetes, chronic lung disease, cancer, Parkinson’s, cataracts, hip fracture, other fractures, cognitive impairment, affective/emotional disorder, any arthritis, anti-hypertensives, drugs for pain, psychotropic drugs, living alone, difficulty in making ends meet, poor vision, poor hearing, country dummies, area (rural versus urban), functional status; Model 3: all co-variates of Model 2 and physical activity and maximum grip strength; Model 4: all co-variates of Model 3 and falls at baseline and dizziness, faints or blackout at baseline. Total number of participants reporting falls at follow-up: all 1,865
n number, OR odds ratios, CI confidence intervals, ref reference
After full adjustment, participants perceiving vandalism or crime had an increased falls risk of 1.16 (1.02–1.31), compared to those not perceiving these (Table 3). In contrast, lack of cleanliness, feeling part of the neighbourhood or that neighbours would be helpful and difficult accessibility to services, respectively, were not associated with falls risk. After full adjustment, vandalism or crime was associated with an increased falls risk among women, while not in men; no sex-interaction was observed (supplementary Tables S12-S14 and Fig. 2).
Fig. 2.
Associations between neighbourhood characteristics and falls risk at 2-year follow-up: stratified analyses. Odds ratios and 95% confidence intervals are calculated by binary logistic regression; analyses were adjusted for (Model 4): age, sex, vandalism or crime, clean area, feeling part, helpful neighbours, easy accessibility to services, education, self-rated health, heart attack, hypertension, high cholesterol, stroke, diabetes, chronic lung disease, cancer, Parkinson’s, cataracts, hip fracture, other fractures, cognitive impairment, affective/emotional disorder, any arthritis, anti-hypertensives, drugs for pain, psychotropic drugs, living alone, difficulty in making ends meet, poor vision, poor hearing, country dummies, area (rural versus urban), functional status, physical activity, maximum grip strength; falls at baseline and dizziness, faints or blackout at baseline. Panel A corresponds to supplementary Table S14; panel B to supplementary Table S16; panel C to supplementary Table S17; panel D to Table 4
The associations between co-variates and falls risk at follow-up are shown in supplementary Tables S15. In fully adjusted analyses, older age, poorer self-reported health, falls at baseline, dizziness, faints or blackouts at baseline and functional impairment were all associated with an increased falls risk at follow-up. In contrast, living with someone and greater maximum grip strength were associated with a decreased falls risk at follow-up.
Stratified analyses at follow-up
After full adjustment, no association between neighbourhood and falls was observed among participants aged 50 to 64 years, at follow-up; vandalism or crime tended to be associated with an increased falls risk among participants aged ≥ 65 years (supplementary Table S16). Age tended to modify the association between perceived vandalism or crime and falls risk at follow-up (p for interaction = 0.102, supplementary table S16 and Fig. 2).
After full adjustment, vandalism or crime was associated with an increased falls risk, at follow-up, among participants without functional impairment (supplementary Table S17). Functional status did not modify the associations between neighbourhood characteristics and falls risk at follow-up (all p for interaction > 0.05, supplementary Table S17 and Fig. 2).
In urban areas, vandalism or crime and difficult accessibility to services were associated with an increased falls risk, at follow-up, after full adjustment, while the other neighbourhood characteristics were not (Table 4). In rural areas, lack of cleanliness was prospectively associated with an increased falls risk, after full adjustment (Table 4). The area (urban versus rural) modified the association between cleanliness and falls risk at follow-up (p for interaction = 0.036, after full adjustment, Table 4 and Fig. 2).
Table 4.
Association between neighbourhood characteristics and falls risk at 2-year follow-up by area (fully adjusted)
| Urban area residents (n = 17,164) | Rural area residents (n = 8,303) | P for interaction | |||||||
|---|---|---|---|---|---|---|---|---|---|
| n of pts | n of falls | OR [95% CI] | P value | n of pts | n of falls | OR [95% CI] | P value | ||
| Vandalism or crime | |||||||||
| No | 13,708 | 959 | 1 (ref) | 7,219 | 463 | 1 (ref) | |||
| Yes | 3,456 | 361 | 1.23 [1.06; 1.42] | 0.006 | 1,084 | 82 | 0.97 [0.74; 1.28] | 0.842 | 0.214 |
| Clean area | |||||||||
| Yes | 14,445 | 1,076 | 1 (ref) | 7,553 | 473 | 1 (ref) | |||
| No | 2,719 | 244 | 0.98 [0.83; 1.16] | 0.846 | 750 | 72 | 1.48 [1.11; 1.98] | 0.008 | 0.036 |
| Feeling part | |||||||||
| Yes | 15,963 | 1,206 | 1 (ref) | 7,857 | 500 | 1 (ref) | |||
| No | 1,201 | 114 | 1.01 [0.81; 1.28] | 0.906 | 446 | 45 | 1.15 [0.81; 1.65] | 0.438 | 0.394 |
| Helpful neighbours | |||||||||
| Yes | 14,890 | 1,119 | 1 (ref) | 7,564 | 478 | 1 (ref) | |||
| No | 2,274 | 201 | 1.00 [0.83; 1.19] | 0.959 | 739 | 67 | 1.10 [0.81; 1.49] | 0.550 | 0.438 |
| Easy accessibility | |||||||||
| Yes | 13,986 | 900 | 1 (ref) | 5,527 | 304 | 1 (ref) | |||
| No | 3,178 | 420 | 1.19 [1.04; 1.38] | 0.015 | 2,776 | 241 | 0.99 [0.80; 1.21] | 0.887 | 0.053 |
Odds ratios and 95% confidence intervals are calculated by binary logistic regression; analyses were adjusted for (Model 4, without area): age, sex, vandalism or crime, clean area, feeling part, helpful neighbours, easy accessibility to services, education, self-rated health, heart attack, hypertension, high cholesterol, stroke, diabetes, chronic lung disease, cancer, Parkinson’s, cataracts, hip fracture, other fractures, cognitive impairment, affective / emotional disorder, any arthritis, anti-hypertensives, drugs for pain, psychotropic drugs, living alone, difficulty in making ends meet, poor vision, poor hearing, country dummies, functional status, physical activity, maximum grip strength, falls at baseline and dizziness, faints or blackout at baseline. P values for interaction were calculated in the total sample of 25,467 participants, using a Model that included all five neighbourhood characteristic simultaneously, all co-variates of Model 4 (including area) and a specific interaction term (computed for each neighbourhood characteristic at a time, by multiplying each neighbourhood characteristic variable by area (rural versus urban)). Number of participants reporting falls at follow-up in each category: urban area residents: 1,320; rural area residents: 545
n number, pts participants, OR odds ratios, CI confidence intervals, ref reference
In Northern/Western Europe, vandalism or crime tended to be associated with an increased falls risk, at follow-up, after full adjustment (supplementary Table S18).
Further analyses
Most participants did not change residence during follow-up (19,155, 88.8% of those with known information on this, supplementary Table S19). When restricting to these participants not changing residence, vandalism or crime and difficult accessibility to services were associated with an increased falls risk at follow-up (supplementary Table S20).
When restricting to participants not reporting falls at baseline (n = 23,748), vandalism or crime was associated with an increased falls risk, at follow-up, after full adjustment (supplementary Table S21).
Supplementary Tables S22 to S38 present associations between neighbourhood and falls risk, as estimated by using alternative Models of adjustment, including each neighbourhood characteristic one at a time. In cross-sectional analyses, vandalism or crime, lack of cleanliness and difficult accessibility to services, respectively, were each associated with an increased falls risk in the total population, after full adjustment (supplementary Table S22). After full adjustment, at 2-year follow-up, vandalism or crime was consistently associated with increased falls risks, in women (supplementary Table S26 and S27), adults aged ≥ 65 years (supplementary table S33), adults without functional impairment (supplementary Table S34), residents in Northern and Western Europe (supplementary Table S35) and urban areas residents (supplementary Table S36).
Discussion
This large study shows that perceived vandalism or crime was associated with an increased falls risk in community-dwelling adults, independent of individual falls risk factors, at 2-year follow-up. Both perceived vandalism or crime and difficult accessibility to services were associated with an increased falls risk, at 2-year follow-up, among residents in urban areas. Moreover, perceived lack of cleanliness was associated with an increased falls risk, at 2-year follow-up, among residents in rural areas.
Our estimates of association between these neighbourhood characteristics and falls risk seemed attenuated in analyses at follow-up, compared to analyses at baseline. This may imply that the impact of contemporary neighbourhood on falls risk is higher than that of the neighbourhood of two years previous; our analyses at follow-up may be conservative.
This cross-European study adds to the literature—mainly from the USA—on neighbourhood and falls. In the Health and Retirement Study, neighbourhood social cohesiveness and better physical environment (absence of rubbish; safety walking alone at night) were associated with lower falls risk in the two years prior to the interview, among older adults (Nicklett et al. 2017). In the National Health and Aging Trends Study, older adults residing in areas with higher environmental barriers on sidewalks/streets and uneven walking surfaces or broken steps had a higher falls risk at follow-up (Lee et al. 2019). Among community-dwelling older adults in Alabama, greater neighbourhood disadvantage was longitudinally associated with a higher falls risk (Lo et al. 2016). Among older adults in Sao Paulo, Brazil, moderate homicide rate in the local area was associated with indoor falls (do Nascimento et al. 2017).
The novelty of our study is the exploration of the link between neighbourhoods and falls across Europe. Findings from the USA may not be generalizable to Europe; health inequalities, neighbourhood segregation and crime may be more pronounced in the USA than Europe (Avendano et al. 2009). Yet, we showed considerable differences in perceived neighbourhood characteristics across Europe.
Consistent with previous reports (Lovasi et al. 2014; Shenassa et al. 2006; Stafford et al. 2007; Piro et al. 2006; Lorenc et al. 2013), in our study, women were more likely to report vandalism or crime than men. At follow-up, vandalism or crime was associated with falls in women, while not in men. This could be due to lack of statistical power in men, rather than to sex-differences in the association between vandalism or crime and falls risk. Notably, we failed to show any interaction by sex in these associations.
Previous literature suggested that older adults may be more vulnerable to adverse neighbourhood effects, than middle-aged adults. Consistent with this, in our study, vandalism or crime tended to be associated with an increased falls risk in older adults, while not in middle-aged adults (p for interaction tended to be significant). Yet, we showed associations of vandalism or crime with an increased falls risk in adults without functional impairment. We failed to show any interaction of functional status on the association between any neighbourhood characteristic and falls risk.
A novelty and strength of our study is to provide analyses that are stratified by area of residence and European region. Our study confirms previous reports suggesting that crime and lack of social cohesion may be more prevalent in urban versus rural areas (World Health Organization 2016); on the other hand, accessibility to services is easier in urban areas. Our novel finding is that the importance of neighbourhood characteristics may vary by area; crime and difficult accessibility to services were associated with falls in urban areas and lack of cleanliness in rural areas. Notably, the area of residence modified the association between cleanliness and falls risk; the negative impact of lack of cleanliness was consistent among residents of rural areas (i.e. interaction was significant). Vandalism or crime tended to be associated with an increased falls risk in Northern/Western Europe; however, the analyses stratified by European region may lack statistical power.
Different factors may mediate the association between adverse neighbourhood environment and increased falls risk. First, vandalism and lack of cleanliness may directly influence the risk of outdoor falls, through slips on uneven or wet surfaces or objects (Li et al. 2006). Environmental barriers to safe walking outdoors include litter, broken steps, putrescent leaves, dog excrements on sidewalks and streets, uneven walking surfaces, inadequate (winter) street maintenance, unsafe traffic patterns, uncontrolled dogs, delivery men and poor lighting (Lee et al. 2019; Chippendale and Boltz 2015).
Second, perceived neighbourhood unsafety may lead to decreased physical activity, deconditioning and loss of physical function (Shenassa et al. 2006; Balfour and Kaplan 2002). Signs of vandalism and lack of cleanliness may contribute to perceived unsafety, by suggesting lack of social control (“broken window theory”) (Lorenc et al. 2013). On the other hand, better neighbourhood environments may act as motivators and facilitators of outdoor physical activity; outdoor walking and exercise are promoted by aesthetic features of the environment, amenities, natural blue and green spaces and opportunities for socialising (Chippendale and Boltz 2015). Spacious even continuous walking paths, rails on stairs, fences to hold for support, presence of supportive people, availability of public bathrooms, benches for resting and transport are all facilitators of outdoor activity (Chippendale and Boltz 2015). In our study, each negative neighbourhood characteristic was associated with physical inactivity; yet, a few neighbourhood-falls associations remained consistent after adjusting for physical inactivity.
Third, adults residing in disadvantaged neighbourhoods, compared to affluent neighbourhoods, may have lower socio-economic status, which is a risk factor for physical and mental diseases (Diez Roux and Mair 2010; Pickett and Pearl 2001). However, our associations between neighbourhood characteristics and falls were independent of co-morbidities, medications, subjective health and ability to make ends meet. Ability to make ends meet, as measured with a simple question, reflects individual socio-economic status and psychological stress if inadequate and is more directly comparable across countries. Similarly, adults in disadvantaged neighbourhoods may experience more loneliness, another falls risk factor (Bu et al. 2020). However, our associations were independent of living alone. Fourth, adults who experience falls, especially if related to environmental hazards, may develop a worse perception of their neighbourhood, compared to adults not falling (reverse causation). However, crime or vandalism and lack of cleanliness were associated with an increased falls risk at follow-up, in the sample of participants not falling at baseline.
Finally, adverse neighbourhood impacts the global well-being of adults. Indeed, individual well-being may be more affected by perceived—rather than objective—neighbourhood characteristics. Perceived neighbourhood disorder may lead to chronic stress response, poorer muscle strength and physical performance (Duchowny et al. 2020).
A major strength of our longitudinal population-based study is the large and diverse sample of participants; we included men and women, middle-aged and older adults, adults with and those without functional impairment. The neighbourhoods were also diverse and located in urban versus rural areas and across Europe. Furthermore, we examined various dimensions of the neighbourhood environment—physical, social and service. Rather than creating summary indexes, we assessed the association between each neighbourhood characteristic and falls risk separately, to explore which could be more relevant for falls risk. Through stratified analyses, we could explore whether neighbourhood characteristics may have different relevance for different residents. By stratifying our analyses according to area or European region, we could take into account unmeasured co-variates that are related to historical and cultural factors. Our definition of European Regions was based on the historical and socio-political background. Furthermore, we carried out both cross-sectional analyses and analyses at follow-up. Of note, the association between vandalism or crime and falls risk was significant in both cross-sectional and follow-up analyses. In this way, we could show temporality of this association; perceived vandalism or crime preceded the falls. Another strength of our study is to show that vandalism or crime was associated with increased falls risk, in the total population, at 2-year follow-up, even when simultaneously adjusting for the other neighbourhood characteristics. A further strength is the adjustment for many individual-level co-variates, including socio-demographic factors, subjective health, co-morbidities, medications, physical inactivity, living alone, ability to make ends meet and sensory impairment. As expected, negative neighbourhood characteristics were associated with many of these co-variates, which are markers of poor health and falls risk factors. After adjustment, our estimates of the associations between neighbourhood and falls became attenuated; this suggests that the associations are partly accounted for by these co-variates, yet, a few exist beyond these. Moreover, standardized, harmonised questionnaires were used across the fourteen countries.
This study has some limitations. First, falls were retrospectively self-reported and this could result in underreporting of falls, especially non-injurious falls (Hoffman et al. 2018). Second, information was available only on falls in the six months prior to the baseline and follow-up interviews; in this way, we may have missed falls that occurred immediately following the baseline interview and likely underestimated our associations. Third, we lacked information on frequency and circumstances of falls (indoors versus outdoors). Although cleanliness of neighbourhoods may affect outdoor more than indoor falls, we postulate that neighbourhood environments affect the global health of individuals and, thus, the risk of both outdoor and indoor falls. Fourth, we included only those participants who were direct respondents on neighbourhood and co-variates; although we excluded many participants, we could focus on subjective perceptions of neighbourhoods. Fifth, the participants who dropped out between Wave 5 and 6 were more likely to be men and report lack of social cohesiveness, difficult accessibility to services and falls at baseline; we may have lost statistical power in men and attrition may have attenuated our estimates at follow-up. Sixth, we measured perception of neighbourhood contemporary to the baseline interview, rather than measuring cumulative life-time exposures to adverse neighbourhoods. Of note, our main findings remained consistent when restricting to participants not changing residence during follow-up. Yet, adults may change neighbourhood; neighbourhoods may change. Seventh, we included only community-dwelling adults in our study. Neighbourhood may be more relevant for the falls risk of community-dwelling rather than institutionalised adults; our findings may not be generalizable to institutionalised populations. Eighth, we did not formally compare coefficient estimates between Models of adjustment. Finally, we cannot exclude residual confounding in our estimates of associations between perceived neighbourhood characteristics and falls risk.
Previous research on falls prevention focussed on individual risk factors or home safety. Future research should explore whether urban planning and other policy intervention aimed at improving walkability, safety and cleanliness of residential neighbourhoods and service accessibility may reduce falls risk in adults. Environmental adjustments may reduce the falls risk of many adults, in the long term. Previous studies combining environmental adjustments with other interventions yielded mixed results (McClure et al. 2005). Future research could include “living in adverse neighbourhoods” into falls risk stratification tools.
In conclusion, perceived neighbourhood environment is associated with falls risk in community-dwelling adults, independent of individual risk factors. Perceived crime or vandalism and difficult accessibility to services in urban areas and lack of cleanliness in rural areas were associated with an increased falls risk, at follow-up, while social cohesiveness was not. Further research should explore whether urban policies aiming at improving safety, cleanliness and service accessibility may reduce falls risk inequalities .
Supplementary Information
Below is the link to the electronic supplementary material.
Acknowledgements
This paper uses data from SHARE Waves 5 and 6 for methodological details (Börsch-Supan A, Brandt M, Hunkler C, Kneip T, Korbmacher J, Malter F, Schaan B, Stuck S, Zuber S; SHARE Central Coordination Team. Data Resource Profile: the Survey of Health, Ageing and Retirement in Europe (SHARE). Int J Epidemiol. 2013 Aug;42(4):992-1001.. Epub 2013 Jun 18. PMID: 23778574; PMCID: PMC3780997). The SHARE data collection has been funded by the European Commission through FP5 (QLK6-CT-2001-00360), FP6 (SHARE-I3: RII-CT-2006-062193, COMPARE: CIT5-CT-2005-028857, SHARELIFE: CIT4-CT-2006-028812), FP7 (SHARE-PREP: GA N°211909, SHARE-LEAP: GA N°227822, SHARE M4: GA N°261982, DASISH: GA N°283646) and Horizon 2020 (SHARE-DEV3: GA N°676536, SHARE-COHESION: GA N°870628, SERISS: GA N°654221, SSHOC: GA N°823782) and by DG Employment, Social Affairs & Inclusion. Additional funding from the German Ministry of Education and Research, the Max Planck Society for the Advancement of Science, the U.S. National Institute on Aging (U01_AG09740-13S2, P01_AG005842, P01_AG08291, P30_AG12815, R21_AG025169, Y1-AG-4553-01, IAG_BSR06-11, OGHA_04-064, HHSN271201300071C) and from various national funding sources is gratefully acknowledged (see www.share-project.org).
Authors’ contributions
GO contributed to conceptualization, methodology, formal analysis, writing—original draft. JR contributed to conceptualization, supervision, writing—review & editing. KA-R contributed to conceptualization, supervision, writing—review & editing. LLS-H contributed to conceptualization, methodology, writing—review & editing. TM contributed to conceptualization, supervision, writing—review & editing.
Funding
Dr Giulia Ogliari was supported by Nottingham Hospitals Charity, Nottingham, UK (grant APP2380/N7359 (N7359 Osteoporosis & Falls Research for “Improving Quality of Life In Older Patients”).
Data availability
Data are available upon request from the SHARE website (see http://www.share-project.org/data-access/user-registration.html).
Code availability
Not applicable.
Declarations
Conflict of interest
On behalf of all authors, the corresponding author states that there is no conflict of interest.
Ethics approval
SHARE obtained ethical approval by the University of Mannheim and the Ethics Council of the Max Planck Society.
Consent to participate
This article uses anonymised data.
Consent for publication
Not applicable. Personal data were not identifiable during the analysis.
Footnotes
Responsible Editor: Morten Wahrendorf
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Contributor Information
Giulia Ogliari, Email: giulia.ogliari@virgilio.it, Email: Giulia.Ogliari1@nottingham.ac.uk.
Jesper Ryg, Email: jesper.ryg@rsyd.dk.
Karen Andersen-Ranberg, Email: karanberg@health.sdu.dk.
Lasse Lybecker Scheel-Hincke, Email: llscheel-hincke@health.sdu.dk.
Tahir Masud, Email: tahir.masud@nuh.nhs.uk.
References
- Auchincloss AH, Mujahid MS, Shen M, Michos ED, Whitt-Glover MC, Diez Roux AV. Neighborhood health-promoting resources and obesity risk (the multi-ethnic study of atherosclerosis) Obesity (silver Spring) 2013;21(3):621–628. doi: 10.1002/oby.20255. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Avendano M, Glymour MM, Banks J, Mackenbach JP. Health disadvantage in US adults aged 50 to 74 years: a comparison of the health of rich and poor Americans with that of Europeans. Am J Public Health. 2009;99(3):540–548. doi: 10.2105/AJPH.2008.139469. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Balfour JL, Kaplan GA. Neighborhood environment and loss of physical function in older adults: evidence from the Alameda County Study. Am J Epidemiol. 2002;155(6):507–515. doi: 10.1093/aje/155.6.507. [DOI] [PubMed] [Google Scholar]
- Börsch-Supan A, Brandt M, Hunkler C, et al. Data resource profile: the survey of health, ageing and retirement in Europe (SHARE) Int J Epidemiol. 2013;42(4):992–1001. doi: 10.1093/ije/dyt088. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Börsch-Supan A (2020a) Survey of health, ageing and retirement in Europe (SHARE) Wave 5. Release version: 7.1.0. SHARE-ERIC. Data Set. 10.6103/SHARE.w5.710
- Börsch-Supan A (2020b) Survey of health, ageing and retirement in Europe (SHARE) Wave 6. Release version: 7.1.0. SHARE-ERIC. Data set. 10.6103/SHARE.w6.710
- Bu F, Abell J, Zaninotto P, Fancourt D. A longitudinal analysis of loneliness, social isolation and falls amongst older people in England. Sci Rep. 2020;10(1):20064. doi: 10.1038/s41598-020-77104-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Chippendale T, Boltz M. The neighborhood environment: perceived fall risk, resources, and strategies for fall prevention. Gerontologist. 2015;55(4):575–583. doi: 10.1093/geront/gnu019. [DOI] [PubMed] [Google Scholar]
- Christine PJ, Auchincloss AH, Bertoni AG, et al. Longitudinal associations between neighborhood physical and social environments and incident type 2 diabetes mellitus: the multi-ethnic study of atherosclerosis (MESA) JAMA Intern Med. 2015;175(8):1311–1320. doi: 10.1001/jamainternmed.2015.2691. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Deandrea S, Lucenteforte E, Bravi F, Foschi R, La Vecchia C, Negri E. Risk factors for falls in community-dwelling older people: a systematic review and meta-analysis. Epidemiology. 2010;21(5):658–668. doi: 10.1097/EDE.0b013e3181e89905. [DOI] [PubMed] [Google Scholar]
- Diez Roux AV. Invited commentary: places, people, and health. Am J Epidemiol. 2002;155(6):516–519. doi: 10.1093/aje/155.6.516. [DOI] [PubMed] [Google Scholar]
- Diez Roux AV, Mair C. Neighborhoods and health. Ann N Y Acad Sci. 2010;1186:125–145. doi: 10.1111/j.1749-6632.2009.05333.x. [DOI] [PubMed] [Google Scholar]
- Do Nascimento CF, Duarte YA, Lebrão ML, Chiavegatto Filho AD. Individual and contextual characteristics of indoor and outdoor falls in older residents of São Paulo Brazil. Arch Gerontol Geriatr. 2017;68:119–125. doi: 10.1016/j.archger.2016.10.004. [DOI] [PubMed] [Google Scholar]
- Duchowny KA, Glymour MM, Cawthon PM. Is perceived neighbourhood physical disorder associated with muscle strength in middle aged and older men and women? Findings from the US health and retirement study. J Epidemiol Community Health. 2020;74(3):240–247. doi: 10.1136/jech-2019-213192. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Evenson KR, Block R, Diez Roux AV, McGinn AP, Wen F, Rodríguez DA. Associations of adult physical activity with perceived safety and police-recorded crime: the Multi-ethnic Study of Atherosclerosis. Int J Behav Nutr Phys Act. 2012;9:146. doi: 10.1186/1479-5868-9-146. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hoffman GJ, Ha J, Alexander NB, Langa KM, Tinetti M, Min LC. Underreporting of fall injuries of older adults: implications for wellness visit fall risk screening. J Am Geriatr Soc. 2018;66(6):1195–1200. doi: 10.1111/jgs.15360. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kaiser P, Diez Roux AV, Mujahid M, et al. Neighborhood environments and incident hypertension in the multi-ethnic study of atherosclerosis. Am J Epidemiol. 2016;183(11):988–997. doi: 10.1093/aje/kwv296. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Katz S, Ford AB, Moskowitz RW, Jackson BA, Jaffe MW. Studies of illness in the aged. The index of ADL: a standardized measure of biological and psychological function. JAMA. 1963;185:914–919. doi: 10.1001/jama.1963.03060120024016. [DOI] [PubMed] [Google Scholar]
- Kelsey JL, Procter-Gray E, Hannan MT, Li W. Heterogeneity of falls among older adults: implications for public health prevention. Am J Public Health. 2012;102(11):2149–2156. doi: 10.2105/AJPH.2012.300677. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lawton MP, Brody EM. Assessment of older people: self-maintaining and instrumental activities of daily living. Gerontologist. 1969;9(3):179–186. doi: 10.1093/geront/9.3_Part_1.179. [DOI] [PubMed] [Google Scholar]
- Lee S, Lee C, Ory MG. Association between recent falls and changes in outdoor environments near community-dwelling older adults' homes over time: findings from the NHATS study. Int J Environ Res Public Health. 2019;16(18):3230. doi: 10.3390/ijerph16183230. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Li W, Keegan TH, Sternfeld B, Sidney S, Quesenberry CP, Jr, Kelsey JL. Outdoor falls among middle-aged and older adults: a neglected public health problem. Am J Public Health. 2006;96(7):1192–1200. doi: 10.2105/AJPH.2005.083055. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Li W, Procter-Gray E, Lipsitz LA, et al. Utilitarian walking, neighborhood environment, and risk of outdoor falls among older adults. Am J Public Health. 2014;104(9):e30–e37. doi: 10.2105/AJPH.2014.302104. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lo AX, Rundle AG, Buys D, et al. Neighborhood disadvantage and life-space mobility are associated with incident falls in community-dwelling older adults. J Am Geriatr Soc. 2016;64(11):2218–2225. doi: 10.1111/jgs.14353. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lorenc T, Petticrew M, Whitehead M, et al. Fear of crime and the environment: systematic review of UK qualitative evidence. BMC Public Health. 2013;13:496. doi: 10.1186/1471-2458-13-496. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lovasi GS, Goh CE, Pearson AL, Breetzke G. The independent associations of recorded crime and perceived safety with physical health in a nationally representative cross-sectional survey of men and women in New Zealand. BMJ Open. 2014;4(3):e004058. doi: 10.1136/bmjopen-2013-004058. [DOI] [PMC free article] [PubMed] [Google Scholar]
- McClure R, Turner C, Peel N, Spinks A, Eakin E, Hughes K. Population-based interventions for the prevention of fall-related injuries in older people. Cochrane Database Syst Rev. 2005 doi: 10.1002/14651858.CD004441.pub2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Mohd Hairi F, Mackenbach JP, Andersen-Ranberg K, Avendano M. Does socio-economic status predict grip strength in older Europeans? Results from the SHARE study in non-institutionalised men and women aged 50+ J Epidemiol Community Health. 2010;64(9):829–837. doi: 10.1136/jech.2009.088476. [DOI] [PubMed] [Google Scholar]
- Nevitt MC, Cummings SR, Kidd S, Black D. Risk factors for recurrent nonsyncopal falls. A prospective study. JAMA. 1989;261(18):2663–2668. doi: 10.1001/jama.1989.03420180087036. [DOI] [PubMed] [Google Scholar]
- Nicklett EJ, Lohman MC, Smith ML. Neighborhood environment and falls among community-dwelling older adults. Int J Environ Res Public Health. 2017;14(2):175. doi: 10.3390/ijerph14020175. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Organisation for Economic Co-operation and Development. Classifying educational programmes: manual for ISCED-97 implementation in OECD countries. Paris: OECD, (1999)
- Pickett KE, Pearl M (2001) Multilevel analyses of neighbourhood socioeconomic context and health outcomes: a critical review. J Epidemiol Community Health 55(2):111–122. 10.1136/jech.55.2.111 [DOI] [PMC free article] [PubMed]
- Piro FN, Noss O, Claussen B. Physical activity among elderly people in a city population: the influence of neighbourhood level violence and self perceived safety. J Epidemiol Community Health. 2006;60(7):626–632. doi: 10.1136/jech.2005.042697. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Rubenstein LZ. Falls in older people: epidemiology, risk factors and strategies for prevention. Age Ageing. 2006;35(2 Suppl):37ii–ii41. doi: 10.1093/ageing/afl084. [DOI] [PubMed] [Google Scholar]
- Shenassa ED, Liebhaber A, Ezeamama A. Perceived safety of area of residence and exercise: a pan-European study. Am J Epidemiol. 2006;163(11):1012–1017. doi: 10.1093/aje/kwj142. [DOI] [PubMed] [Google Scholar]
- Stafford M, Chandola T, Marmot M. Association between fear of crime and mental health and physical functioning. Am J Public Health. 2007;97(11):2076–2081. doi: 10.2105/AJPH.2006.097154. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Stafford M, Cummins S, Macintyre S, Ellaway A, Marmot M. Gender differences in the associations between health and neighbourhood environment. Soc Sci Med. 2005;60(8):1681–1692. doi: 10.1016/j.socscimed.2004.08.028. [DOI] [PubMed] [Google Scholar]
- Stevens JA, Corso PS, Finkelstein EA, Miller TR. The costs of fatal and non-fatal falls among older adults. Inj Prev. 2006;12(5):290–295. doi: 10.1136/ip.2005.011015. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Tinetti ME, Speechley M, Ginter SF. Risk factors for falls among elderly persons living in the community. N Engl J Med. 1988;319(26):1701–1707. doi: 10.1056/NEJM198812293192604. [DOI] [PubMed] [Google Scholar]
- Tinetti ME, Williams CS. Falls, injuries due to falls, and the risk of admission to a nursing home. N Engl J Med. 1997;337(18):1279–1284. doi: 10.1056/NEJM199710303371806. [DOI] [PubMed] [Google Scholar]
- Vaz C, Andrade AC, Silva U, et al. Physical disorders and poor self-rated health in adults living in four latin american cities: a multilevel approach. Int J Environ Res Public Health. 2020;17(23):8956. doi: 10.3390/ijerph17238956. [DOI] [PMC free article] [PubMed] [Google Scholar]
- World Health Organization . Global report on urban health: Equitable healthier cities for sustainable development. Geneva: WHO; 2016. [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Data are available upon request from the SHARE website (see http://www.share-project.org/data-access/user-registration.html).
Not applicable.