http://www.diva-portal.org
This is the published version of a paper published in Journal of Transport and Health.
Citation for the original published paper (version of record):
Gustavsson, J., Nilson, F., Bonander, C. (2020)
Individual and contextual factors associated with the use of anti-slip devices according to a Swedish national survey
Journal of Transport and Health, 17: 1-8 https://doi.org/10.1016/j.jth.2020.100865
Access to the published version may require subscription.
N.B. When citing this work, cite the original published paper.
© 2020 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
Permanent link to this version:
http://urn.kb.se/resolve?urn=urn:nbn:se:kau:diva-78942
Journal of Transport & Health 17 (2020) 100865
Available online 22 May 2020
2214-1405/© 2020 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Protocol
Individual and contextual factors associated with the use of anti-slip devices according to a Swedish national survey
Johanna Gustavsson
a,b,*, Finn Nilson
a,b, Carl Bonander
a,caCentre for Public Safety, Faculty of Health Science and Technology, Karlstad University, Sweden
bRisk- and Environmental Studies, Faculty of Health Science and Technology, Karlstad University, Sweden
cHealth Metrics Unit, Sahlgrenska Academy, University of Gothenburg, Sweden
A B S T R A C T
Introduction: Walking as a means of transportation can enforce a more active lifestyle and constitutes an environmentally friendly option to motor vehicles. However, in Northern countries, ice and snow tend to increase the risk of fall injuries among pedestrians during the winter. Therefore, the use of anti-slip devices, such as ice cleats or “studded footwear”, has been suggested as a viable intervention in promoting an active lifestyle whilst reducing injury risk. We investigate the usage of anti-slip devices, focusing on people 50 years and above living in Sweden.
Method: We used nationally representative survey data for men and women aged 18–79 years and residing in Sweden (n ¼ 23,168), focusing primarily on middle-aged to older adults (50þ years). We used logistic regression to identify predictors of use.
Results: Overall, our estimates suggest that 28.5 (95% CI: 27.0, 29.2) percent of the Swedish population use anti-slip devices during snowy and slippery road conditions, with usage rates increasing strongly with age (from roughly 10 percent at 20 years to 60 percent at 79 years). In addition, the results show that being female, experiencing a fear of falling, living in a municipality with a high number of snow days, and using other types of personal safety equipment increases the probability of being an anti-slip device user.
Conclusions: Our results imply that people at risk for outdoor fall injuries are high users. Even so, the number of pedestrian injuries due to slipping on snow and ice are still substantial and there are a several potential target groups for future intervention. More research is needed to determine if the devices are used correctly, and to determine the barriers to anti-slip device use in low-use populations.
1. Introduction
Walking as a means of transportation can enforce a more active lifestyle (WHO, 2002) and constitutes an environmentally friendly option to motor vehicles. However, although the health benefits of active transportation exceed the increased risk of injury (Buehler et al., 2016), pedestrians are vulnerable road users. Historically, road safety policies have prioritized a reduction of motor vehicle crashes ahead of other incident types. Whilst a natural prioritization, given the mortality distribution in road traffic fatalities, the issue of other road traffic related incidents, such as pedestrian falls, has largely been ignored. Indeed, official road safety statistics rarely include pedestrian falls globally (OECD, 2017; WHO, 2018). However, lately there have been signs of a rising interest to improve road safety for vulnerable road users (WHO, 2018), perhaps as a consequence of the reduced risk in motor vehicle crashes, that are now also targeted in the sustainable development goals (WHO, 2017).
As is the case for road traffic incidents in general, contextual factors are crucial to understand in order to improve safety. In countries with colder climates, snow and ice dominate as causes to road traffic injuries, regardless of whether these are related to motor vehicles (Malin et al., 2019) or vulnerable road users (Berntman, 2015). As can be seen in Fig. 1, the risk for snow- and ice-related fall-injuries requiring medical attendance in Sweden is unevenly distributed in the population. Specifically, two groups are visible;
* Corresponding author. Centre for Public Safety, Faculty of Health Science and Technology, Karlstad University, Sweden.
E-mail address: johanna.gustavsson@kau.se (J. Gustavsson).
Contents lists available at ScienceDirect
Journal of Transport & Health
journal homepage: http://www.elsevier.com/locate/jth
https://doi.org/10.1016/j.jth.2020.100865
Received 2 August 2019; Received in revised form 24 February 2020; Accepted 21 April 2020
Journal of Transport & Health 17 (2020) 100865
2
teenagers and older adults. In terms of older adults, a risk increase can be seen at around the age of 50 years, especially for women (Fig. 1). Whilst the teenage group is important to study, previous research has shown that the injury severity of outdoor falls is significantly greater in the older age group (Berntman, 2015), therefore motivating a focus on this group.
Disregarding the injury perspective, slippery road conditions can also have a limiting effect on older adults, 29% of which report avoiding going outside during wintertime due to road conditions (NTF, 2015). In terms of risk factors for outdoors falls among older adults, an active life and high mobility skills are critical (Bergland et al., 2003). However, it is important to promote these factors to prolong the duration of independent life. Therefore, the use of anti-slip devices, such as ice cleats or “studded footwear”, has been suggested as a viable intervention in promoting an active lifestyle whilst reducing injury risk. Previous studies have shown that they have the potential to decrease the risk of outdoor falls and their subsequent injuries (Berggård and Johansson, 2010; McKiernan, 2005) and recently municipalities in Sweden have started showing an interest in intervention programs with the purpose of increasing the use of anti-slip devices.
However, there is little research on current usage rates in high risk populations, and, to our knowledge, no studies have documented the predictors of use. Therefore, in this paper, we use data from a nationally representative survey to investigate the usage of anti-slip devices, focusing on people 50 years and above living in Sweden.
2. Materials and methods 2.1. Data collection
We used nationally representative survey data for men and women aged 18–79 years and residing in Sweden (n ¼ 23,168), collected by Statistics Sweden in 2014 on behalf of the Swedish Civil Contingencies Agency (MSB). They used a complex stratified random sampling design which included one national sample (n ¼ 10,000) and municipal samples for 58 municipalities (n ¼ 34,800) where data was oversampled to allow for municipality-specific estimates. The sampling frame was also stratified into eight strata based on age groups (18–29, 30–49, 50–64 and 65–79 years) and sex. Individuals were drawn from the Swedish total population register, and socioeconomic register data from Statistics Sweden was linked to the respondents to supplement the data collected from the survey questionnaire. We used design weights to account for the stratified sampling scheme and obtain nationally representative estimates. To correct for bias due to survey nonresponse (the response rate was 52%), Statistics Sweden calibrated the design weights using a post- stratification estimator that adjust the weights to match known population totals within the levels of the variables sex, country of birth, age, marital status, income, educational attainment, region and 2011 municipality group classification used by the Swedish Associ- ation of Local Authorities and Regions (Deville and Sarndal, 1992; SKL, 2016). We also obtained estimates of the yearly number of days with snow cover per municipality from the Swedish Meteorological and Hydrological Institute (SMHI). The values were interpolated from 338 measurement stations (collecting data on snow depth) and correspond to estimates for the regional center in each munic- ipality (averaged over the period 2003 to 2018).
Table 1 details all measures used in this study. They include a battery of sociodemographic characteristics, details on fall history and fear of falling, general safety attitudes, safety behaviors and a climate indicator.
2.2. Analysis
Most of our analyses were focused on high risk age groups (50–79 years, n ¼ 14,130). Partial nonresponse on the included variables
Fig. 1. Snow-related fall-injury patients per 100.000 person-years (ICD-10 external cause code: W00) by age group and sex. The data is based on publicly available, aggregated statistics from the National Patient Register (The National Board of Health and Welfare), and reflects the period 2008 to 2017.
J. Gustavsson et al.
Journal of Transport & Health 17 (2020) 100865
was fairly low (1–10%), but only 70% of the sample (n ¼ 9811) had complete covariate data on all variables. To handle this, we imputed missing data using multiple imputation with chained equations (White et al., 2011), including all variables to be used in the following analyses in the imputation model. To account for the uncertainty introduced in this procedure, all estimates presented below were pooled from ten imputed datasets. The imputation was performed using the mice package for R (van Buuren and Groothuis-Oudshoorn, 2011).
The remaining analysis was performed in several steps. First, we investigated the use of studded footwear in different population groups using generalized additive models for continuous variables and prevalence estimates for categorical data (Hastie and Tib- shirani, 1987). Next, we fitted a logistic regression model to identify predictors of use for individuals aged 50–79 years. To enhance the interpretability of the coefficients from this model, we converted the output to estimated marginal effects (in percentage points), keeping all other covariates constant at their mean (this analysis was performed in Stata version 15.1 using the margins command). In all instances, we used robust (sandwich) variance estimators to account for the use of survey sampling and non-response weights (Robins et al., 2000).
Ethical approval was received from the regional ethical board (Nr, 2018/480).
3. Results
Overall, our estimates suggest that 28.5 (95% CI: 27.0, 29.2) percent of the Swedish population aged 18–79 years use studded footwear during snowy and slippery road conditions. As seen in Fig. 2a, usage rates increase strongly with age (from roughly 10 percent at 20 years to 60 percent at 79 years). The average number of snow days in a municipality also increased use, up to approximately 150 snow days/year when the correlation declined (Fig. 2b). Another clear pattern is that women report higher usage rates than men across most ages. On average, the prevalence is approximately 20 percentage points higher among women.
Estimated usage rates for different population characteristics in all ages, and in the high-risk age groups (50–79 years), are pre- sented in Table 2. We present characteristics associated with significant deviations (p < 0.05) from the average prevalence in bold face.
For all ages, we can see that many of the included socioeconomic indicators appear to be associated with use. However, many of these are also associated with age, and few of them still stand out after controlling for age, as can be seen in the age stratified results (also presented in Table 2).
Several expected patterns emerge in the age-stratified results besides the association with age and sex (Table 2). For instance, fear of falling, using other types of personal safety equipment related to transportation and living in a municipality with a high number of snow days per year appears to affect usage rates positively. Interestingly, there is no positive correlation with educational attainment or income levels; if anything, the results imply that the prevalence is lower among high income individuals. We also found other unexpected patterns; being an immigrant is associated with lower usage rates, but only in the oldest age group. There is no such indication in other age groups. Moreover, living in a rural area is associated with a small, but significant increase in usage rates, and living with children <20 years is associated with lower usage rates.
In Table 3, we present how these factors affect the use of anti-slip devices in the age group 50–79 years according to a logistic regression model. Similar to the unadjusted results, we found that older age, being female, experiencing a fear of falling, living in a municipality with a high number of snow days, and using other types of personal safety equipment increases the probability of being a Table 1
Variables included in the study.
Variable/Question Type of measure Categories Data source
Age group Sociodemographic 18–29 years; 30–49 years; 50–64 years; 64–79 years Register data
Female Sociodemographic Yes; No Register data
Educational attainment Sociodemographic Primary; Secondary; Tertiary Register data
Occupational status Sociodemographic Employed; Student; Retired; Unemployed; Other Register data
Income group Sociodemographic <165 k SEK/year; 165-310 k SEK/year; 310 k SEK/year Register data
Immigrated to Sweden Sociodemographic Yes; No Register data
Marital status Sociodemographic Married; Not married (ever); Divorced; Widowed Register data
Living situation Sociodemographic Alone; With others (adults only); With children <20
years Survey data
Area of residence Sociodemographic Regional center; Other urban area; Rural area Survey data
Fallen in the last five years? Fall history Yes; No Survey data
Fear of falling Risk perception Never; Sometimes; Often/Always Survey data
To which degree can you influence fall risk? Safety attitudes Little; Either or; Greatly Survey data
Are you a safety conscious person? Safety attitudes Yes; No; Don’t know Survey data
Safety of others is more important than my own? Safety attitudes Yes; No; Don’t know Survey data
Individuals are responsible for their own safety? Safety attitudes Yes; No; Don’t know Survey data
Accidents happen no matter what? Safety attitudes Yes; No; Don’t know Survey data
Do you use studded footwear when it is slippery
outside? Safety behaviors
(primary) Yes; No Survey data
Do you use a bicycle helmet when cycling? Safety behaviors (other) Yes; No Survey data
Do you use a life jacket when boating? Safety behaviors (other) Yes; No Survey data
Do you use reflective clothing when it is dark outside? Safety behaviors (other) Yes; No Survey data
Snow days in municipality of residence Climate indicator Quartile groups Weather
data J. Gustavsson et al.
JournalofTransport&Health17(2020)100865
4
Fig. 2. Proportion of studded footwear users in Sweden by (a) age and sex, and (b) snow exposure and sex (in high risk age groups; 50–79 years), as estimated by generalized additive models with cubic splines.
J. Gustavsson et al.
Journal of Transport & Health 17 (2020) 100865 Table 2
Studded footwear in the Swedish population by different characteristics, for the entire age range 18–79 years and for high risk groups (50–64 years;
65–79 years).
Studded footwear use by age group (%, with 95% CI)
Characteristic All ages, 18–79 years 50–64 years 65–79 years
Overall 28.5 (27.0, 29.2) 33.3 (31.2, 35.5) 60 (57.7, 62.4)
Female 35.4 (33.8, 37.1) 44.8 (41.7, 48) 74.8 (71.9, 77.6)
Educational attainment
Primary 38.2 (35.3, 41.2) 36.2 (30.6, 42.2) 60.1 (56, 64.1)
Secondary 27.6 (25.9, 29.3) 34.3 (31.1, 37.5) 59.5 (55.8, 63.1)
Tertiary 24.5 (22.9, 26.2) 31.1 (27.8, 34.5) 60.9 (56.1, 65.4)
Occupational status
Employed 20.2 (18.9, 21.6) 30.1 (27.7, 32.5) 51.4 (39.2, 63.5)
Student 11.2 (8.3, 14.9) 11.9 (2.9, 38.2) *
Retired 57.8 (55.5, 60) 50.9 (44.8, 56.9) 61 (58.6, 63.4)
Unemployed 19 (14.4, 24.7) 28.3 (19.3, 39.5) 82.6 (15.7, 99.2)
Other 25.8 (20.2, 32.3) 37.3 (26.5, 49.7) 31.3 (17.9, 48.8)
Income group
<165 k SEK/year 26.5 (24.4, 28.6) 35.6 (29.6, 42.1) 66.9 (62.6, 70.9)
165-310 k SEK/year 34.9 (33, 36.9) 40.9 (37.1, 44.8) 62.6 (59.2, 65.9)
>310 k SEK/year 22.9 (21.3, 24.6) 27.5 (24.9, 30.4) 47.5 (42.7, 52.3)
Immigrated to Sweden 28.9 (25.9, 32.1) 35.2 (29.3, 41.6) 47.3 (40.2, 54.6)
Martial status
Married 34.6 (33, 36.3) 34.8 (32, 37.8) 60.2 (57.1, 63.1)
Not married (ever) 15.7 (14.3, 17.3) 28.9 (24.7, 33.6) 47.6 (39.6, 55.6)
Divorced 37 (33.6, 40.4) 32.4 (27.5, 37.7) 60.9 (55.1, 66.4)
Widowed 62 (56, 67.8) 44.1 (28.5, 61) 67.2 (60.8, 73.1)
Living arrangement
Alone 31.9 (29.3, 34.5) 31.7 (27, 36.9) 62.0 (57.5, 66.3)
With others (adults only) 33.4 (31.7, 35) 36.6 (33.9, 39.5) 59.9 (57, 62.6)
With children <20 years 18.6 (16.8, 20.4) 24.1 (19.7, 29.2) 45.7 (31.6, 60.4)
Area of residence
Regional center 25.3 (23.8, 26.7) 29.3 (26.5, 32.3) 59.0 (55.7, 62.2)
Other urban area 26.5 (24.3, 28.9) 32.4 (27.9, 37.2) 58.9 (53.9, 63.8)
Rural area 37.3 (34.8, 39.9) 42.5 (38.2, 46.9) 63.5 (58.8, 68)
Fallen in the last 5 years? 34.1 (30.5, 37.9) 36.6 (30.3, 43.4) 66.4 (59.7, 72.6) Fear of falling
Never 23.5 (22.2, 24.9) 29.1 (26.5, 31.8) 54.2 (50.8, 57.5)
Sometimes 35 (33, 37.1) 39.2 (35.4, 43.1) 64.9 (61.3, 68.3)
Often 40.4 (35.2, 45.8) 44.5 (34.4, 55.2) 73.3 (64.6, 80.4)
To which degree can you influence fall risk?
Little 34.6 (31.6, 37.7) 34.1 (28.2, 40.5) 57.7 (52.8, 62.3)
Either or 31 (28.8, 33.3) 36.5 (32.3, 40.9) 63.3 (58.9, 67.4)
Greatly 25 (23.7, 26.5) 31.7 (29, 34.5) 59.2 (55.7, 62.7)
Are you a safety conscious person?
Yes 29.7 (28.5, 31) 34.2 (31.8, 36.6) 61.4 (58.7, 63.9)
No 18.8 (16.5, 21.2) 27.4 (22.1, 33.4) 53.2 (46.4, 59.9)
Don’t know 30.7 (25.5, 36.5) 33.8 (23.2, 46.2) 55.9 (46.5, 64.8)
Safety of others is more important than your own?
Yes 27.5 (26.3, 28.8) 32.4 (30, 34.9) 60 (57.3, 62.7)
No 30.1 (27.2, 33.1) 37.3 (31.9, 43) 61.3 (55.3, 67)
Don’t know 29 (25.1, 33.1) 33.4 (25.8, 42) 58 (49.9, 65.7)
Individuals are responsible for their own safety?
Yes 28.2 (26.9, 29.5) 33.1 (30.7, 35.7) 60 (57.3, 62.7)
No 27.3 (24.9, 29.8) 33.8 (29, 38.9) 61.7 (56.1, 67)
Don’t know 28.8 (24.6, 33.3) 33.8 (25.5, 43.2) 56.7 (48.4, 64.7)
Accidents happen no matter what?
Yes 27.9 (26.2, 29.7) 34.3 (30.8, 38) 59.8 (56.1, 63.4)
No 26.7 (25.1, 28.2) 32.1 (29.2, 35.2) 59.8 (56.1, 63.4)
Don’t know 33.7 (30.5, 37.1) 35.1 (28.8, 41.9) 61.1 (55.5, 66.4)
Uses bicycle helmet when cycling 36.7 (34.8, 38.6) 42.9 (39.3, 46.6) 71.6 (68, 74.9)
Uses lifevest when boating 29.9 (28.7, 31.1) 34.3 (32, 36.8) 64.5 (62, 66.9)
Uses reflective clothing when dark outside 34 (32.6, 35.3) 38.3 (35.8, 40.8) 65.1 (62.6, 67.5) Snow days in municipality of residence
20-52/year (Q1) 23.8 (21.9, 25.7) 27.2 (23.6, 31.1) 56.9 (52.6, 61.1)
53-72/year (Q2) 26.6 (24.4, 28.9) 33.7 (29.2, 38.5) 60.6 (55.6, 65.3)
73-98/year (Q3) 26.1 (24.2, 28.1) 30.7 (27.1, 34.4) 53.3 (48.9, 57.7)
99-197/year (Q4) 44.1 (41, 47.3) 52.5 (46.7, 58.2) 78.4 (73.6, 82.6)
Notes: Studded footwear use was defined as a positive response to the question: “Do you use slip protection (e.g., studded footwear) when it is slippery or icy outside?“. Bold-faced differences are statistically significant at the 0.05-level. Confidence intervals account for the complex survey design and are pooled across 10 multiply imputed datasets to account for the uncertainty that arises when missing values are imputed. SEK ¼ Swedish krona.
J. Gustavsson et al.
Journal of Transport & Health 17 (2020) 100865
6 Table 3
Estimates of the marginal effect of different individual and contextual factors on studded footwear use among middle-aged and older adults in Sweden (50–79 years), derived from a logistic regression model.
Characteristics Marginal effect estimate (percentage points)a Lower 95% CI Upper 95% CI
Age group
50–64 years Ref
65–79 years 12.7 7.5 17.9
Female 19.4 16.2 22.7
Educational attainment
Primary Ref
Secondary 1.3 5.2 2.7
Tertiary 0.9 5.3 3.4
Occupational status
Employed Ref
Student 20.6 42.2 1.0
Retired 11.1 5.6 16.6
Unemployed 2.9 14.1 8.3
Other 2.7 6.2 11.7
Income
<165 k SEK/year Ref
165-310 k SEK/year 1.7 2.7 6.1
>310 k SEK/year 4.6 9.8 0.5
Immigrated to Sweden 0.5 5.5 4.5
Marital status
Married Ref
Not married (ever) 5.8 ¡10.8 ¡0.9
Divorced 1.5 6.2 3.2
Widowed 0.5 6.8 7.7
Living situation
Alone Ref
With others (adults only) 0.3 5.2 4.7
With children <20 years 6.7 ¡12.4 ¡0.9
Area of residence
Regional center Ref
Other urban area 3.0 0.7 6.8
Rural area 5.3 1.7 8.9
Fallen in the last five years? 1.4 3.3 6.2
Fear of falling
Never Ref
Sometimes 6.8 3.4 10.1
Often/Always 11.2 3.9 18.5
To which degree can you influence fall risk?
Little Ref
Either or 2.9 2.3 8.1
Greatly 1.5 3.1 6.0
Are you a safety conscious person?
Yes Ref
No 2.7 2.1 7.6
Don’t know 0.7 9.2 7.8
Safety of others more important than own?
Yes Ref
No 4.1 8.4 0.2
Don’t know 0.3 7.2 6.6
Individuals responsible for their own safety?
Yes Ref
No 2.7 6.7 1.3
Don’t know 5.8 12.9 1.4
Accidents happen no matter what?
Yes Ref
No 2.1 1.2 5.4
Don’t know 4.2 0.9 9.2
Uses bicycle helmet when cycling 13.4 10.0 16.8
Uses life jacket when boating 7.6 3.2 12.0
Uses reflective clothing when dark outside 15.2 11.1 19.2
Snow days in municipality of residence
20-52/year (Q1) Ref
53-72/year (Q2) 5.4 1.2 9.6
73-98/year (Q3) 2.8 1.0 6.5
99-197/year (Q4) 22.4 17.9 26.9
aEstimated using States margins postestimation command after logistic regression, keeping all other covariates constant at their mean. Estimates have been adjusted using sampling and differential non-response weights, and missing covariate data was imputed using multiple imputation with J. Gustavsson et al.
Journal of Transport & Health 17 (2020) 100865
studded footwear user. Living with children <20 years and never having been married decrease the probability. The results do not change appreciably when splitting the age range into two groups (50–64 years, 65–79 years; data not shown).
4. Discussion
Our results imply that almost 30% of the Swedish population aged 18–79 years use anti-slip devices during slippery road condi- tions. Older people have the highest risk of falls and, as expected, usage rates increase with age (approximately 75% in the oldest age group). Apart from age and sex, the number of snow days were also strong contextual predictor. Likewise, fear of falling was a strong predictor for use of anti-slip devices among older adults. Both of these results support theories regarding the influence of (subjective) risk on safety equipment use (Gielen and Sleet, 2003; Sleet et al., 2010). Previous research show that women perceive a higher level of risk and also take more safety precautions then men (Enander, 2002, 2018), therefore perhaps explaining the differences between men and women.
As was mentioned in the introduction, many municipalities in Sweden have recently introduced programs offering free, or heavily subsidized studded footwear aimed at older adults. To our knowledge, these interventions tend to focus on the general population of older adults. However, our results imply that older men may require targeted intervention and campaigns in order to increase usage.
This is further supported by the fact that mortality from severe fall injuries are higher among men (Abrahamsen et al., 2009). However, creating effective interventions for men requires a better understanding as to why their usage rates are lower. Potential mechanisms could be a lower fear of falling (Pohl et al., 2015) and gender differences in risk perception (Hughes et al., 2008). Also, walking as a mode of transportation is less common among Swedish men (Trafikanalys, 2018) and older men have greater access to motorized transport than women (Frandberg and Vilhelmson, 2011). Hence, they may not experience the same need for anti-slip devices as women. Finally, as showed in Fig. 1, the actual risk of injury for men is lower meaning that the risk analysis of lesser usage amongst men may actually be appropriate. However, future shifts towards active transportation are likely to change this, should they occur.
Given that low usage is associated with lower use of other types of personal safety equipment, it seems that barriers to use may be related to general safety behaviors. Targeting safety attitudes and knowledge may therefore be a potential approach. However, in- terventions targeting these factors are often difficult to implement successfully and tools from behavioral science are needed to facilitate change (Sleet et al., 2010). In regards to low-usage groups, other potential specific target groups are older individuals who live alone, those who live in urban areas or have high incomes.
The latter result is particularly interesting given that the economic gradient in health and safety usually favors higher income groups (WHO, 2014). A potential explanation could be that individuals with high income have greater access to other means of transportation and physical activity than walking. Even so, this result raises the question of whether subsidy programs are needed for economic reasons. However, such programs may have an important symbolic value, often including information and other components that generally promote use.
4.1. Strengths and limitations
This is to our knowledge the first study describing usage pattern for studded footwear based on a large, nationally representative sample in a high risk context. We have accounted for survey nonresponse in an attempt to ensure that the results are valid for the Swedish population (aged 18–79 years) and used multiple imputation to account for missing variable information. These methods rely on the assumption that the data are missing at random (i.e., recoverable by observed variables), which is an untestable assumption.
Information bias may also be a concern, e.g., if some respondents do not answer the questions truthfully.
The data allowed us to quantify usage rates during icy and snowy conditions, but lack in details regarding the situations and ac- tivities they are used or if they are used all or some of the time. Additionally, the lack of respondents older than 79 years of age means that usage rates, as well as predictors of use, among the oldest adults is still unknown.
5. Conclusions
In summary, our results imply that people at risk for outdoor fall-injuries are high users. Even so, the number of injuries are substantial and there are a number of potential target groups for future intervention. However, more research is needed to determine the most important barriers to overcome in order to design effective interventions, especially in high-risk groups. Perceived (or actual) ability to perform a certain behavior (e.g., applying anti-slip devices to shoes) may be an important barrier for frail individuals, and social norms (in combination with low perceived risk) could possibly explain the lower usage rates among younger adults. These issues needs to be further explored in order to promote active transportation in areas with snow and ice.
Financial disclosure
This work was supported by The Kamprad Family Foundation for Entrepreneurship, Research and Charity (Nr. 20180067).
chained equations (m ¼ 10). Confidence intervals are based on robust standard errors to account for the weights (and pooled over the imputed datasets).
J. Gustavsson et al.
Journal of Transport & Health 17 (2020) 100865
8 CRediT authorship contribution statement
Johanna Gustavsson: Conceptualization, Methodology, Writing - original draft, Writing - review & editing, Project administra- tion, Funding acquisition. Finn Nilson: Methodology, Writing - original draft, Writing - review & editing. Carl Bonander: Concep- tualization, Methodology, Validation, Formal analysis, Data curation, Writing - original draft, Writing - review & editing, Project administration, Funding acquisition.
References
Abrahamsen, B., Van Staa, T., Ariely, R., Olson, M., Cooper, C., 2009. Excess mortality following hip fracture: a systematic epidemiological review. Osteoporos. Int. 20 (10), 1633–1650.
Berggård, G., Johansson, C., 2010. Pedestrians in wintertime—effects of using anti-slip devices. Accid. Anal. Prev. 42 (4), 1199–1204.
Bergland, A., Jarnlo, G.B., Laake, K., 2003. Predictors of falls in the elderly by location. Aging Clin. Exp. Res. 15 (1), 43–50. Retrieved from. http://www.ncbi.nlm.nih.
gov/pubmed/12841418.
Berntman, M., 2015. Fotg€angares olyckor och skador i trafikmilj€o med fokus på fallolyckor. In: Lunds Universitet, LTH, Instutionen F€or Teknik Och Samh€alle, Trafik Och V€ag, Bulletin 295. Bulletin 295/3000.
Buehler, R., G€otschi, T., Winters, M., 2016. In: Moving toward Active Transportation: How Policies Can Encourage Walking and Bicycling. Active Living Research, San Diego, CA.
Deville, J.C., Sarndal, C.E., 1992. Calibration estimators in survey sampling. J. Am. Stat. Assoc. 87 (418), 376–382. https://doi.org/10.2307/2290268. Retrieved from
<Go to ISI>://WOS:A1992HW62200013.
Enander, A.J., 2002. In: S€akerhet och risker i vardagen: en studie av uppfattningar, v€arderingar och beteenden hos allm€anheten i Sverige. R€addningsverket, Karlstad rapport P21-406/02).
Enander, A., 2018. M€anniskors f€orhållningss€att till risker, olyckor och kriser. In: Karlstad: Myndigheten F€or Samh€allsskydd Och Beredskap.
Frandberg, L., Vilhelmson, B., 2011. More or less travel: personal mobility trends in the Swedish population focusing gender and cohort. J. Transport Geogr. 19 (6), 1235–1244.
Gielen, A.C., Sleet, D., 2003. Application of behavior-change theories and methods to injury prevention. Epidemiol. Rev. 25, 65–76.
Hastie, T., Tibshirani, R., 1987. Generalized additive-models - some applications. J. Am. Stat. Assoc. 82 (398), 371–386. https://doi.org/10.2307/2289439. Retrieved from <Go to ISI>://WOS:A1987J105700001.
Hughes, K., van Beurden, E., Eakin, E.G., Barnett, L.M., Patterson, E., Backhouse, J., Newman, B., 2008. Older persons’ perception of risk of falling: implications for fall-prevention campaigns. Am. J. Publ. Health 98 (2), 351–357.
Malin, F., Norros, I., Innamaa, S., 2019. Accident risk of road and weather conditions on different road types. Accident Analysis & Prevention, 122, 181–188.
McKiernan, F.E., 2005. A simple gait-stabilizing device reduces outdoor falls and nonserious injurious falls in fall-prone older people during the winter. J. Am. Geriatr.
Soc. 53 (6), 943–947.
NTF, 2015. Livr€add f€or att halka! En unders€okning om €aldres upplevelse av fallolyckor på sn€o och is. Solna:NTF.
OECD, 2017. In: Road Safety Annual Report 2017. OECD Publishing, Paris.
Pohl, P., Ahlgren, C., Nordin, E., Lundquist, A., Lundin-Olsson, L., 2015. Gender perspective on fear of falling using the classification of functioning as the model.
Disabil. Rehabil. 37 (3), 214–222.
Robins, J.M., Hernan, M.A., Brumback, B., 2000. Marginal structural models and causal inference in epidemiology. Epidemiology 11 (5), 550–560.
SKL, 2016. Kommungruppsindelning 2017. Omarbetning av sveriges kommuner och landstings kommungruppsindelning. In: Stockholm: Sveriges Kommuner Och Landsting.
Sleet, D.A., Carlson Gielen, A., Diekman, S., Ikeda, R., 2010. Preventing unintentional injury: a review of behavior change theories for primary care. Am. J. Lifestyle Med. 4 (1), 25–31.
Trafikanalys, 2018. In: Så Reser Vi Baserat På Socioekonomi - Resm€onster F€or 37 Grupper, vol. 9. Trafikanalys, Stockholm.
van Buuren, S., Groothuis-Oudshoorn, K., 2011. Mice: multivariate imputation by chained equations in R. J. Stat. Software 45 (3), 1–67.
White, I.R., Royston, P., Wood, A.M., 2011. Multiple imputation using chained equations: issues and guidance for practice. Stat. Med. 30 (4), 377–399.
WHO, 2002. A Physically Active Life through Everyday Transport with a Special Focus on Children and Older People and Examples and Approaches from Europe (Retrieved from).
WHO, 2014. Injuries and inequities - guidance for addressing inequities in unintentional injuries. In: Copenhagen: WHO Regional Office for Europe.
WHO, 2017. Fact sheets on sustainable development goals: health targets. Road safety. In: Copenhagen: World Health Organization Regional Office for Europe.
WHO, 2018, 2018. In: Global Status Report on Road Safety 2018. World Health Organization, Geneva. Licence: CC BYNC-SA 3.0 IGO.
J. Gustavsson et al.
