This is an open-access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in the Journal of Medical Internet Research, is properly cited. The complete bibliographic information, a link to the original publication on http://www.jmir.org/, as well as this copyright and license information must be included.
Use of information and communication technologies (ICT) among seniors is increasing; however, studies on the use of ICT by seniors at the highest risk of health impairment are lacking. Frail and prefrail seniors are a group that would likely benefit from preventive nutrition and exercise interventions, both of which can take advantage of ICT.
The objective of the study was to quantify the differences in ICT use, attitudes, and reasons for nonuse among physically frail, prefrail, and nonfrail home-dwelling seniors.
This was a population-based questionnaire study on people aged 65-98 years living in Northern Finland. A total of 794 eligible individuals responded out of a contacted random sample of 1500.
In this study, 29.8% (237/794) of the respondents were classified as frail or prefrail. The ICT use of frail persons was lower than that of the nonfrail ones. In multivariable logistic regression analysis, age and education level were associated with both the use of Internet and advanced mobile ICT such as smartphones or tablets. Controlling for age and education, frailty or prefrailty was independently related to the nonuse of advanced mobile ICT (odds ratio, OR=0.61,
Physical frailty status is associated with older peoples’ ICT use independent of age, education, and opinions on ICT use. This should be taken into consideration when designing preventive and assistive technologies and interventions for older people at risk of health impairment.
The growing role of information and communication technologies (ICT) in our daily lives has led to concerns of increasing inequality between those who can and those who cannot take advantage of new technologies. The divide has been observed not only between younger and older people [
ICT use can also be beneficial for older adults. Use of ICT for direct interaction with people or indirectly for other tasks has also been shown to contribute to the well-being and quality of life of older persons [
The question remains, however, how greatly the digital divide among older adults hampers the potential reach of beneficial e-services such as eHealth within the groups that would most benefit from them. Previous studies on the ICT use of older adults with impaired health have used rather crude markers such as existing disability or a medical condition [
Frailty, as a medical concept, is the loosely defined collection of physiological changes that results in an increased risk of adverse effects in response to stressors. Frail persons are at an elevated risk of falls, declining mobility or disability in daily activities, hospitalization, and death. For a frail person, any minor infection, trauma, or other events can cause a major change in health status and result in loss of independency [
In addition to the general benefits of ICT use for seniors, it could be utilized in various ways in the care of frail seniors in particular. Nutrition such as caloric and protein support [
In this population-based study, we aimed to compare the frail, prefrail, and nonfrail senior citizens regarding (1) ICT use, defined as use of Internet or advanced mobile devices (smartphones and tablet computers) and (2) attitudes toward the use of advanced mobile devices. We hypothesized that there would be lower ICT use among the frail seniors, and that the difference in ICT use would not be explained by age and education alone. Our aim was to produce new knowledge that can be utilized when developing effective interventions for persons at risk of impaired health and loss of independence.
The study was conducted from November 2014 to January 2015 in Oulu. Oulu is Finland’s fifth largest city with a population of approximately 200,000, and 14% of the population is aged 65 years or more [
The study protocol was approved by the Ethics Committee of Human Sciences in the University of Oulu (statement 6/2014). Questions on ICT use, health, lifestyle, and sociodemographic factors were posed as described in the following sections. The study was part of the GASEL (Gamified Services for Elderly) project.
Operational indicators of frailty are usually based on weight loss, fatigue, slowness, low physical activity, number of illnesses, and physical attributes such as low grip strength, inability to rise from a chair, or walk upstairs [
To identify the frail study participants, we used a modified version of the 3-item SOF (Study of Osteoporotic Fractures) index, validated for the prediction of falls, disability, fracture, and mortality [
The modified frailty index.
SOFa index [ |
Modified item | Reason for modification |
Weight loss of 5% in 3-4 years | At least 1 kg weight loss in 3 months | Easier to remember. Part of other scales used in the GASEL study. |
Inability to rise from a chair 5 times without using arms | Inability to rise from a chair without using arms, as identified by “Can you rise from a chair independently and without using hands?” answered on a 5-point scale. | Risk of falling if the participants do the test unattended at home. |
Poor energy as identified by “No” to “Do you feel full of energy?” | Poor energy as identified by “Which of the following best describes how energetic you have felt in the last month?” and response of “I feel moderately tired, exhausted or weak” or worse. | Fitting in with other scales. Adapted from the 15D instrument [ |
aSOF: Study of Osteoporotic Fractures.
Internet use in the last 3 months was assessed with the question “How often on average did you use the Internet in the last 3 months?” with 3 frequency response options, and 1 “Not at all.” Internet access at home was assessed with the question “Do you or anyone in your household have access to the Internet at home? (by any device)” with the options “Yes,” “No,” and “I don’t know.” These questions are slightly modified for brevity from the ones used in Eurostat surveys [
Use of specific devices was asked with the question “In the last 12 months have you used following devices” with the response options “Yes, without major difficulties,” “Yes, with difficulties,” “No,” and “I don't know.” The value of differentiating between use with difficulty and use without difficulty has been previously shown in the development of the everyday technology use questionnaire [
Smartphone use was asked with the question “Have you used a mobile phone with a touch display?” Although not all smartphones are touch phones or vice versa, this simplification made responding easier for the less technologically inclined.
Technology acceptance was asked based on the constructs of the Unified Theory of Acceptance and Use of Technology (UTAUT) [
UTAUT includes 3 constructs that predict behavioral intention to use the specific technology. Performance expectancy (PE) is defined as the degree to which technology provides benefits to users in performing certain activities; effort expectancy (EE) refers to the degree of ease associated with the use of technology; and social influence (SI) refers to the extent to which users perceive that important others (eg, family and friends) believe they should use a particular technology [
We used UTAUT as a list of factors affecting technology adoption of mobile technologies such as smartphones or tablets. Facilitating conditions could not be assessed because of the many possible types and the length of the questionnaire. Outside of the questions rooted in the UTAUT framework, the questionnaire was supplemented with a general interest question, a question regarding cost, and a question about privacy concerns. Only 1 question per determinant was included and answered with “Yes,” “No,” or “I don’t know.” For regression analyses, “No or I don’t know” was used as the reference category [
The type of habitation of the participants was assessed with the question “Which of the following best describes your current form of residence?,” with 3 options corresponding to single- and multistory buildings with or without elevator, and one “assisted living building, rehabilitation center, or nursing home.”
Economic situation was assessed with the question “Is your household income sufficient for your needs?” with the options of “Yes, very well”; “Yes, reasonably well”; “Barely”; and “No (I am using up my savings or being supported by my friends and family).” The latter 2 were classified in the analyses as financial concerns.
Sensory problems were assessed with the dichotomous questions “Do you have difficulties in close activities (eg, handcrafts, reading) due to poor vision?” and “Do you have difficulties hearing (eg, using the phone)?” with the guidance to answer according to experiences in daily life, using the assistive devices usually used.
Mobility outside of the house was assessed with a combination of 2 questions. The Mini Nutritional Assessment (MNA) embedded within the survey included the question “Which of the following best describes your current mobility?” The response options were “I am bed or chair bound”; “I am able to get out of bed/chair, but do not go out”; and “I am able to move outside the house.” This was supplemented with the question “Do you use mobility aids?” with a list of various mobility aids and the multiple options “Indoors,” “Outdoors,” and “Not at all.” The first 2 MNA response options were considered to indicate indoors only, with the exception of a “Bed or chair bound” person who used mobility aids outdoors.
Loneliness was assessed with the 6-item Gierveld scale, which scores emotional and social loneliness on a scale of 0-6 [
Medical conditions were asked with a dichotomous answer to a list of possible diagnosed conditions. Dementia was asked separately as “mild dementia” or “moderate or severe dementia.” In cases where a respondent had answered at least one medical condition question, it was assumed blank answers correspond to “No.”
Education was asked on a 4-level scale, with the options “Less than primary education”; “Primary education”; “High school or college”; and “University degree or similar.” High school or above was considered higher education.
Participation on organized activities outside of the home environment was asked with the question “Do you participate in the activities of a club, union, society, hobby group or spiritual or religious society (eg, sports group, political party, choir, congregation)?” with the answer options “Yes, frequently,” “Yes, occasionally,” and “No.”
The data of each respondent were included in this study if the respondent (1) answered to all the frailty index questions; or scored 2 points in the index (answer to the additional question would not have changed the classification); (2) did not have moderate or severe dementia; and (3) lived in an environment that was not an assisted living facility or nursing home.
Kendall tau was used for crosstabs with ordinal variables and all 3 responder groups. Chi square tests were used to analyze the statistical significance of the differences in dichotomous frailty categories (“nonfrail” and “frail or prefrail”; or “nonfrail or prefrail” and “frail”). Independent-samples Kruskal–Wallis test was used to evaluate the significance of differences in continuous variables (age, number of daily medications) between the groups.
Univariate associations between explanatory variables and ICT use were first analyzed using cross-tabulation. Factors associated with ICT use in univariate analyses were entered into the multivariable logistic regression analysis. Age was first entered into the model, followed by the significant variables using forward stepwise (likelihood ratio) method as a second block. Two models were built and tested, 1 for each of the 2 different dichotomous classifications of frailty.
All statistical analyses were performed using SPSS Statistics version 22, 64-bit edition (IBM). The results were presented as odds ratios (ORs) with 95% CIs. The statistical significance was set at
By January 2015, 918 responses had been received, resulting in a response rate of 61.2%. After checking against inclusion criteria, a total of 794 responses (52.9% of the original sample) were included in the analysis. The responders were approximately 2.5 years younger than nonresponders or those who were excluded (
Age (
The use of ICT was different across the frailty groups. Frailty level was significantly and inversely associated with having Internet connection at home (78.6% of nonfrail, 70.2% of prefrail, 46.4% of frail,
In univariate analyses, age, higher education, prevalence of mild dementia, poor near vision, and dichotomous frailty status were associated with Internet use, whereas gender was not (
Among those participants who had used the Internet during the last 12 months (n=556), there were no statistically significant differences in the type of Internet use between the different frailty level groups. The most common types of Internet use were e-services such as bank, social services, taxes, and tickets (86.2%); information such as timetables, health information, or recipes (83.3%); news (81.1%); communication such as email or Skype (70.1%); entertainment such as movies, books, and music (40.6%); and gaming (34.9%). Shopping, hobbies or studying, social media, and following or posting on forums were less frequent uses.
Socioeconomic predictors of Internet use in the last 3 months among Finnish seniors aged 65+ years, according to multivariable regression analysis adjusted for age (n=738).
Socioeconomic predictors | Model 1a | Model 2b | ||||
Odds ratio | 95% CI | Odds ratio | 95% CI | |||
Age (years) | 0.87 | 0.84-0.90 | <.001 | 0.87 | 0.84-0.90 | <.001 |
High education | 5.14 | 3.28-8.07 | <.001 | 4.97 | 3.17-7.80 | <.001 |
Poor near vision | 0.65 | 0.45-0.94 | .02 | 0.67 | 0.47-0.98 | .04 |
Frailty | 0.45 | 0.22-0.91 | .03 |
aNonfrail or prefrail is the reference category.
bFrail is the reference category.
The majority of all responders, including the frail participants, used a regular mobile phone. Smartphones and tablets were used by less than a third of the responders and a small minority of frail persons, and the frail group experienced more difficulties in the use of all items. Only a small minority of responders in any group used a computer or mobile phone designed for seniors or other users with special needs (
In univariate analyses, higher education, prevalence of mild dementia, and dichotomous frailty levels were all associated with use of advanced mobile ICT (smartphones or tablets), but gender or poor near vision was not (
When exploring opinions of advanced mobile ICT in ordinal analyses including all 3 response options, we found that nonusers had more negative opinions on performance expectancy (
The frail nonusers in particular were more negative than nonfrail nonusers on opinion scales (
Socioeconomic predictors of the use of advanced mobile information and communication technologies (ICT) in the last 12 months among Finnish seniors aged 65+ years, according to multivariable regression analysis adjusted for age (n=664).
Socioeconomic predictors | Odds ratio | 95% CI | |
Age (years) | 0.92 | 0.89-0.95 | <.001 |
High education | 2.64 | 1.89-3.69 | <.001 |
Frailty or prefrailty | 0.61 | 0.42-0.89 | .01 |
Attitude predictors of the use of advanced mobile information and communication technologies (ICT) in the last 12 months among Finnish seniors aged 65+ years, according to multivariable regression analysis adjusted for age and education level (n=626).
Attitude predictors | Odds ratio | 95% CI | |
Age (years) | 0.95 | 0.92-0.98 | .002 |
High education | 1.90 | 1.31-2.77 | .001 |
Performance expectancya | 2.56 | 1.70-3.98 | <.001 |
Effort expectancya | 2.39 | 1.48-3.86 | .001 |
Frailty or prefrailty | 0.60 | 0.40-0.92 | .02 |
aReference category is those who answered “No” or “I don’t know.”
Potential factors on the adoption of advanced mobile ICT, including questions regarding UTAUT performance expectancy (PE), effort expectancy (EE), and social influence (SI). Users: nnonfrail=227; nprefrail=57; nfrail=7. Nonusers: nnonfrail=254; nprefrail=101; nfrail=38.
Frail seniors represent an important target group for ICT-assisted interventions for the elderly. In this study, we, for the first time, compared the ICT use between frail and nonfrail older people at the population level. The observed lesser ICT use among the frail or prefrail people was not explained by age, education, memory disorders, visual impairments, or opinions on performance or effort to use.
Frailty and low ICT use share risk factors. In this study, the association of Internet use and frailty was only partially explained by the different demographics. The types of Internet use between the frail and nonfrail people were similar. The association of lower use of advanced mobile ICT and frailty was not explained by demographics or attitudes toward the technology.
In general, the frail nonusers were more negative toward advanced mobile ICT than the nonfrail nonusers; there was no such difference in attitudes among users. While the frail respondents reported more difficulties with making ends meet financially, cost was not rated as an obstacle to ICT use among frail nonusers.
Our findings are in agreement with previous research results on the prevalence of frailty and general ICT use of the senior population. The prevalence of frailty in this study (7.1%) is similar to previous estimates of 8.6-12.4% of community-dwelling people aged more than 65 years in northern Europe [
Multiple previous studies have looked into the ICT use of seniors with impaired health, but using different definitions of health. The presence of disability has been shown to be associated with lower Internet use, having fast Internet access at home, less ownership of major digital devices, and less use of email or SMS messages [
The original TAM model [
Our findings offer further evidence that the correlation between low ICT use and health impairment reflects health status and risk, and not simply disability, self-perception, or the personality of the respondent. While older people at large are increasingly using computers, mobile technologies, and the Internet, it is still difficult to reach those most in the need of health care support through such means. The observed similar attitudes among users of mobile ICT regardless of frailty status, as well as the similar uses of the Internet, suggest that there are barriers of entry, rather than different needs. Previous research has divided barriers of ICT adoption with regards to eHealth into (1) physical and psychological attributes; (2) provision; (3) support from others; and (4) economic barriers [
The response rate was high in this study; however, the survey was voluntary. It is likely that the frail are underrepresented in the sample, compared with the general home-dwelling senior population. In particular, the group of frail seniors using advanced mobile ICT was very small, which limited the statistical analyses.
We modified the UTAUT survey for the purposes of this study. The UTAUT model has been built for use with specific use cases rather than broad categories such as “mobile ICT,” and we did not use the standard question set. Thus, the results should only be used for the context of this paper, not as validation of the UTAUT model for this purpose.
Due to the constraints of the survey format and the large number of questions, there were some limitations in the depth of assessment. In particular, controlling for more detailed assessments of cognitive functioning, depressive symptoms, and social network would have been interesting, and should be taken into account in future studies.
This study provides early evidence that community-dwelling nonfrail and frail older people are highly different as ICT users, both in actual use rate and in opinions affecting adoption among the nonusing population. These differences are only partly explained by more advanced age, lower education, and higher prevalence of visual or cognitive impairments. As such the research also highlights the clustering of various psychological and social risk factors of social exclusion to the same physically vulnerable population, and the continuing need of measures to prevent an increasing digital divide in society. Future research is needed to further assess the ICT-related needs, barriers to adoption, and facilitating conditions of this interesting and important population.
Characteristics of the study participants according to frailty status.
Socioeconomic predictors of Internet use in the last 3 months among Finnish seniors aged 65+ years in univariate regression analysis.
Information and communication technologies (ICT) device use during the last 12 months and difficulties among users across the frailty categories.
Socioeconomic predictors of the use of advanced mobile information and communication technologies (ICT; smartphones or tablets) in the last 12 months among Finnish seniors aged 65+ years in univariate regression analysis.
Attitude predictors of the use of advanced mobile information and communication technologies (ICT) in the last 12 months among Finnish seniors aged 65+ years in univariate regression. “I don’t know” answers counted as negative.
effort expectancy
facilitating conditions
information and communication technologies
performance expectancy
social influence
Unified Theory of Acceptance and Use of Technology
Study of Osteoporotic Fractures
This study was performed as part of the GASEL project, funded by the Finnish Funding Agency for Innovation (Grant No. 40270/13), BelleGames Oy, BonWell Intelligence Oy, Caritas foundation, Iscom Oy, Mawell Oy, Northern Ostrobothnia Hospital District, Luleå University of Technology, and Oulu Deaconess Institute. The study was also supported by the ASTS joint project by Academy of Finland (Grant No. 270816) and Japan Science and Technology Agency; the iKaaS joint project by the European Union (Grant No. 643262) and Ministry of Internal Affairs and Communication, Japan; Infotech Oulu; and the Finnish Cultural Foundation.
None declared.