Abstract
Findings indicate that occupational therapy practitioners should consider executive functioning and dexterity of older adults with Type 2 diabetes in future mobile self-management programs.
The burden of chronic disease on health care resources and costs can be reduced by using self-management, which has been described as what a person does to manage a disease (American Diabetes Association, 2017; De Silva, 2011). Many self-management programs for people with diabetes have been developed and studied (Simmons et al., 2013). These programs include recommendations on monitoring blood sugar and medications and adopting a healthy lifestyle.
Type 2 diabetes is characterized by hyperglycemia that is caused by a relative lack of insulin, insulin action, or both and is associated with a high risk of serious chronic disease (Longo et al., 2011; Melmed et al., 2011). Diabetes is a risk factor for eye and kidney disease, neuropathy, and cardiovascular and cerebrovascular morbidity and mortality (Longo et al., 2011; Melmed et al., 2011). It is also a risk factor for cognitive dysfunction, dementia, and sarcopenia, with overall reduced quality of life (Umegaki et al., 2017). Self-management programs may reduce risk factors, improve health, prevent secondary complications, maintain independence in daily living, and enhance the quality of life and participation of older adults with diabetes (Hunt et al., 2014).
An increasing number of self-management programs can now be facilitated by the use of mobile devices (i.e., smartphones and touchscreen tablets), which offer the opportunity to use a wide range of applications (apps), the majority of which can be downloaded for free. Mobile devices have been used in research for diverse health purposes such as self-management and monitoring of blood sugar for people with diabetes, memory enhancement and wandering safety for people with dementia, symptom management for people undergoing chemotherapy, and detection of early symptoms of disease to prevent hospitalization in people with congestive heart failure (Ventola, 2014).
The use of mobile devices entails sensorimotor abilities for finger and hand movements on the touchscreen (gestures) and cognitive abilities to understand and learn how to operate apps (Givon Shaham et al., 2018). Older adults with diabetes might have difficulty using touchscreen tablets as a result of neuropathy or cognitive dysfunction and therefore might not be suitable candidates for tablet self-management. Thus, the aim of this study was to explore and identify the motor and cognitive abilities associated with touchscreen tablet app performance of older adults with diabetes. Results might help pave the way for identification of which older adults with diabetes would be suitable candidates for using touchscreen tablet apps for self-care management.
Method
This cross-sectional study was conducted as part of a multidisciplinary evaluation at the Center for Successful Aging With Diabetes (Ramat Gan, Israel). Participants were either self-referred or referred by their treating physician as a result of difficulties in managing their disease.
Participants
Forty-five older adults with diabetes were included in this study according to the following inclusion criteria: age 60 yr or older with a diagnosis of Type 2 diabetes and able to understand and read Hebrew. Those with a significant visual, hearing, motor, or cognitive impairment that may have precluded neuropsychological testing and self-report questionnaires were excluded. For this exploratory study, we applied rules of thumb to determine our sample size (i.e., a ratio of 10 participants for each variable in the model; Wilson VanVoorhis & Morgan, 2007).
Measures of Sensorimotor Abilities
The Purdue Pegboard Test (PPT; Tiffin & Asher, 1948) was used to assess dexterity of both hands. The pegboard has holes arranged in two parallel rows; participants are given 30 s to insert as many pegs as possible into the pegboard under three conditions (tasks): using the dominant hand, using the nondominant hand, and using both hands together. Under a fourth condition, a 60-s assembly task, participants are requested to work continuously and simultaneously with both hands to insert pegs, washers, and collars into the pegboard. The score for each task is the number of inserted pegs, collars, and washers. The PPT is a valid and reliable test for assessing dexterity (Ben Shahar et al., 1998; Causby et al., 2014; Tiffin & Asher, 1948) with normative data (Hamm & Curtis, 1980; Mathiowetz et al., 1984).
Semmes–Weinstein monofilaments (North Coast Medical, 2011; Semmes et al., 1960) were used to assess level of touch sensation by ability to feel the monofilament on the skin. Median and ulnar nerve distributions of each hand were assessed. The shortened version of the test consists of five filaments: 2.83-mm diameter (normal sensation), 3.61-mm diameter (diminished superficial sensation), 4.31-mm diameter (diminished protection sense), 4.56-mm diameter (loss of protection sense), and 6.65-mm diameter (lack of a superficial sensation). This test is reliable and valid for assessing sensation (Bell-Krotoski et al., 1995; Bell-Krotoski & Tomancik, 1987; Radomski & Latham, 2014).
The B&L Engineering® (Santa Ana, CA) pinch gauge was used to measure pinch strength (in kilograms) in three positions—lateral pinch, tip pinch, and tripod pinch—while the participant was seated in a standard position (shoulders in adduction, elbow at 90° of flexion, forearm in a natural position, and wrist at 0°–15° of extension; Mathiowetz et al., 1984). The mean of three trials was calculated for each pinch position.
Measures of Cognitive Abilities
The Montreal Cognitive Assessment (MoCA; Nasreddine et al., 2005) is a cognitive screening tool that assesses eight cognitive domains: Attention and Concentration, Executive Functions, Memory, Language, Visuo-Constructional Skills, Conceptual Thinking, Calculations, and Orientation. Scores range from 0 to 30 points. Scores above 25 are considered normal cognition, and scores between 19 and 25 indicate mild cognitive impairment (Lifshitz et al., 2012; Nasreddine et al., 2005).
The Trail Making Test (TMT; Reitan, 1958) is a widely used valid and reliable (Sánchez-Cubillo et al., 2009) neuropsychological paper-and-pen assessment that measures speed of processing, sequencing, and mental flexibility and is therefore considered to measure executive functioning. The test consists of two parts: Part A (TMT–A) requires the examinee to use a pen to sequentially connect numbers from 1 to 25, and Part B (TMT–B) requires the examinee to alternately connect 13 numbers and 13 letters in numerical and alphabetical order, respectively. The scores for each part include the time in seconds needed to complete the test, including error correction time (Bowie & Harvey, 2006).
Apps
Two touchscreen tablet apps were used. The Dexteria app (BinaryLabs, San Diego, CA) consists of three tasks (“tap it,” “pinch it,” and “write it”). The tap-it task was used in this study. This task requires isolated finger movements in which the participant holds the thumb on a virtual anchor (a blue circle) and touch targets (colored shapes) with the other four fingers. The targets appear and disappear in different locations on the screen (for different fingers). The targets disappear after a touch or after a few seconds without the correct response. Participants performed this task twice with each hand, and the average completion time and percentage of accuracy for each hand were calculated.
SuCare (Sanofi-Aventis U.S., Bridgewater, NJ) is a diabetes self-management app. It entails a menu screen to fill in the following information in the indicated windows: the date and time, the value of blood sugar of the current testing, and a reminder for the next sugar testing. Participants were asked to enter the level of sugar measured at the time of the assessment and to set a reminder for the next sugar testing. An occupational therapist observed each participant’s performance and scored it according to the amount of cues (help) needed to complete the task (more cues indicate less independence). Cueing was based on the cueing guidelines of the Executive Function Performance Test (EFPT; Baum et al., 2003), which is a performance-based assessment to assess executive functioning. Scores can range from 0 (was independent and did not need cues) to 5 (done by the therapist) points. Performance time (in seconds) was also recorded.
Demographics and Medical Information
Medical information was obtained from an interview, a physical examination, and medical records. Data regarding weight, hypertension, smoking status, and dyslipidemia status (lipid profile conducted routinely every several months in people with diabetes) were collected. The severity of diabetes was measured using glucose control as assessed by the hemoglobin A1c (HbA1c). Normoglycemia is defined as a value below 5.7%, and diabetes is defined as a value above 6.5% (American Diabetes Association, 2017). Diabetes duration and diabetes complications (retinopathy, nephropathy, neuropathy, cerebrovascular disease, cardiovascular disease, peripheral vascular disease) were registered. In addition, participants filled in a sociodemographic questionnaire that included information such as age, education, and prior use of mobile devices.
Procedure
This study was approved by the Sheba Medical Center (Ramat Gan, Israel) research and university ethics committee. After signing the informed consent form, participants underwent the clinical assessments, which were conducted by an experienced occupational therapist (Neta Kravitz) and took approximately 1 hr.
Statistical Analysis
Descriptive statistics were used to describe participants’ demographic and diabetes-related information. The Shapiro–Wilk test confirmed that the dependent variables were normally distributed (p > .05); therefore, parametric statistics were used. Pearson correlations were used to assess the associations between app performance and sensorimotor and cognitive abilities. Correlations ranging from .25 to .49 are considered fair and those from .50 to .75 are considered moderate to good (Portney & Watkins, 2015). Spearman correlations were used to assess the associations between cueing (Levels 0–5), an ordinal variable, and performance time in the SuCare app and other research variables.
A multiple linear regression model (enter method in blocks) was used for each app (SuCare; Dexteria tap-it, dominant hand; and Dexteria tap-it, nondominant hand) to determine the contribution of the sensorimotor and cognitive assessments for explaining app performance after controlling for age and diabetes severity (HbA1C). To ensure that the assumptions of multiple regressions were met, we inspected scatterplots of residuals against the model data as well as outliers and influential data points and the variance inflation factor for multicollinearity. All analyses were conducted using IBM SPSS Statistics (Version 24.0, IBM Corp., Armonk, NY).
Results
Study Sample
Forty-five participants (23 women and 22 men; M age = 71.75 yr, SD = 6.29) were included in this study. Participants’ mean diabetes duration was 16.6 yr (SD = 9.04), and their mean HbA1C level was 7.87% (SD = 1.67; Table 1). The majority of the participants (68.9%) reported daily use of a smartphone. Participants varied in their cognitive abilities and executive functioning on the basis of MoCA and TMT scores (see Table 1).
Participant Information (N = 45)
Note. CVA = cerebrovascular accident; HbA1C = hemoglobin A1C; MoCA = Montreal Cognitive Assessment; TMT–A = Trail Making Test–Part A; TMT–B = Trail Making Test–Part B.
The sensorimotor differences between participants’ dominant and nondominant hands appear in Table 2. Significant differences between hands were found for tripod and tip pinch strength. In addition, diminished light touch (as assessed with the monofilaments) was found among 25 participants (55.6%) on the ulnar distribution of both hands.
Comparison of Sensorimotor Measures and App Performance Between the Dominant and Nondominant Hand (N = 45)
Note. PPT = Purdue Pegboard Test; — = not applicable.
p < .05. ** p < .01
Significant moderate correlations were found between SuCare and Dexteria performance and the following variables: age, diabetes severity, cognitive abilities and executive functioning, and dominant and nondominant hand dexterity (Table 3). Sensation, pinch strength, education, and diabetes duration were not found to significantly correlate with app performance.
Correlations Between App Performance and Completion Time and Age, Diabetes Severity, and Sensorimotor and Cognitive Abilities
Note. D = dominant hand; HbA1C = hemoglobin A1C; MoCA = Montreal Cognitive Assessment; ND = nondominant hand; PPT = Purdue Pegboard Test; TMT–B = Trail Making Test–Part B; — = not applicable.
p < .05. **p < .01.
During SuCare performance, participants were assisted in completing the task through cueing. Six participants (13.3%) scored 0 (were independent and did not need cues), 16 (35.6%) scored 1 (needed verbal general cues), 8 (17.8%) scored 2 (needed gestural cues), 9 (20%) scored 3 (needed direct verbal cues), and 6 (13.3%) scored 4 (needed physical assistance to complete the SuCare task). A significant moderate correlation (r = .642, p < .01) was found between the level of cueing and SuCare task completion time, indicating that participants who demonstrated slower SuCare performance needed more cueing. In addition, moderate significant correlations were found between the level of cueing and age (r = .576, p < .01) and level of cueing, cognitive abilities, and executive functioning as measured with the MoCA (r = –.527, p < .01) and TMT–B (r = .597, p < .01).
In other words, older participants and participants with cognitive and executive functioning deficits required more cues to complete the SuCare app. Previous experience using a mobile device was moderately significantly correlated with the level of cueing (r = .46, p < .001) and time to complete (r = .44, p < .001) the SuCare app. The level of cueing was weakly correlated to dexterity of the dominant hand (r = –.37, p < .001; measured with the PPT), indicating that participants who inserted fewer pegs (less dexterity) needed more cues.
On the basis of the significance and strength of these correlations, executive functioning (TMT–B) and dexterity (PPT) were entered into the linear regression model to determine their contribution to the variance in performance time for the SuCare and Dexteria (dominant and nondominant hand) apps. To control for age and diabetes severity (HbA1C), these variables were entered into the model first (see Table 3). Prior mobile device use was also considered for the SuCare model to determine its contribution to performance time on this app.
Dexteria With the Dominant Hand
HbA1C and age accounted for 9.7% and 29.8%, respectively, of the total variance in Dexteria performance time with the dominant hand. The addition of executive functioning (TMT–B) was not significant in Dexteria performance time, but dexterity of the dominant hand (PPT) accounted for an additional 5.4% of the total variance of 45.1%, F(4, 40) = 10.021, p < .001 (Table 4).
Contribution of Executive Functioning and Dexterity to Performance Time for SuCare, Dexteria Dominant Hand, and Dexteria Nondominant Hand
Note. D = dominant hand; HbA1C = hemoglobin A1c; ND = nondominant hand; PPT = Purdue Pegboard Test; SE = standard error; TMT−B = Trail Making Test−Part B.
Dexteria With the Nondominant Hand
HbA1C and age accounted for 6.2% and 31.8%, respectively, of the total variance in Dexteria performance time for the nondominant hand. The addition of executive functioning (TMT–B) accounted for an additional 9.8% of the total variance of 43.9%, F(4, 40) = 9.429, p < .001. The contribution of dexterity was not significant in Dexteria performance time with the nondominant hand (see Table 4).
SuCare
Prior mobile device use, age, and HbA1C accounted for 6.4%, 13.8%, and 26.4%, respectively, of the total variance of SuCare performance time. The addition of executive functioning (TMT–B) and dexterity of the dominant hand (PPT) accounted for an additional 9.5% and 9.4%, respectively, of the total variance of 61.0%, F(5, 39) = 14.75, p < .001 (see Table 4).
Discussion
Optimal diabetes care requires patients to make daily therapeutic decisions based on information about their condition that they collect and process. Self-management programs enhance optimization of diabetes care, which may reduce risk factors and secondary complications, improve health, and help maintain independence in daily living (Grady & Gough, 2014). Self-management can be facilitated by the use of mobile devices (smartphones and tablets), especially because an increasing number of older adults have reported using this technology for daily living (Gell et al., 2015). Therefore, this study aimed to explore and identify the sensorimotor and executive functioning and cognitive abilities that are required in the use of touchscreen tablet apps among older adults with diabetes. This information can guide clinicians when aiming to determine suitable candidates for self-management using touchscreen tablet apps.
Chronic exposure of the brain to high levels of glucose may accelerate cognitive decline (Cukierman-Yaffe et al., 2009; Xia et al., 2013) and lead to executive functioning deficits, such as difficulties in decision making (Miranda-Félix et al., 2016). Therefore, people with diabetes are approximately 1.5 times more likely to experience cognitive decline than those without diabetes (Cukierman-Yaffe et al., 2009). Our cohort varied in their cognitive abilities and executive functioning, with 43.3% of the participants with a MoCA score below 24 points indicating mild or moderate cognitive deficits (Nasreddine et al., 2005). Neuropathies such as sensory impairments are also common in older adults with diabetes (Ennis et al., 2016), and 48.9% of our cohort had neuropathy, as assessed with the Semmes–Weinstein monofilaments.
Sixty-nine percent of the 45 older participants with Type 2 diabetes reported that they use smartphones or touchscreen tablets for a variety of purposes, which highlights the relevance of this study. All participants were able to use and complete the Dexteria app with their dominant and nondominant hands. Participants also succeeded in completing the SuCare app, which requires self-testing of blood glucose and adding a reminder for the next testing. However, more than 80% of participants needed some therapist support (cueing), which was highly correlated with cognition and executive functions and moderately correlated with prior mobile device use. These findings emphasize the need for professional guidance in implementing a successful self-management program for people with diabetes and cognitive decline and for their prior mobile device use to be taken into consideration.
App performance was significantly correlated with age, diabetes severity, dexterity, and cognitive and executive functioning abilities. For both apps, including Dexteria with the dominant and nondominant hand, age contributed to performance by 13.8% to 31.8%. Similar findings regarding the influence of age and executive functioning were reported for three touchscreen tablet app puzzles for older adults without diabetes compared with younger adults (Givon Shaham et al., 2018). The older adults completed fewer puzzle levels and had longer completion times than the younger adults, which was in accordance with their executive functioning, as assessed using the TMT–B (Givon Shaham et al., 2018). In another study conducted by our group (Kizony et al., 2016), 172 people without a disability from three age groups performed the Nine-Hole Peg Test to assess dexterity and used the tap-it Dexteria app. Significant differences (p < .001) in the time and accuracy of the dominant and nondominant hands for tapping task performance were found between the young adults (n = 79, M age = 26.2 yr) and the middle-aged adults (n = 61, M age = 55.9 yr) and the older adults (n = 32, M age = 68.7 yr).
In this study, app performance was not correlated with participants’ sensory impairments, possibly because the majority of them detected the normal touch or diminished light touch monofilaments (Shah et al., 2015). The mean pinch strength of this cohort was also close to the norms of the general population and was therefore also not found to be correlated to app performance. Moreover, the gestures used for the app do not require pinch strength such as is necessary for other daily functions (e.g., using a key, holding and zipping a zipper).
The linear regression model revealed that app performance was explained by diabetes severity. HbA1C levels contributed 26.4% to the total variance of SuCare performance and 9.7% and 6.2% of Dexteria performance with the dominant and nondominant hand, respectively. These findings demonstrate the strong association between diabetes severity and app performance that specifically requires executive functioning (i.e., SuCare) and highlight the importance of self-management of the disease symptoms for this population. The addition of cognitive deficits to the analysis contributed an additional 9.5% and 9.8% to the total variance of SuCare and Dexteria nondominant-hand performance, respectively. Interestingly, cognitive deficits did not significantly explain the variance in Dexteria dominant-hand performance, possibly because isolated movements of the dominant hand are more automatic and easier to perform than such movements of the nondominant hand (where planning is required).
The limitations of this study include the use of participants who are older adults and are relatively high functioning in their everyday life. Therefore, these findings might not generalize to people with diabetes who are less independent in daily living. A larger sample would have allowed us to investigate additional factors that may influence app performance, such as years of education and literacy level. Prior use of mobile devices should be taken into consideration when planning an intervention that includes such devices because participants with no prior experience may need a longer learning process. Future research should include these variables and use experimental designs to assess the feasibility and effectiveness of using touchscreen tablets for self-management of older adults with diabetes.
Implications for Occupational Therapy Practice
The results of this study have the following implications for occupational therapy practice:
Many older adults already use mobile devices for daily living; therefore, clinicians can have them use tablet apps for self-management of Type 2 diabetes to improve health and prevent secondary complications.
Clinicians should assess higher cognitive abilities and hand dexterity (but not sensation or pinch strength) of older adults to decide who can successfully use mobile devices for self-management of Type 2 diabetes.
Conclusion
Beyond age and diabetes severity, higher cognitive abilities and hand dexterity (but not sensation or pinch strength) contributed to explaining the variance in app performance of older adults with Type 2 diabetes. Because older adults are already using mobile devices for daily living, using tablet apps has potential for future self-management of Type 2 diabetes. Findings from this study can assist clinicians when selecting suitable candidates for self-management tablet programs.
Footnotes
Acknowledgments
We thank all the participants who agreed to participate in the study. Data collection and analysis were made possible by a donation from the Israeli Association for the Study of Diabetes and grants from the Israeli Diabetes Association, the Israeli Ministry of Health’s Chief Scientist Office, and the Otzma Diabetes Care Initiative. This work was performed in partial fulfillment of the requirements for a master of science degree in occupational therapy for Neta Kravitz, Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel. Yafi Levanon and Debbie Rand declare equal contribution to this study.