User characteristics associated with use of wrist-worn wearables and physical activity apps by adults with and without impaired speech-in-noise recognition: a cross-sectional analysis

Abstract

Objective

To study weekly use of smartwatches, fitness watches and physical activity apps among adults with and without impaired speech-in-noise (SIN) recognition, to identify subgroups of users.

Design

Cross-sectional study.

Study sample

Adults (aged 28–80 years) with impaired (n = 384) and normal SIN recognition (n = 341) as measured with a web-based digits-in-noise test, from the Netherlands Longitudinal Study on Hearing. Multiple logistic regression analyses were used to study differences and to build an association model.

Results

Employed adults in both groups are more likely to use each type of fitness technology (all ORs >3.4, all p-values < 0.004). Specific to fitness watch use, adults living with others use it more (OR 2.5, 95%CI 1.1;5.8, p = 0.033) whereas those abstaining from alcohol (OR 0.3, 95%CI 0.1;0.6) or consuming >2 glasses/week (OR 0.4, 95%CI 0.2;0.81, overall p = 0.006) and hearing aid users (OR 0.5, 95%CI 0.2;0.9, p = 0.024) make less use.

Conclusions

Subgroups of adults more and less likely to use fitness technology exist, but do not differ between adults with and without impaired SIN recognition. More research is needed to confirm these results and to develop interventions to increase physical activity levels among adults with hearing loss.

Keywords:

• Hearing impairment
• speech-in-noise
• smartwatch
• fitness tracker
• physical activity app

Introduction

Worldwide, over 1.5 billion people live with some degree of hearing impairment. Its prevalence will rise to 2.5 billion people in 2050 due to an increasing number of adults with age-related hearing loss (ARHL) (GBD Hearing Loss Collaborators 2021). ARHL is associated with the prevalence of chronic health conditions like hypertension, cardiovascular disease, diabetes and cerebrovascular disease (Besser et al. 2018). A number of studies have also shown an association between low levels of physical activity and hearing impairment (Curhan et al. 2013; Gispen et al. 2014; Loprinzi et al. 2014; Haas et al. 2016; Vancampfort et al. 2017; Tsimpida et al. 2019; Wells et al. 2020; Kuo et al. 2021). Thus, many adults do not reach the recommended levels for physical activity, but adults with a hearing impairment appear to be even less active than their normal hearing peers. Given their higher prevalence of chronic conditions, many of which could be alleviated through physical activity, increasing physical activity in adults with impaired hearing is imperative.

Several systematic reviews have shown that physical activity can successfully be promoted using smartwatches, fitness trackers and smartphone applications (Gal et al. 2018; Brickwood et al. 2019; Laranjo et al. 2021), with maintained long-term effects (Mönninghoff et al. 2021). Smartwatches act as an extension to a mobile phone, e.g. by showing call and text notifications, and also track physical activity and other health metrics (Henriksen et al. 2018; Jakeman 2022). Fitness trackers measure physical activity but, depending on the brand and model, can monitor other health metrics as well. Some of the current wrist-worn fitness trackers include smartwatch-like features (Jakeman 2022). To access their full functionality, smartwatches and fitness trackers come with an accompanying smartphone application. Some of these wearables also connect to third party apps such as Strava and Runkeeper (Henriksen, Haugen Mikalsen et al. 2018; Jakeman 2022). There are also autonomous smartphone apps aimed at physical activity. These range from apps with a collection of workout routines and apps that use the smartphone sensors to measure physical activity, to feature-rich e-coaching apps (Mollee et al. 2017). Other lifestyle behaviours, such as diet and sleep, may also be addressed but in this paper, we focus on apps primarily aimed at physical activity. We use the term “fitness watch” for wrist-worn fitness trackers and “fitness technology” to describe both types of wrist-worn wearables and physical activity apps.

New on the fitness technology market are ear-worn devices with sensors that measure vital signs such as heart rate, oxygen saturation, as well as movement (Caduff et al. 2020). Currently, two hearing aid models feature physical activity tracking (Starkey Livio AI: Starkey, Eden Prairie, MN; Phonak Audéo Fit^TM: Sonova AG, Stäfa, Switzerland), but more will be released. At the same time the hearing aid market is changing. Consumer brands like Apple and Samsung are marketing some of their earphones as listening devices and are on course to release over-the-counter hearing aids for mild to moderate hearing impairment, while also considering the inclusion of health sensors in these devices (Winkler 2021). These developments could make hearing aids more attractive and less stigmatising, resulting in earlier adoption of a hearing aid. They could also motivate regular wearing of the hearing aids. Moreover, sensor-enhanced hearing aids and connected apps could be used to stimulate physical activity in adults with impaired hearing, similar to wrist-worn wearables, and may in time support early detection of conditions associated with hearing impairment (Caduff, Feldman et al. 2020).

Hearing aids with connected apps that track and promote physical activity could potentially reach a lot of inactive adults, but their impact depends on actual uptake and use. For that reason, current use of fitness technology by people with a hearing impairment, as well as their characteristics are of interest. This could signify which adults with impaired hearing would acquire sensor-enhanced hearing aids, and who would need more encouragement to obtain them or might never use them. Only a few studies have investigated the use of fitness technology in real world conditions and, to the best of our knowledge, none of these studies have been done in adults with impaired hearing, with the exception of an earlier study by our research group (van Wier et al. 2021). We studied the use of smart devices, apps and social media and found that adults with impaired speech-in-noise (SIN) recognition did not differ from adults with normal SIN recognition in their use of a smartphone or tablet, nor in their weekly use of physical activity apps. So far, we have not compared the use of a smartwatch and fitness watch between these groups, nor studied characteristics of fitness technology users.

The aim of this study was to provide the first data on use of fitness technology among adults with impaired hearing, and on the characteristics of these users. We used data collected in a long running online cohort, the Netherlands Longitudinal Study on Hearing (NL-SH). We had the following research questions (RQs):

RQ1: How many adults with impaired SIN recognition make weekly use of fitness technology, namely a smartwatch, fitness watch or physical activity app, and does this use differ from adults with normal SIN recognition?

RQ2: What user characteristics are associated with weekly use of these types of fitness technology in adults with impaired SIN recognition and do their characteristics differ from adults with normal SIN recognition who use this technology?

Materials and methods

Study sample and data collection procedures

The NL-SH is an ongoing prospective cohort study in which both normally hearing and adults with hearing impairment aged 18–70 years at inclusion participate. Inclusion started in 2006, and still continues. Participants are measured every five years using an online questionnaire and SIN test. For the current study, the 10-year follow-up data collected up to 1 April 2020 were used. Participants were included if they had complete data. This concerns the same sample used in our previous study (van Wier, Urry et al. 2021).

The NL-SH study protocol has been approved by the IRB (Medical Ethics Committee; METC) of the Amsterdam Medical Centre, location VUmc in Amsterdam, The Netherlands (METC number 2006/83; NL12015.029.06). All participants provided informed consent.

Questionnaire

User characteristics consisted of demographics and health related items. These characteristics and use of fitness technology were collected through the online questionnaire.

Demographics

Demographics concerned current age, sex, marital status, living situation, education and working status. Marital status was categorised as married (married, registered partnership) and not-married (never married or having a registered partnership, divorced, widowed). Living situation was classified as living with others (with partner, with partner and children, with parents, with children) and living alone. Highest attained education was divided in low (elementary school or attended high school, but no degree), mid (high school graduate or having an associate’s degree), and high (having a bachelor’s degree, master’s degree, or doctoral degree). Working status was classified as employed (paid job; own a business) and other (unemployed and looking for a job; partly unfit to work; fully unfit to work; voluntary work; house wife/man; student; early retired; other situation).

Health-related items

Health-related items concerned hearing aid use, the presence of chronic medical conditions, self-reported health, mobility, visual problems, BMI, smoking and alcohol consumption. Hearing aid use was asked as: “Do you use a hearing aid?” (yes, no).

Chronic medical conditions present in the previous 12 months were assessed by a list of 28 conditions (Mootz and Van Den Berg 1989). For the current study only those conditions which may improve from increased physical activity and those that might hamper physical activity were selected: asthma, COPD; heart disease, myocardial infarct; hypertension; chronic bowel disease; diabetes; thyroid disease; chronic back pain, hernia; osteoarthritis; rheumatoid arthritis (RA), other arthritis; dizziness with falling; migraine; malignancies; depression, burnout, anxiety; and chronic skin disease. Of these, only the conditions present in at least 15 participants of each hearing status group were selected for analysis.

Self-rated health was asked with a Numeric Rating Scale (NRS) with options 0 (worst possible health) to 100 (best possible health). Mobility consisted of no mobility problems and some problems.

Visual problems concerned farsightedness and nearsightedness, corrected with glasses if needed. Self-reported height and body weight were used to calculate BMI.

Smoking was categorised to current smoker (smoke every day, smoke occasionally), former smoker (used to smoke every day, used to smoke occasionally) and never smoker. Alcohol consumption concerned total number of glasses per week.

Fitness technology

From a list of devices, including a tablet, smartphone, smartwatch and a fitness watch, participants ticked those they use at least once a week. A smartwatch was described as “a watch with many functions like a computer, Apple watch or similar” and a fitness watch as “a watch that is worn to, for instance, measure your heart rate or to count steps; Fitbit or similar”. Next, participants who indicated to use a smartphone, smartwatch or tablet were asked what types of apps they use at least once a week, among which (in Dutch) “fitness apps” (hereafter: physical activity apps).

Speech-in-noise test

To measure the ability to understand speech-in-noise, the procedures of the validated National Hearing Test (NHT) were followed (Smits et al. 2006). This test was developed as a functional hearing-screening test. Participants are instructed to perform the test in a quiet room. Hearing aid users are instructed to do the test without their hearing aid(s). Either headphones or speakers are allowed, but participants have to indicate which transducer they use. The test is diotic, presenting identical stimuli binaurally, and predominantly measures the better ear (De Sousa et al. 2020). A total of 23-digit triplets (e.g. 6-2-5) are presented against a background of masking noise in an adaptive manner. The speech-reception-threshold in noise (SRTn) is calculated by taking the average signal-to-noise ratio (SNR) of the last 20 presentations, corresponding to a score of 50% of the presented triplets understood correctly. A higher SRTn indicates worse ability to distinguish speech-in-noise. SRTn values can range from −13.4 to a ceiling level of +4 dB signal-to-noise ratio (dB SNR). A cut-off of −5.5 dB is used to divide the group into adults with sufficient ability to distinguish speech-in-noise (further indicated as normal SIN recognition) and those with insufficient ability to distinguish speech-in-noise (further indicated as impaired SIN recognition). Adjusted for listening with two ears (−1.4 dB SRTn), (Smits, Merkus et al. 2006) this cut-off point agrees with an approximate pure-tone average loss of 23 dB (95%CI 14.1–38.6 dB) at the frequencies 0.5, 1, 2 and 4 kHz (based on Table 3 in (Smits et al. 2004)). The correlation between SRTn and PTA_{0.5, 1, 2, 4} is 0.770 (Smits, Kapteyn et al. 2004). The NHT discriminates between normal hearing adults and those with SNR loss as measured with the sentence SRTn test of Plomp and Mimpen. The NHT has a high sensitivity and specificity of 0.91 and 0.93, respectively, and a measurement error below 1 dB (Smits, Kapteyn et al. 2004).

Statistical analysis

Given the dichotomous outcome, namely use of type of fitness technology yes/no, and the need to include several independent variables in the same model, all RQs were answered using multiple logistic regression. For RQ1, all user characteristics were considered as potential confounders. They were included in the model if (1) the potential confounder had influence (p < 0.10) on both the outcome and the independent variable and (2) the regression coefficient of the independent variable changed by ≥10% after adding the potential confounder to the model. To answer RQ2, all variables were evaluated separately for their association with weekly use of the different types of fitness technology. First an interaction with impaired SIN recognition yes/no was made in each univariable model, to analyse whether this modified their association. In case effect modification for one or more variables was found, a model would be built specific to adults with impaired SIN recognition. If no effect modification was found it meant that user characteristics do not differ between the groups and a model could be built using data from the whole sample. This increases the power of the study compared to only using data from participants with impaired SIN recognition. After the univariable analyses, the characteristics for which the association had p < 0.1 were added to the multiple logistic regression. The final model for RQ2 was built using forward selection. Linearity was checked for all continuous factors. If this assumption was violated, these factors were converted to quartiles and analysed as a categorical factor. Results were considered statistically significant if p < 0.05 or in case the 95% confidence interval (CI) for the odds ratio (OR) does not include the value 1. Analyses were conducted using SPPS Statistics version 26 (IBM Corp., Armonk, NY).

Results

Out of 885 participants responding to the T2 measurements, 725 had complete data. Most participants (620/725; 85.5%; Supplementary Table 1) had filled out the questionnaire between September 2016 and December 2017.

Table 1. Weekly use of fitness technology; descriptive data and results from the comparison between groups with normal and impaired SIN recognition.

	Total group	Normal SIN^a	Impaired SIN	Impaired SIN compared with normal SIN
	N = 725	n = 341	n = 384
	n (%)	n (%)	n (%)	OR	95%CI	p-Value
Smartwatch	18 (2.5)	8 (2.3)	10 (2.6)	1.46^c	0.44; 3.82	0.437
Fitness watch	55 (7.6)	36 (10.6)	19 (4.9)	0.74^d	0.37; 1.48	0.395
Physical activity apps^b	59 (8.1)	33 (9.7)	26 (6.8)	0.87^e	0.50; 1.51	0.619

^aSIN: speech-in-noise recognition; ^bUse of a physical activity app on a smartphone, tablet or smartwatch; ^cModel adjusted for working status; ^dModel adjusted for working status, age and hearing aid use; ^eModel adjusted for working status and age.

The majority of participants were women (442/725, 61.0%). Mean age was 57.7 (SD 11.4) years and 60.0% (435/725) had a high level of education (Supplementary Table 2). Of the total group, 384/725 (53.0%) were classified as having impaired SIN recognition. Supplementary Table 2 also shows the characteristics of these participants and of the participants with normal SIN recognition separately. Of note is that 46/341 (13.5%) of the latter used a hearing aid. This was 253/384 (65.9%) in those with impaired SIN recognition.

Table 2. Multivariable association model of the odds ratios for fitness technology use.

	Smartwatch		Fitness watch		Physical activity app
	OR (95%CI)	p-Value	OR (95%CI)	p-Value	OR (95%CI)	p-Value
Working situation, employee or business owner vs other	5.69 (1.82; 17.74)	0.003	3.44 (1.81; 6.55)	<0.001	3.87 (2.11; 7.10)	<0.001
Sex, female vs male	0.18 (0.06; 0.52)	0.002	–	–	–	–
Alcohol consumption, reference group Q2 (1–2 glasses/week)	–	–		0.006	–	–
Q1 (0 glasses/week)			0.30 (0.14; 0.64)
Q3 (3–7 glasses/week)			0.36 (0.16; 0.81)
Q4 (8–84 glasses/week)			0.36 (0.16; 0.81)
Hearing aid use	–	–	0.47 (0.24; 0.90)	0.024	–	–
Living situation	–	–	2.50 (1.08; 5.79)	0.033	–	–

RQ1: weekly use of fitness technology

Across all participants, 2.5% (18/725) made weekly use of a smartwatch. This was 7.6% (55/725) for use of a fitness watch whereas 8.1% (59/725) used physical activity apps (Table 1). Weekly use of fitness technology did overlap, with 50.0% (9/18) of smartwatch users and 47.3% (26/55) of fitness watch users using a physical activity app as well (Supplementary Table 3). Six participants (0.8%) were using all three technologies on a weekly basis.

Unadjusted odds for the comparisons between adults with normal and impaired SIN recognition can be found in the tables with the univariable associations of all characteristics with use of fitness technology (Supplementary Tables 4–6). It should also be understood that one of the conditions for possible confounding can also be derived from these tables, namely a possible association of the characteristic with the use of fitness technology when p < 0.10. After adjustment for various confounders, no statistically significant differences were found between the groups in use of a smartwatch, nor a fitness watch and physical activity apps (Table 1).

RQ2: user characteristics associated with weekly use of fitness technology

Analysis of interactions between smartwatch user characteristics and impaired SIN recognition was not possible because the number of smartwatch users was too low. For weekly use of a fitness watch and of physical activity apps no statistically significant interactions of all separately assessed variables with impaired SIN recognition were found. This means that, in our sample, the user characteristics for these types of fitness technology did not differ between adults with and without impaired SIN recognition. It was therefore not necessary to develop a separate model for adults with this impairment.

Analysis of the stand-alone relationship of the various demographic and health characteristics with participants’ use of a smartwatch resulted in two potentially associated factors, namely sex and working status (Supplementary Table 4). Both identified features remained in the final model. Adjusted for working status, female participants had 5.6 times lower odds (OR 0.2, 95%CI 0.1; 0.5, p = 0.002) of using a smartwatch on a weekly basis (Table 2). Participants who described their working situation as being employed, had 5.7 times higher odds of using a smart watch compared with participants who identified their working situation differently (OR 5.7, 95%CI 1.8; 17.7, p = 0.003), while correcting for sex.

Seven characteristics were potentially associated with weekly use of a fitness watch (Supplementary Table 5). In order of lowest to highest p-value, these were: working status; age; impaired SIN recognition (yes/no); alcohol consumption; hearing aid use, SRTn and living status. As impaired SIN recognition and SRTn are overlapping constructs, these were tested in two different models. The final full association model is shown in Table 2. Four characteristics remained, in order of the strength of their association: working status (p < 0.001), alcohol consumption (p = 0.006), hearing aid use (p = 0.024) and living situation (p = 0.033). Participants who mainly identified as employed had 3.4 times higher odds of using a fitness watch compared with participants who identified differently (OR 3.4, 95%CI 1.8; 6.6), while correcting for alcohol consumption, hearing aid use and living situation. Participants who consume 1–2 glasses alcoholic beverages/week (2nd quartile; Q2) had the highest odds of using a fitness watch so this category was chosen as the reference category. Participants who consume no alcohol have 3.3 times lower odds of using a fitness watch as compared to those who consume 1–2 glasses per week, keeping all other characteristics the same (OR 0.3, 95%CI 0.1; 0.6). The other two levels of alcohol consumption each have 2.7 times lower odds of using a fitness watch (OR 0.4, 95%CI 0.2; 0.8) (Table 2). Hearing aid users had 2.0 times lower odds of using a fitness watch (OR 0.5, 95%CI 0.2; 0.9), while controlling for the other characteristics, whereas those living together with others had 2.5 higher odds (OR 2.5, 95%CI 1.1; 5.8).

Four characteristics showed a potential association with the use of physical activity apps. In order of their p-value these were: working status, age, migraine and osteoarthritis (Supplementary Table 6). After forward selection only working status remained in the model, with an almost four times higher odds for using physical activity apps when identifying as employed (OR 3.9, 95%CI 2.1; 7.1, p < 0.001).

Discussion

The weekly use of a fitness watch, smartwatch or physical activity app did not differ between adults with and without impaired SIN. The demographic and health-related characteristics associated with using any type of fitness technology were the same for both groups. Working status was associated with the use of all fitness technology, with employed adults having higher odds for using them than those identifying differently. In addition, sex was associated with smartwatch use, with women having lower odds for using it than men. Besides working status, fitness watch use was associated with alcohol consumption, hearing aid use and living status. Adults who consume no alcohol or more than two glasses per week had lower odds of using a fitness watch, as compared to those who consume 1–2 glasses per week. Hearing aid users had lower odds for using a fitness watch, while people who live with others as compared to living alone had higher odds of using this device.

RQ1: weekly use of fitness technology

Smartwatch use and use of physical activity apps did not differ between the two SIN groups. The number of smartwatch users was quite low, which leads to unreliable results. The analysis of use of physical activity apps is less affected by low numbers, but differences in features used, such as tracking of physical activity or educational content, are still possible. Our current results on use of these apps, with correction for age and working status did confirm previously reported results, in which we only corrected for age (van Wier, Urry et al. 2021). Fitness watch use was lower among adults with impaired recognition of speech-in-noise in the unadjusted analyses, but this disappeared after the confounders working status, age and hearing aid use were added to the model. Our results suggest that there are no differences in the use of fitness technology between adults with and without impaired SIN recognition, taking confounding factors into account. Given that our data were collected when wearable fitness technology was still new and the overall uptake in the Netherlands was low, this result could change over time and might not apply to other regions.

RQ2: user characteristics associated with weekly use of fitness technology

Impaired SIN recognition did not modify any of the associations that could be tested. This implies that, in our sample, the same user characteristics are associated with use of a fitness watch and physical activity apps among adults with and without this impairment. In the univariable analyses, worse SIN recognition as well as use of a hearing aid were significantly associated (p < 0.05) with lower odds for using a fitness watch, but not for using a smartwatch or physical activity apps. In the full association model worse SIN recognition disappeared, likely because other variables in the model which are strongly associated with SIN recognition, namely working status and use of a hearing aid, absorbed its association and added more to the explained variance. To illustrate, 84% of the non-working adults were categorised to Q3 and Q4 of the SRTn, having a SRTn > −5.2 dB SNR. In the Netherlands, a hearing aid is reimbursed when mean PTA_1,2,4kHz loss ≥ 35 dB in at least one ear, and its use is indicative of moderate to profound hearing loss. Various studies found that less physically active adults are less likely to use a fitness tracker (Dool et al. 2017; Shen et al. 2017; Macridis et al. 2018). Considering that adults with hearing loss are less physically active than normal hearing adults, lower use of a fitness watch by hearing aid users is perhaps explained by an association with lower physical activity in this group. An additional reason could be the age distribution among hearing aid users: 56% of them was aged ≥ 60 years or older vs. 46% in non-users.

A number of previous studies have investigated user characteristics associated with fitness technology in a comparable way as we have, in the general population, using models with mutual adjustment for variables. The type and goals of the health technology studied varied, i.e. all health-related apps (Shen, Wang et al. 2017); apps and wearable trackers for physical activity (Macridis, Johnston et al. 2018; Strain et al. 2019); apps and wearable devices tracking physical activity for weight management (Strain, Wijndaele et al. 2019); or apps and devices tracking all health-related behaviours and outcomes (Paré et al. 2018; Chandrasekaran et al. 2020; Rising et al. 2020).

All studies found an association with age, with older respondents less likely to use health technology. In the univariable analyses of fitness watch use and physical activity app use we also found an association with age, which disappeared in the multivariable analyses, perhaps because working status was added. Working status is highly correlated with age, with in our study 65% of 60–66 year old’s (Q3) and 98% of those of 67 years and older (Q4) belonging to the non-working group.

Working status was associated with use of all types of fitness technology. One other study that included working status (Paré, Leaver et al. 2018) also found that working participants were more likely to digitally track one or more aspect of their health or wellbeing compared to retirees. Besides age, other underlying reasons for the association between working status and use of fitness technology could be a higher income among employed adults, which in three studies was found to increase health technology use (Shen, Wang et al. 2017; Paré, Leaver et al. 2018; Chandrasekaran, Katthula et al. 2020). Another reason could be higher digital competency among working adults (Ertl et al. 2020). Other studies have reported that lower digital competency (eHealth literacy) reduces the use of digital interventions (Western et al. 2021).

We did not find a relation with education, whereas almost all studies found that people with a higher level of education were more likely to track aspects of their health (Shen, Wang et al. 2017; Macridis, Johnston et al. 2018; Paré, Leaver et al. 2018; Chandrasekaran, Katthula et al. 2020). Not finding an association with education may have been caused by the low number of low-educated adults participating in our study, namely 13%.

Our results showed that sex was only related to using a smartwatch, but not to other fitness watch technology. It should be kept in mind that the low number of smartwatch users makes this result less reliable. Two other studies also found no relationship between sex and tracking of physical activity and other health behaviours (Paré, Leaver et al. 2018; Strain, Wijndaele et al. 2019). In comparison, three studies found that women were more likely than men to track their physical activity (Macridis, Johnston et al. 2018) or multiple health/wellness outcomes (Chandrasekaran, Katthula et al. 2020; Rising, Jensen et al. 2020), one study found that men were more likely to track physical activity but not more likely to track other health measures (Shen, Wang et al. 2017) and one study found women more likely to track activity for weight management purposes (Strain, Wijndaele et al. 2019). Similarly, two other studies found that people with a higher BMI were more likely to track their physical activity (Macridis, Johnston et al. 2018) or health-related behaviours (Rising, Jensen et al. 2020). The latter results suggest that using certain health technology, irrespective of sex, depends on how the particular technology aligns with personal goals and needs. This corroborates with results from qualitative studies reporting that wanting to achieve specific goals was important for the adoption and sustained use of health and fitness wearables (Canhoto and Arp 2017; Moore et al. 2021).

Our finding that living with others is associated with higher odds for using a fitness watch concurs with results from one other study which found that people who are married more likely to use a fitness tracker (Macridis, Johnston et al. 2018). Two other studies found no association (Shen, Wang et al. 2017; Chandrasekaran, Katthula et al. 2020). It is therefore uncertain if living status is associated with tracking of physical activity.

We found no influence of having a chronic condition, in contrast to two other studies that found a negative association (Paré, Leaver et al. 2018; Rising, Jensen et al. 2020). Both studies analysed the presence of several chronic conditions as one characteristic, whereas we analysed the chronic conditions separately. This might explain the lack of statistical significance in our study, although a study that also treated presence of chronic diseases as a single variable did not find an association either (Chandrasekaran, Katthula et al. 2020). It is important to mention that our study was not powered to study associations for all chronic diseases separately. When statistical significance is ignored, many of the chronic conditions in our study had adverse associations, but a few were in the other direction. This could indicate that the specific chronic condition matters to whether or not fitness technology is used, perhaps because some chronic diseases limit physical activity while others call for exercise to alleviate complaints or prevent progression. This may affect use of fitness technology in opposite ways, but further research is needed to clarify this. This is relevant to adults with hearing impairment as they often have other chronic diseases (Besser, Stropahl et al. 2018).

The association of alcohol consumption with use of a fitness watch is an interesting finding, although the confidence intervals are fairly wide. Adults who consume 1–2 glasses of alcohol/week might represent a group who in general have healthier behaviour. For instance, this group had the lowest percentage of smokers (7% vs. 12%, 10% and 13% in the other groups). Alcohol consumption has in many studies been linked to higher physical activity levels, with more studies reporting a positive relationship of light to moderate alcohol use than studies reporting this for heavy alcohol use (Dodge et al. 2017). Moderate drinkers in our study could therefore also be more physically active and this in turn could be associated to their higher odds for using a fitness watch, as mentioned before.

Implications for practice and research

According to the Diffusion of Innovations theory, individuals who pick up an innovation early after its introduction differ from those who adopt it later (Berwick 2003). Considering that less than 10% of our sample was using each of fitness technologies, results are more representative of the early adopters of physical activity apps and wrist-worn wearables than of subsequent adopters (Canhoto and Arp 2017). We argue that user characteristics of these early adopters of wrist-worn wearables are comparable to those of early adopters of ear-worn fitness technology. Our results suggest that working adults might the first ones to be interested in hearing aids that track physical activity. Research is warranted if working status and currently not using a hearing aid indeed reflect underlying factors such as lower age, higher income, higher digital competency and higher levels of physical activity. Moreover, research on whether they predict use of ear-worn activity trackers is needed.

If age is indeed an underlying factor in the association of working status with any fitness technology use, its role in retired individuals, who comprise the majority of people with hearing loss, cannot completely be dismissed. It is still possible that older adults in this group are less interested in using fitness technology than younger retired adults, although another study reports that this depends on the usability of the device and the context in which it is offered (Moore, O'Shea et al. 2021). They conclude that older adults are interested in wearables but require devices adjusted to their abilities, as well as a support structure to encourage their use. This suggests that older people need differently designed fitness technology than the leading commercial companies supply. The hearing aid industry, audiologists and hearing aid dispensers have extensive experience with older clients, and could be well positioned to offer more suitable devices and support.

Our finding that hearing aid use was associated with lower use of a fitness watch does pose a challenge if this applies to ear-worn fitness trackers as well. Adults looking for a hearing aid may need encouragement to obtain one that tracks their activity. This has to start with a conversation about the interest the client has in becoming more physically active, and if and how they want to use physical activity tracking to achieve this (Otmanowski and Chase 2020). Hearing care professionals that offer these hearing aids need to think about business models that include such support, or seek collaboration with other parties that share an interest in improving physical activity in adults with impaired hearing. The hearing aid industry should build their proprietary apps in such a way that the app coaches clients through the process of behaviour change, based on the theory and evidence in this field (Mollee, Middelweerd et al. 2017; Domin et al. 2021). Specific to adults aged ≥50 years, the most significant improvements in behavioural and health outcomes are seen when the digital solution comprises goal setting, self-monitoring/tracking, feedback and social support (Stockwell et al. 2019).

It should also be noted that fitness technology may not increase physical activity in all adults. For instance, no effectiveness was shown in individuals of low socioeconomic status (Western, Armstrong et al. 2021). Moreover, fitness wearables and physical activity apps are not to everybody’s liking, nor is it necessary to use them to become more physically active. To address physical inactivity in adults with impaired hearing a more detailed picture is needed. Cross-sectional and longitudinal national cohort studies can be used to identify specific subgroups among adults with impaired hearing who are insufficiently active, preferably through measuring physical activity with accelerometers. A theory-based tool such as the Behaviour Change Wheel can then be applied to identify effective physical activity interventions which are already available, or to develop new interventions (Michie et al. 2011). Among these interventions (ear-worn and wrist-worn) fitness trackers and mobile applications may well have a place.

Strengths and limitations

Strengths of the study include the large sample, its diversity in age and the inclusion of adults with normal and impaired SIN recognition. Finally, we are the first study to evaluate whether weekly use of fitness technology among adults with impaired hearing ability differs from adults with normal hearing ability.

Our study has several limitations. All data, except SIN recognition, were self-reported and could therefore suffer from bias. As we did not ask which smartwatch functionalities participants used, we cannot be certain they used this device to track their physical activity. Furthermore, it would have been useful to analyse physical activity apps according to features used, but this was not included in the questionnaire. Another limitation is the relatively low number of participants using fitness technology. This particularly hampers the analyses with smartwatch use, of which the results should be interpreted with great caution. It should also be taken into account that understanding of speech-in-noise is less compromised in conductive and mixed hearing losses. This means that participants with these types of hearing impairment, even those using a hearing aid, may be classified as having normal SIN recognition. Use of the diotic testing protocol results in underestimation of unilateral or asymmetric (sensorineural) hearing loss (De Sousa, Swanepoel et al. 2020). Participants using speakers may perform slightly worse in the test than if they had used headphones, resulting in misclassification to the group with impaired SIN recognition for some of them. Furthermore, random measurement error in the speech-in-noise test might also result in misclassification of a participant. These non-differential misclassifications lead to a bias towards the null, i.e. to a reduction in the strength of an estimated association. Finally, all analyses were cross-sectional which does not allow for inferences on causal relationships.

Conclusions

In adults with impaired SIN recognition the same characteristics are associated with using fitness technology as in adults without this impediment. They also use this technology to the same extent, although hearing aid users make less use of a fitness watch. Employees or business owners are more likely to use all types of fitness technology. Women may be less likely to use a smartwatch than men. Furthermore, adults who consume no alcohol or more than two glasses per week are less likely to use a fitness watch than adults who consume 1–2 glasses per week. Among adults living with others, higher use of a fitness watch is noted. Research is warranted if these characteristics reflect underlying factors such as lower age, higher income, digital competency and better health. Future research should aim to differentiate physical activity patterns among adults with hearing impairment, to recognise which specific groups are insufficiently active. Theory-based tools can then be used to identify and develop interventions, among which digital solutions, to address physical inactivity in these groups.

Acknowledgements

The authors thank the participants of the Netherlands Longitudinal Study on Hearing.

Disclosure statement

The authors have no conflict of interest to disclose.

Additional information

Funding

Funding for data collection of the NLSH 10-year measurement round came from the EMGO + Institute for Health and Care Research, The Netherlands and Sonova AG, Switzerland. EU is affiliated with Sonova AG. The current study was financially supported by Sonova AG, Switzerland.