Reliable and valid measurement of free-living physical activity (PA) is pivotal in the examination of population PA levels and dose–response effects of PA on health (Migueles et al., 2017). Triaxial accelerometers capture high-resolution raw data, which can be processed into, for example, sedentary time and duration, intensity, and frequency of PA (Migueles et al., 2017). Such data are richer than data from, for example, uniaxial pedometers providing only step counts or PA questionnaires giving very uncertain outcome estimates (Dowd et al., 2018). There are several methods available for attaching accelerometers onto the subject, and taping them directly onto the skin has been used in several studies (Duncan et al., 2018; Schneller et al., 2017). Taping devices onto the skin is challenging; several elements of the taping protocol are crucial for the integrity of the attachment. Identifying the optimal solution for attaching accelerometers onto the skin is needed to ensure high wear compliance and to ensure that the subject is not harmed doing the PA recording.
Accelerometers are biomechanical movement sensors that record bodily movement by changes in acceleration in up to three axes (Migueles et al., 2017). In PA research, accelerometer technology emerged in the 1980s and 1990s (Troiano et al., 2014). The most common procedure has been to apply a single-accelerometer system using a belt to strap the device onto the hip (Migueles et al., 2017). Using this setup, raw data can be converted into duration, intensity, and frequency of PA as well as sedentary time, whereas body posture and specific PA behaviors cannot be derived (Brønd et al., 2020). A critical aspect of accelerometer measurement is to capture as complete a recording as possible, and therefore it is a drawback of belt-worn accelerometers that participants should take them off during sleep and water activity (because they are not waterproof) and subsequently have to remember to put them on again. Wear compliance is thus known to be a challenge when using belt-mounted accelerometers (Troiano et al., 2014). Another attachment option is skin mounting, which has received increased attention over the past few years as it improves wear compliance, and thus data quality (Schneller et al., 2017; Troiano et al., 2014). Skin mounting has been applied using different types of accelerometers, for example, activPAL (Edwardson et al., 2017), ActiGraph (Gupta et al., 2015; Johansson et al., 2022), and Axivity AX3 (Duncan et al., 2018; Schneller et al., 2017). The triaxial Axivity AX3) is very suitable for skin mounting because it is a small, light-weight, waterproof device, which can be attached directly onto the skin for uninterrupted recording 24 hr/day (Duncan et al., 2018). When used as a dual-accelerometer system, that is, as a combination of two devices capturing triaxial acceleration of the truncus and of a lower limb respectively, the recordings can be processed into sedentary time and intensities, duration, and frequency of PA, as well as detailed information on body posture and specific PA behaviors (Duncan et al., 2018). Axivity AX3 accelerometers can be worn in removable solutions such as belts (Brønd et al., 2020) or wrist bands (Lopes et al., 2021) but, as mentioned, they can also be taped directly onto the skin (Brønd et al., 2020; Duncan et al., 2018; Schneller et al., 2017). Taping devices directly onto the skin may have potential negative side effects such as skin reactions (Duncan et al., 2018). Schneller et al. (2017) examined accelerometer-derived wear time in 903 children using skin taping of a dual-accelerometer Axivity AX 3 system and reported that 65.7% of the children wore the thigh-worn accelerometer and 59.5% wore the lower back accelerometer uninterrupted for 24 hr/day for seven consecutive days. In a small sample of 16 adults who also combined a thigh-worn with a lower back-worn Axivity AX3 unit for seven consecutive days using medical tape fixation, Duncan et al. (2018) examined self-reported adhesion and found that none of the participants removed the units prematurely. The authors also collected self-reported data on skin irritation, which was found in three out of the 16 adult individuals (18.8%). This study also included children, for whom neither adhesion nor skin irritation were investigated (Duncan et al., 2018). Nor were such data included in the study by Schneller et al. (2017). In both of these studies, attachment of the accelerometers onto the skin of the participants was undertaken by researchers (Duncan et al., 2018; Schneller et al., 2017), illustrating that this is a human resource-consuming procedure. No other information was given about ease of handling of the accelerometers.
On this basis, the aim of the present study was to evaluate different attachment protocols with respect to: (a) skin reactions, (b) adhesion problems, and (c) wear time in children and adults. The second aim was to study the ease of accelerometer handling associated with different attachment protocols.
Methods
Study Design, Setting, and Sample
The present study is cross-sectional using data from the Lolland-Falster Health Study (LOFUS). LOFUS is a household-based population study that recruited randomly selected households on the two Danish islands Lolland and Falster with a combined total of 103,000 inhabitants from February 8, 2016 to February 13, 2020. The design, procedures, and entire data collection of LOFUS are described in detail elsewhere (Egholm et al., 2020; Jepsen et al., 2020). In total, 18,949 individuals aged 0–96 years participated in LOFUS. Data collection on PA using accelerometers was implemented for subsamples of LOFUS participants starting February 1, 2017. From February 1, 2017 to November 30, 2018, individuals belonging to families with at least one child aged ≤17 years were eligible for inclusion. From December 1, 2018 through February 13, 2020, all LOFUS participants were eligible for inclusion. However, individuals who were not able to walk, such as very young toddlers or wheelchair users, were excluded from accelerometer measurement (Petersen et al., 2020).
Accelerometers
The LOFUS used a dual system of Axivity AX3 accelerometers to measure free-living PA. The Axivity AX3 accelerometer is a triaxial accelerometer containing a 512Mb NAND nonvolatile memory and a battery that provides logging for 14 days at 100 Hz. It combines a microelectromechanical system for sampling of raw acceleration data with a real-time clock, a temperature sensor, and a light sensor, for example, for categorization of wear time versus nonwear time. The device is a small case (23 × 32.5 × 7.6 mm) weighing 11 g. The external material is made of polycarbonate, and the device is dust tight and water resistant to 1.5 m. A Micro USB serves as connector. More details can be found in the Axivity AX3 data sheet (2015). The edges of the devices were polished before use in order to make them less sharp. LOFUS staff prepared kits for attachment of accelerometers, equipped participants with the two devices, and instructed the participants to wear them for seven consecutive, uninterrupted days, including during sleep and water activity. Long hair on the attachment site was removed with a single-use shaver with respect to skin integrity, and greasy skin was cleaned gently with an alcohol wipe. One accelerometer was attached onto the skin that covered the right musculus latissimus dorsi at the lower part of the lumbar curve at a distance of 1–5 cm (depending on body size) horizontally from the spine (Figure 1). The x-axis of the accelerometer turned horizontal, and the y-axis pointed distal (USB port distal and label side of the accelerometer invisible). The other accelerometer was attached to the skin that covered the right musculus quadriceps femoris in the medial position and centrally between the hip and the knee (Figure 1). The x-axis turned horizontal, and the y-axis pointed distal toward the patella (USB port distal and label side of the accelerometer visible).
—Attachment sites on the body for the dual-Axivity AX3-accelerometer system. Note. (a) One accelerometer was attached onto the skin that covered the right musculus latissimus dorsi at the lower part of the lumbar curve at a distance of 1–5 cm (depending on body size), horizontally from the spine. (b) The other accelerometer was attached to the skin that covered the right musculus quadriceps femoris in the medial position and centrally between the hip and the knee.
Citation: Journal for the Measurement of Physical Behaviour 5, 4; 10.1123/jmpb.2022-0024
—Attachment sites on the body for the dual-Axivity AX3-accelerometer system. Note. (a) One accelerometer was attached onto the skin that covered the right musculus latissimus dorsi at the lower part of the lumbar curve at a distance of 1–5 cm (depending on body size), horizontally from the spine. (b) The other accelerometer was attached to the skin that covered the right musculus quadriceps femoris in the medial position and centrally between the hip and the knee.
Citation: Journal for the Measurement of Physical Behaviour 5, 4; 10.1123/jmpb.2022-0024
—Attachment sites on the body for the dual-Axivity AX3-accelerometer system. Note. (a) One accelerometer was attached onto the skin that covered the right musculus latissimus dorsi at the lower part of the lumbar curve at a distance of 1–5 cm (depending on body size), horizontally from the spine. (b) The other accelerometer was attached to the skin that covered the right musculus quadriceps femoris in the medial position and centrally between the hip and the knee.
Citation: Journal for the Measurement of Physical Behaviour 5, 4; 10.1123/jmpb.2022-0024
Before the start of the accelerometer data collection, we did a search of the literature for specific, evidence-based attachment protocols for tape mounting of Axivity AX3 accelerometers. The search was unsuccessful, and therefore, our approach to attachment protocols became open to change throughout the data collection period. The optimal solution should be skin-friendly, prevent accelerometers from falling off, and be staff-friendly. We ended up having used four different attachment protocols, which are described and justified in detail in Table 1. All attachment protocols included use of only medical tape except for Protocol 1, which also included nonmedical tape between the device and the skin, however, not directly onto the skin. Protocol 1 (see Table 1 and Figure 2) and Protocol 2 (see Table 1 and Figure 2) were based on experience shared by other researchers. Both included fixation of the device onto the skin using stretch tape and complete coverage of the device and the surrounding skin using medical tape. The attachment protocol used in Protocol 1 was later published by Schneller et al. (2017). In Protocol 3 (see Table 1 and Figure 2) and Protocol 4 (see Table 1 and Figure 2), the accelerometer itself was not taped onto the skin. The differences between these two protocols were the length of the two narrow pieces of adhesive film that were attached horizontally and vertically to fixate the accelerometer onto the skin (longest in Protocol 4), and that the device was completely covered on all sides by the cotton gauze pad in Protocol 4 (see Table 1 and Figure 3).
Attachment Protocols for a Dual-Accelerometer System in the Lolland-Falster Health Study, 2017–2020
| Attachment protocol | Period | Procedure | Input from | Justification | Observations |
|---|---|---|---|---|---|
| 1 | February 1, 2017–October 1, 2017 | (a) A 30 × 40-mm piece of Fixomull stretch tapea was adhered to the skin. (b) On top of the stretch tape, a 3M double-sided adhesive tapeb was adhered. (c) On top of the double-sided adhesive tape, the accelerometer was attached. (d) The accelerometer and the surrounding skin were covered with an 80 × 100-mm piece of Opsite Flexifix transparent adhesive film.c | Oral information from University of Southern Denmark—later published in Schneller et al. (2017) | A starting point, which had been used by others. There were no available data on the feasibility for participants and staff. | Regarding participants: We suspected skin reactions. Regarding staff: Cleaning the accelerometers for glue was a time-consuming and strenuous task for hands, arms, and shoulders. |
| 2 | October 2, 2017–June 3, 2018 | (a) A 30 × 40-mm piece of Opsite Flexifix transparent adhesive filmc was adhered to the skin. (b) On top of the transparent film, the accelerometer was placed. (c) The accelerometer and the surrounding skin were covered with an 80 × 100-mm piece of Opsite Flexifix transparent adhesive film.c | Oral information from Trøndelag Health Study, Norway (www.hunt.no) | Learning from the experiences of others in an attempt to prevent skin reactions. There were no available data on the feasibility for participants and staff. | Regarding participants: We suspected skin reactions. Regarding staff: Cleaning the accelerometers for glue was a time-consuming and strenuous task for hands, arms, and shoulders. |
| 3 | June 4, 2018–November 30, 2018 | (a) A 50 × 50-mm Mesoft unsterile four-layer cotton gauze padd was folded once and placed on the skin. (b) The accelerometer was placed on top of the gauze pad. (c) Two 20 × 100-mm pieces of Opsite Flexifix transparent adhesive filmc were attached vertically and horizontally, respectively, onto the accelerometer and the skin. | Advice from a dermatologist | We started using cotton gauze pads to allow humidity to vaporize from the skin and thus prevent skin reactions. | Regarding participants: Skin reactions seemed to be reduced. Regarding participants: We suspected that accelerometers fell off more frequently than with attachment Protocols 1 and 2. Regarding staff: Cleaning the accelerometers for glue was a time-consuming and strenuous task for hands, arms, and shoulders. |
| 4 | December 1, 2018–February 13, 2020 | (a) A 50 × 50-mm Mesoft unsterile four-layer cotton gauze padd was cut to fit the vertical length of the device and folded around the accelerometer with a double layer on the skin side. (b) The gauze pad covered accelerometer was placed on the skin. (c) Two 28 × 150-mm pieces of Opsite Flexifix transparent adhesive filmc were attached vertically and horizontally, respectively, onto the accelerometer and the skin. | Synthesis of our experience from attachment Protocols 1–3 | In order to save staff time and decrease the risk of occupational injuries, we wished to avoid cleaning of accelerometers for glue. | We considered this the best method for both participants and staff. |
aBSN Medical. b3M. cSmith and Nephew. dMölnlycke Health Care.
—Illustration of the taping of Axivity AX3 accelerometers in Protocol 1, 2, and 3. Note. (a) Protocol 1: A piece of Fixomull stretch tape was adhered to the skin. The accelerometer was attached on top of the stretch tape using 3M double-sided adhesive tape. The accelerometer and the surrounding skin were covered with Opsite Flexifix transparent adhesive film. (b) Protocol 2: A piece of Opsite Flexifix transparent adhesive film was adhered to the skin. The accelerometer was placed on top of the film and covered with a piece of Opsite Flexifix transparent adhesive film. (c) Protocol 3: A Mesoft cotton gauze pad was folded and placed on the skin. The accelerometer was placed on top of the pad and fixated first vertically and then horizontally with two pieces of Opsite Flexifix transparent adhesive film (Table 1).
Citation: Journal for the Measurement of Physical Behaviour 5, 4; 10.1123/jmpb.2022-0024
—Illustration of the taping of Axivity AX3 accelerometers in Protocol 1, 2, and 3. Note. (a) Protocol 1: A piece of Fixomull stretch tape was adhered to the skin. The accelerometer was attached on top of the stretch tape using 3M double-sided adhesive tape. The accelerometer and the surrounding skin were covered with Opsite Flexifix transparent adhesive film. (b) Protocol 2: A piece of Opsite Flexifix transparent adhesive film was adhered to the skin. The accelerometer was placed on top of the film and covered with a piece of Opsite Flexifix transparent adhesive film. (c) Protocol 3: A Mesoft cotton gauze pad was folded and placed on the skin. The accelerometer was placed on top of the pad and fixated first vertically and then horizontally with two pieces of Opsite Flexifix transparent adhesive film (Table 1).
Citation: Journal for the Measurement of Physical Behaviour 5, 4; 10.1123/jmpb.2022-0024
—Illustration of the taping of Axivity AX3 accelerometers in Protocol 1, 2, and 3. Note. (a) Protocol 1: A piece of Fixomull stretch tape was adhered to the skin. The accelerometer was attached on top of the stretch tape using 3M double-sided adhesive tape. The accelerometer and the surrounding skin were covered with Opsite Flexifix transparent adhesive film. (b) Protocol 2: A piece of Opsite Flexifix transparent adhesive film was adhered to the skin. The accelerometer was placed on top of the film and covered with a piece of Opsite Flexifix transparent adhesive film. (c) Protocol 3: A Mesoft cotton gauze pad was folded and placed on the skin. The accelerometer was placed on top of the pad and fixated first vertically and then horizontally with two pieces of Opsite Flexifix transparent adhesive film (Table 1).
Citation: Journal for the Measurement of Physical Behaviour 5, 4; 10.1123/jmpb.2022-0024
—Illustration of the recommended attachment protocol for skin mounting of Axivity AX3 accelerometers. Note. (a) A 50 x 50-mm Mesoft unsterile four-layer cotton gauze pad was cut to fit the vertical lenght of the device and folded around the accelerometer with a double layer on the skin side. The gauze pad covered accelerometer was placed on the skin. (b and c) Two 28 × 150-mm pieces of Opsite Flexifix transparent adhesive film were attached first vertically and then horizontally on top of the accelerometer and onto the skin.
Citation: Journal for the Measurement of Physical Behaviour 5, 4; 10.1123/jmpb.2022-0024
—Illustration of the recommended attachment protocol for skin mounting of Axivity AX3 accelerometers. Note. (a) A 50 x 50-mm Mesoft unsterile four-layer cotton gauze pad was cut to fit the vertical lenght of the device and folded around the accelerometer with a double layer on the skin side. The gauze pad covered accelerometer was placed on the skin. (b and c) Two 28 × 150-mm pieces of Opsite Flexifix transparent adhesive film were attached first vertically and then horizontally on top of the accelerometer and onto the skin.
Citation: Journal for the Measurement of Physical Behaviour 5, 4; 10.1123/jmpb.2022-0024
—Illustration of the recommended attachment protocol for skin mounting of Axivity AX3 accelerometers. Note. (a) A 50 x 50-mm Mesoft unsterile four-layer cotton gauze pad was cut to fit the vertical lenght of the device and folded around the accelerometer with a double layer on the skin side. The gauze pad covered accelerometer was placed on the skin. (b and c) Two 28 × 150-mm pieces of Opsite Flexifix transparent adhesive film were attached first vertically and then horizontally on top of the accelerometer and onto the skin.
Citation: Journal for the Measurement of Physical Behaviour 5, 4; 10.1123/jmpb.2022-0024
For time-saving purposes, we developed a “cutting machine” after inspiration from the Norwegian Trøndelag Health Study (www.HUNT.no) for simultaneous cutting of four rolls of the adhesive film used in all four attachment protocols (see Supplementary Figure S1 [available online]).
Predominantly, skin attachment of accelerometers was undertaken by LOFUS staff (registered nurses and biomedical laboratory technicians), but in a few cases, staff handed out an attachment kit including a written instruction with photos of each step of the attachment protocol for the participant to self-administer the attachment. All participants received at least one extra kit of equipment including the abovementioned written instruction for reattachment in case an accelerometer fell off before the completed wear time. LOFUS staff underwent standard operation procedure training in attachment of accelerometers and instruction of the participants; adherence to the procedure was monitored regularly. Participants were either handed a stamped envelope for return of the devices or agreed to personally hand in the accelerometers to LOFUS.
Outcomes
Participants were asked to hand in a completed questionnaire together with the accelerometers, which included one question asking “Have you been wearing the activity devices for seven consecutive days?” If the participant answered no, an open-ended question inquired “Why did you take them off?” Parents reported on behalf of their children. This handwritten information was categorized as yes or no to skin reaction (e.g., reports of itch, burn, sting, redness, swelling, warmth, or unspecific irritation) and yes and no to adhesion problem (accelerometer fell off). Ease of handling was assessed throughout the data collection period and is presented in Table 1. Wear time was derived from the accelerometers considering both acceleration and temperature and computed as number of days (Petersen et al., 2020).
Statistical Analyses
We used Stata (version 16.1, StataCorp) for the statistical analyses. Descriptive statistics characterized the sample by sex and age. Participants were classified into four groups: Protocol 1, Protocol 2, Protocol 3, and Protocol 4. A one-way analysis of variance was conducted separately for children and adults to determine if the prevalence of skin reactions differed between the four attachment protocols. A one-way analysis of variance was also conducted for children and adults to determine if the prevalence of accelerometers that fell off during the wear period differed between the four groups. To compare group means, we performed a Tukey post hoc test. A p value of ≤.05 was considered statistically significant. A two-sided t test was conducted to examine potential differences between children and adults in relation to skin reactions, adhesion problems, and wear time.
Ethics
Region Zealand’s Ethical Committee on Health Research (SJ-421) and the Danish Data Protection Agency (REG-024-2015) approved LOFUS, including accelerometer measurement of PA. Participants aged ≥15 years gave written consent for participation in LOFUS, while parents or other holders of custody gave consent for children aged 0–14 years (Jepsen et al., 2020). LOFUS is registered in Clinicaltrials.gov (NCT02482896).
Results
In total, 5,675 LOFUS participants consented to accelerometer measurement. Of these, 57 returned neither the accelerometers nor the questionnaire, while 230 returned the accelerometers but not the questionnaire, leaving 5,389 individuals (see Table 2) for inclusion in the analysis (Protocol 1, n = 438 [February 1, 2017–October 1, 2017]; Protocol 2, n = 630 [October 2, 2017–June 1, 2018]; Protocol 3, n = 636 [June 2, 2018–November 30, 2018]; and Protocol 4, n = 3,685 [December 1, 2018–February 13, 2020]). The youngest child who could walk and wear accelerometers was 10 months old, and the oldest participant was 95 years of age (1,289 were aged 10 months–17 years [53% girls], and 4,100 were 18–95 years [55% female]). Descriptive statistics on skin reactions, adhesion problems, and wear time are shown in Table 2.
Descriptive Statistics of the Participants Included in the Four Attachment Protocols for Tape-Mounted Axivity AX3 Accelerometers in the Lolland-Falster Health Study
| Attachment protocol | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| 1, n = 438 | 2, n = 630 | 3, n = 636 | 4, n = 3,685 | All, n = 5,389 | ||||||
| Children | Adults | Children | Adults | Children | Adults | Children | Adults | Children | Adults | |
| N | 218 | 220 | 314 | 316 | 314 | 322 | 443 | 3,242 | 1,289 | 4,100 |
| Age (years) | ||||||||||
| Mean (SD) | 9.9 (4.7) | 40.1 (8.3) | 10.3 (4.1) | 41.6 (7.9) | 11.1 (3.9) | 41.9 (9.0) | 11.0 (3.8) | 57.7 (15.7) | 10.7 (4.1) | 54.3 (15.9) |
| Range | 1.1–17.9 | 18.1–61.7 | 0.97–17.9 | 18.1–59.2 | 1.1–17.9 | 18.1–65.4 | 1.8–17.9 | 18.0–95.4 | 0.97–17.9 | 18.0–95.4 |
| Skin reactions, n (%) | 63 (28.9) | 42 (19.1) | 84 (26.8) | 50 (15.8) | 31 (9.9) | 19 (6.0) | 71 (16.0) | 227 (6.0) | 231 (19.0) | 338 (8.2) |
| Adhesion problems, n (%) | 24 (11.0) | 10 (4.5) | 19 (6.1) | 7 (2.2) | 51 (16.2) | 19 (5.9) | 21 (4.74) | 51 (1.6) | 108 (8.9) | 87 (2.1) |
| Wear time (days), mean (SD) | 4.8 (2.7) | 5.4 (2.5) | 5.0 (2.6) | 5.4 (2.5) | 5.4 (2.5) | 5.9 (2.3) | 5.7 (2.6) | 6.5 (2.2) | 5.3 (2.6) | 6.3 (2.3) |
| Completed wear period,a % | 69 | 77 | 71 | 77 | 77 | 84 | 81 | 93 | 76 | 90 |
| Sex | ||||||||||
| Female, n (%) | 122 (56.0) | 124 (58.5) | 151 (48.1) | 177 (58.4) | 164 (52.2) | 180 (60.0) | 247 (55.8) | 1,704 (54.2) | 684 (53.1) | 2,262 (55.2) |
| Male, n (%) | 96 (44.0) | 88 (41.5) | 163 (51.9) | 126 (41.6) | 150 (47.8) | 120 (40.0) | 196 (44.2) | 1,441 (45.8) | 605 (46.9) | 1,838 (44.8) |
aA full wear period was seven consecutive days.
Among children, we found a statistically significant difference in skin reactions between the four groups, F(3, 1289) = 15.50, p < .001. Skin reactions were reported significantly more in Protocols 1 and 2 compared to Protocols 3 and 4 (p < .001). There was no difference between Protocols 1 and 2 (p = .923, 95% confidence interval [CI] [−0.110, 0.067]) or between Protocols 3 and 4 (p = .139, 95% CI [−0.012, 0.135]) respectively. Adhesion problems varied between the four groups, F(3, 1289) = 11.81, p < .001, and the Tukey post hoc analysis revealed statistically significant differences between Protocols 1 and 4 (p < .036), Protocols 2 and 3 (p < .001), and Protocols 3 and 4 (p < .001). Adhesion problems were most prevalent in Protocol 3 and were reported least in Protocols 2 and 4 (p < .036).
In adults, we found a statistically significant difference in skin reactions between the four groups, F(3, 4100) = 22.74, p < .001. Skin reactions were reported significantly more in Protocols 1 and 2 compared to Protocols 3 and 4 (p < .001). There was no difference between Protocols 1 and 2 (p = .522, 95% CI [−0.094, 0.029]) or between Protocols 3 and 4 (p = .901, 95% CI [−0.030, 0.052]) respectively. Adhesion problems varied between the four groups, F(3, 4100) = 11.10, p < .001, and the Tukey post hoc analysis revealed statistically significant differences between Protocols 1 and 4 (p < .016), and Protocols 2 and 3 (p < .007). Most problems with adhesion occurred in Protocol 3, whereas the least of problems were recorded in Protocols 2 and 4.
For children, wear time increased from one protocol to the next (Protocol 1 with 4.8 ± 2.7 days; 69% of the 7 days wear period vs. Protocol 4 with 5.7 ± 2.6 days; 81% of the 7 days; see Table 2). The difference in wear time was statistically significant between the four protocols, F(3, 1289) = 7.48, p < .001.
For adults, wear time was similar in Protocols 1 and 2. It increased in Protocol 3 and was longest in Protocol 4 (6.5 ± 2.2 days; 93% of the 7 days wear period; see Table 2). The difference in wear time between the four protocols was also statistically significant for adults, F(3, 4100) = 84.58, p < .001.
Both skin reactions, t(5, 387) = 11.26, p < .001, and adhesion problems, t(5, 387) = 11.34, p < .001, were more prevalent in children compared to adults. Adults achieved longer wear time than children, t(5, 387) = −13.73, p < .001.
Protocol 4 was the most staff-friendly in terms of ease of handling of the accelerometers (see Table 1).
Discussion
When planning to use Axivity AX3 accelerometers for research purposes, issues related to compliance, potential risk for participants, and ease of handling should be recognized and appraised. The findings of our study indicate that the choice of attachment protocol for skin mounting of accelerometers influences all these factors.
Only few studies have reported on skin reactions (Duncan et al., 2018; Schneller et al., 2017) and adhesion (Duncan et al., 2018) when using tape-mounted Axivity AX3 devices and, to the best of our knowledge, we are the first to compare different attachment protocols. The attachment protocol used by Schneller et al. (2017) in their study on 903 children aged 9–13 years corresponds to Protocol 1 in the present study, but the paper provided no information about skin reactions and adhesion problems. Complete wear time (seven consecutive days) was obtained for 72% of the thigh-worn accelerometers (Schneller et al., 2017), which is in line with our finding of a 69% completed wear period for children in Protocol 1. Our finding of a mean wear time of 5.7 days for children in Protocol 4 is high compared to a study on children using belt-worn GT1M and GT3X ActiGraph accelerometers with a median number of valid days of 4 out of 7 days (Olesen et al., 2013). This supports that tape-mounted devices may improve wear compliance compared to belt-worn accelerometers. Our finding of a mean wear time of 6.5 days for adults in Protocol 4 is similar to a Norwegian study of adults using belt-worn GT1M ActiGraph units that reported a wear time of 6.8 days out of seven (97% completion; Hansen et al., 2012). This is high compared to a U.S. population study where only 72% completed 4 out of 7 days (Troiano et al., 2008). An important difference between the present and the Norwegian study is that the Norwegian study sample consisted only of individuals with a minimum of 4 valid days of accelerometer measurement (Hansen et al., 2012), whereas we included everybody who consented to wear accelerometers independent of wear time, and thus, the results are not directly comparable.
We kept an experience-based and iterative approach throughout the whole data collection period and continued to change the protocol in force if we experienced need for modifications. Hence, we ended up having tested four different attachment protocols as shown in Table 1. Unique to Protocol 1 (see Table 1) which resulted in frequent skin reactions in both children and adults, was the use of a nonmedical, double-sided tape between the medical stretch tape on the skin and the accelerometer. We became concerned that the reported skin reactions could be due to this nonmedical tape and therefore, it was abandoned in Protocol 2 (see Table 1). Nevertheless, it was our impression that skin reactions still occurred frequently, which the present statistical analysis has confirmed. Currently, the knowledge with skin reactions in relation to taping of devices onto the skin is limited, which makes it difficult to further improve the methodology based on previous studies. However, to move forward with reducing adverse effects of taping devices onto the skin, we approached a dermatologist. Based on this expert’s input, we focused on the occluded and humid environment created by the stretch tape that covered the skin. Humidity from the evaporation from the skin and from showers and water activities could not dry up under the tape (Zhai & Maibach, 2001). This problem was reduced in Protocols 3 and 4 (see Table 1) where a cotton gauze pad was mounted between the skin and the accelerometer, and where the fixation tape on top of the accelerometer did not cover the skin completely, thus allowing for ventilation and drying of the surface of the skin. This may explain the reduction in self-reported skin reactions in both children and adults confirmed in the statistical analysis.
However, after the introduction of Protocol 3 (see Table 1), we experienced that accelerometers started to fall off more frequently, which is now confirmed by the statistical analysis. Sweaty skin, high air temperature, exercise, and water activities may challenge tape adhesion; feedback from the participants and LOFUS staff indicated that the pieces of tape for fixation might have been too short. Therefore, we introduced Protocol 4 (see Table 1 and Figure 3) and included one-third longer pieces of fixation tape. The statistical analysis confirmed our experience that this was a good solution as adhesion problems were only minor for both children and adults in Protocol 4.
Overall, our findings indicate that skin mounting of accelerometers is a suitable procedure for both children and adults. Still, there seems to be more challenges when employing the accelerometers in children when compared to adults. In terms of skin reactions, the skin of children may be slightly more sensitive, that is, prone to irritant reactions, than the skin of adults (Patel & Nixon, 2022). In relation to adhesion and wear time, adults may have higher awareness of protecting the device and thus keeping the accelerometer in place compared to children. Furthermore, parents may have lower tolerance for skin reactions in their children than in themselves, resulting in higher reporting for children. A concern related to adhesion may be the size of the tape-solution relative to the size of the body, for example, when young children wear diapers, when getting dressed, and during movement while sleeping.
Regarding ease of handling, Protocols 1–3 (see Table 1) were almost equally problematic. Removing remains of tape glue from the accelerometers required use of alcohol wipes, strenuous manual work, and many working hours. The risk of work-related injuries on hands, arms, and shoulders made it necessary to avoid long sessions of cleaning work. This problem was completely eliminated when we introduced Protocol 4 (see Table 1 and Figure 3) because the way the cotton gauze pad was folded around the device completely prevented remains of glue sticking to the accelerometer. Thus, we achieved a considerable reduction of burden on the LOFUS staff and saved many working hours.
Strengths and Limitations
Strengths of the present study are that minimal data were missing and we had a large sample, including both children and adults. Weaknesses are that skin problems were only self-reported and not assessed by a physician, and that we did not ask the participants to report problems for the thigh-worn and the back-worn accelerometer separately. A recommendation for future studies would be to use a validated questionnaire for reporting of specific skin reactions and severity of reactions instead of open-ended questions. Including meteorological data on air temperature and humidity may reveal potential seasonal variation in skin reactions and adhesion problems.
Conclusion
A procedure for attachment of accelerometers should be part of the planning when using tape-mounted Axivity AX3 devices on the skin to assess free-living PA. In LOFUS, which included children and adults, we tested four different attachment protocols, and our conclusion is that Protocol 4 was superior when taking skin reactions, adhesion, wear time, and ease of handling into consideration. Therefore, we recommend folding a 50 × 50-mm cotton gauze pad around the Axivity AX3 accelerometer with a double layer on the skin side before fixating it with two pieces of 28 × 150-mm medical adhesive tape across the vertical and horizontal direction of the accelerometer.
Acknowledgments
The LOFUS was registered with ClinicalTrials.gov (Identifier NCT 02482896). The LOFUS, Nykøbing Falster Hospital, Denmark, is a collaboration between Region Zealand, Nykøbing Falster Hospital, and Lolland and Guldborgsund Municipalities. The authors are grateful to LOFUS for making its research data available. However, LOFUS bears no responsibility for the analysis or the interpretation conducted within this study. The authors thank the participants and the LOFUS staff for their contributions.
References
AX3 data sheet, version 1.3. (2015). Newcastle University, UK. https://www.axivity.com/files/resources/AX3_Data_Sheet.pdf
Brønd, J.C., Grøntved, A., Andersen, L.B., Arvidsson, D., & Olesen, L.G. (2020). Simple method for the objective activity type assessment with preschoolers, children and adolescents. Children, 7(7), 72. https://doi.org/10.3390/children7070072
Dowd, K.P., Szeklicki, R., Minetto, M.A., Murphy, M.H., Polito, A., Ghigo, E., van der Ploeg, H., Ekelund, U., Maciaszek, J., Stemplewski, R., Tomczak, M., & Donnelly, A.E. (2018). A systematic literature review of reviews on techniques for physical activity measurement in adults: A DEDIPAC study. International Journal of Behavioral Nutrition and Physical Activity, 15(1), 15. https://doi.org/10.1186/s12966-017-0636-2
Duncan, S., Stewart, T., Mackay, L., Neville, J., Narayanan, A., Walker, C., Berry, S., & Morton, S. (2018). Wear-time compliance with a dual-accelerometer system for capturing 24-h behavioural profiles in children and adults. International Journal of Environmental Research and Public Health, 15(7), 1296. https://doi.org/10.3390/ijerph15071296
Edwardson, C.L., Henson, J., Bodicoat, D.H., Bakrania, K., Khunti, K., Davies, M.J., & Yates, T. (2017). Associations of reallocating sitting time into standing or stepping with glucose, insulin and insulin sensitivity: A cross-sectional analysis of adults at risk of type 2 diabetes. BMJ Open, 7(1), Article e014267. https://doi.org/10.1136/bmjopen-2016-014267
Egholm, C.L., Packness, A., Stokholm, J., Rasmussen, K., Ellervik, C., Simonsen, E., Christensen, A.I., & Jepsen, R. (2020). Questionnaire development for the Lolland-Falster Health Study, Denmark: An iterative and incremental process. BMC Medical Research Methodology, 20(1), 52. https://doi.org/10.1186/s12874-020-00931-1
Gupta, N., Christiansen, C.S., Hallman, D.M., Korshøj, M., Carneiro, I.G., & Holtermann, A. (2015). Is objectively measured sitting time associated with low back pain? A cross-sectional investigation in the NOMAD study. PLoS One, 10(3), Article e0121159. https://doi.org/10.1371/journal.pone.0121159
Hansen, B.H., Kolle, E., Dyrstad, S.M., Holme, I., & Anderssen, S.A. (2012). Accelerometer-determined physical activity in adults and older people. Medicine & Science in Sports & Exercise, 44(2), 266–272. https://doi.org/10.1249/MSS.0b013e31822cb354
Jepsen, R., Egholm, C.L., Brodersen, J., Simonsen, E., Grarup, J., Cyron, A., Ellervik, C., & Rasmussen, K. (2020). Lolland-Falster Health Study: Study protocol for a household-based prospective cohort study. Scandinavian Journal of Public Health, 48(4), 382–390. https://doi.org/10.1177/1403494818799613
Johansson, M.S., Holtermann, A., Marott, J.L., Prescott, E., Schnohr, P., Korshøj, M., & Søgaard, K. (2022). The physical activity health paradox and risk factors for cardiovascular disease: A cross-sectional compositional data analysis in the Copenhagen City Heart Study. PLoS One, 17(4), Article e0267427. https://doi.org/10.1371/journal.pone.0267427
Lopes, M.V.V., da Costa, B.G.G., Malheiros, L.E.A., Costa, R.M., Souza, A.C.C., Crochemore-Silva, I., & Silva, K.S. (2021). Correlates of the adherence to a 24-hr wrist-worn accelerometer protocol in a sample of high school students. Journal for the Measurement of Physical Behaviour, 4(4), 277–286. https://doi.org/10.1123/jmpb.2020-0062
Migueles, J.H., Cadenas-Sanchez, C., Ekelund, U., Delisle Nyström, C., Mora-Gonzalez, J., Löf, M., Labayen, I., Ruiz, J.R., & Ortega, F.B. (2017). Accelerometer data collection and processing criteria to assess physical activity and other outcomes: A systematic review and practical considerations. Sports Medicine, 47(9), 1821–1845. https://doi.org/10.1007/s40279-017-0716-0
Olesen, L.G., Kristensen, P.L., Korsholm, L., & Froberg, K. (2013). Physical activity in children attending preschools. Pediatrics, 132(5), e1310–e1318. https://doi.org/10.1542/peds.2012-3961
Patel, K., & Nixon, R. (2022). Irritant contact dermatitis—A review. Current Dermatology Reports, 11(2), 41–51. https://doi.org/10.1007/s13671-021-00351-4
Petersen, T.L., Brond, J.C., Kristensen, P.L., Aadland, E., Grontved, A., & Jepsen, R. (2020). Resemblance in accelerometer-assessed physical activity in families with children: The Lolland-Falster Health Study. International Journal of Behavioral Nutrition and Physical Activity, 17(1), 161. https://doi.org/10.1186/s12966-020-01067-7
Schneller, M.B., Bentsen, P., Nielsen, G., Brønd, J.C., Ried-Larsen, M., Mygind, E., & Schipperijn, J. (2017). Measuring children’s physical activity: Compliance using skin-taped accelerometers. Medicine & Science in Sports & Exercise, 49(6), 1261–1269. https://doi.org/10.1249/MSS.0000000000001222
Troiano, R.P., Berrigan, D., Dodd, K.W., Mâsse, L.C., Tilert, T., & Mcdowell, M. (2008). Physical activity in the United States measured by accelerometer. Medicine & Science in Sports & Exercise, 40(1), 181–188. https://doi.org/10.1249/mss.0b013e31815a51b3
Troiano, R.P., McClain, J.J., Brychta, R.J., & Chen, K.Y. (2014). Evolution of accelerometer methods for physical activity research. British Journal of Sports Medicine, 48(13), 1019–1023. https://doi.org/10.1136/bjsports-2014-093546
Zhai, H., & Maibach, H.I. (2001). Skin occlusion and irritant and allergic contact dermatitis: An overview. Contact Dermatitis, 44(4), 201–206. https://doi.org/10.1034/j.1600-0536.2001.044004201.x
