Quality control methods in accelerometer data processing: defining minimum wear time. — TEI-tagged view

Carly RichpersName, Marco GeracipersName, Lucy GriffithspersName, Francesco SerapersName, Carol DezateuxpersName, Mario Cortina-BorjapersName (2013date). “Quality control methods in accelerometer data processing: defining minimum wear time..” PLoS ONEtitle·j.

Object of study — thresholds for minimum daily wear timeterm and number of wear days needed for reliable accelerometerterm·key estimates of childrenterm·key's physical activityterm

Research question — whether gender and the purposeful inclusion of childrenterm·key with weekendterm·key data change the reliabilityterm of accelerometerterm·key measurement

Contribution — a minimum wear criterion of at least two days lasting at least 10 hours per day for population studies of childrenterm·key

Full text — tagged

The reference list is not part of the text we hold for this paper. Our copy was extracted from the publisher's page and the bibliography did not survive that extraction, so the article below ends before its references. Nothing has been removed by us.

Quality controlterm·key methods in accelerometerterm·key data processing: defining minimum wear timeterm·key.

Abstract When using accelerometersterm·key to measure physical activityterm, researchers need to determine whether subjects have worn their device for a sufficient period to be included in analyses. We propose a minimum wear criterion using population-based accelerometerterm·key data, and explore the influence of gender and the purposeful inclusion of childrenterm·key with weekendterm·key data on reliabilityterm.

Accelerometerterm·key data obtained during the age seven sweep of the UK Millennium Cohort Studyterm·dom were analysed. Childrenterm·key were asked to wear an ActiGraph GT1M accelerometerterm·key for seven days. Reliabilityterm coefficients( r ) of mean daily counts/minute were calculated using the Spearman-Brown formula based on the intraclass correlation coefficientterm. An r of 1.0 indicates that all the variation is between- rather than within-children and that measurement is 100% reliable. An r of 0.8 is often regarded as acceptable reliabilityterm. Analyses were repeated on data from childrenterm·key who met different minimum daily wear timesterm (one to 10 hours) and wear days (one to seven days). Analyses were conducted for all childrenterm·key, separately for boys and girls, and separately for childrenterm·key with and without weekendterm·key data.

At least one hour of wear timeterm·key data was obtained from 7,704 singletons. Reliabilityterm increased as the minimum number of days and the daily wear timeterm increased. A high reliabilityterm ( r = 0.86) and sample size ( n = 6,528) was achieved when childrenterm·key with ≥ two days lasting ≥10 hours/day were included in analyses. Reliabilityterm coefficients were similar for both genders. Purposeful sampling of childrenterm·key with weekendterm·key data resulted in comparable reliabilities to those calculated independent of weekendterm·key wear.

Quality controlterm·key procedures should be undertaken before analysing accelerometerterm·key data in large-scale studies. Using data from childrenterm·key with ≥ two days lasting ≥10 hours/day should provide reliable estimates of physical activityterm. It’s unnecessary to include only childrenterm·key with accelerometerterm·key data collected during weekendsterm·key in analyses.

Introduction Childrenterm·key’s physical activityterm (PAterm) is a difficult behaviour to measure as it is sporadic, intermittent and characterised by substantial inter- and intra-individual variation [1] . In recent years accelerometersterm·key have been regarded as the ‘gold standard’ method to examine PAterm in childhood populations [2] . Childrenterm·key are asked to wear their accelerometerterm·key for a fixed period of time, typically during all waking hours for seven consecutive days [3] , [4] , [5] . Despite various incentives and reminders, childrenterm·key rarely wear their accelerometerterm·key for this entire period. As a result, researchers need to determine whether each child wore their accelerometerterm·key for long enough to provide a reliable estimate of PAterm and be included in analyses. This can be achieved by defining the minimum number of minutes per day and the minimum number of days that the accelerometerterm·key needs to be worn by each child.

Reliabilityterm determines the consistency of a set of measurements or of a measuring instrument [6] . The duration of daily wear timeterm must be long enough to remove days when the accelerometerterm·key was not worn but short enough to prevent unnecessary days being removed from analyses, and the number of days the accelerometerterm·key needs to worn by each child must provide a reliable estimate of childrenterm·key’s habitual PAterm. No single value has been used by large-scale studies in childrenterm·key to define the minimum daily wear timeterm: thresholds have ranged from at least four [7] to at least 10 [3] , [4] , [5] , [8] , [9] hours per day. Large-scale accelerometerterm·key studies in childrenterm·key have also used a range of thresholds to define the minimum number of wear days required by each child to be included in analyses, although at least three days per child has been most commonly used [4] , [5] , [10] , [11] .

Given recent evidence that childrenterm·key’s PAterm varies according to the time of day and day of the week [12] , it is necessary to determine the minimum daily wear timeterm and the number of wear days required to reliably estimate childrenterm·key’s habitual activity. No studies have explored the influence of varying the minimum daily wear timeterm threshold on the reliabilityterm of accelerometer-determined PAterm measurement in pre-pubertal primary school-aged childrenterm·key. In those available, reliabilityterm estimates in preschool [13] , [14] and older childrenterm·key [15] have been investigated. Previous research has shown that the child’s age influences the minimum number of accelerometerterm·key wear days required to reliably estimate PAterm; it is therefore likely that the minimum daily wear timeterm is dependent on age. Researchers have looked at the influence of varying the minimum number of days required by each child to be included in analyses on the reliabilityterm of PAterm measurement but their findings are inconsistent, and study populations tend to be geographically clustered [16] , [17] , [18] .

There are substantial gender differences in childrenterm·key’s PAterm [5] , [8] , [11] ; however, previous research on the influence of varying the thresholds used to define minimum wear timeterm·key have combined data from boys and girls. Although findings are inconsistent, previous research also found gender differences between childrenterm·key who did and did not provide reliable data in childrenterm·key’s accelerometerterm·key studies [3] , [11] , [15] , [19] . Furthermore, studies have found that childrenterm·key’s PAterm varies between weekdays and weekendterm·key days [12] . Despite this, few studies [8] , [12] , [20] have considered whether or not childrenterm·key with week and weekendterm·key wear days are required to reliably estimate habitual PAterm.

Given the impact that these data processing procedures can have on derived activity variables and lack of previous research in pre-pubertal primary school aged childrenterm·key, further clarification on data cleaning methods is needed for researchers using these devices [21] , [22] , [23] . Esliger et al [24] emphasised the need for studies to evaluate within- and between-day variations in PAterm and in particular how these vary by gender. The aim of this study was to propose a threshold for the minimum number of hours per day and the minimum number of days of data required from each child to achieve reliable estimates of PAterm in population-based accelerometerterm·key studies. The influence of gender and the purposeful inclusion of childrenterm·key with and without weekendterm·key day data was also explored.

Methods Study Population We analyse population-based accelerometerterm·key data obtained as part of the Millennium Cohort Studyterm·dom (MCSterm·dom). The MCSterm·dom is a longitudinal UK-wide prospective study of childrenterm·key born in the new century sampled to ensure an adequate representation of all four UK countries, disadvantaged areas, and ethnic minority groups [25] . At age seven years, accelerometersterm·key were used to measure childrenterm·key’s PAterm levels. All childrenterm·key were invited to wear an accelerometerterm·key and written consent was obtained from parents/guardians of those agreeing.

Accelerometerterm·key Protocol Activity was measured using the ActiGraph GT1M (ActiGraph, Florida, USA), a small (3.8×3.7×1.8 cm), lightweight (27 g) uni-axial accelerometerterm·key that measures volumes and patterns of activity. The ActiGraph has been extensively validated in childrenterm·key [26] , [27] , [28] , and is robust when used in large-scale studies in childrenterm·key [3] , [4] , [5] , [8] . A 15-second sampling epoch was selected in order to optimize the ability to capture the sporadic nature of childrenterm·key’s activity [1] . Childrenterm·key were asked to wear the accelerometerterm·key on an elasticated belt on the right hip for seven consecutive days during all waking hours, except during bathing or swimming. Accelerometersterm·key were posted to families who were asked to return it as soon as possible after the monitoring period using a supplied pre-paid envelope. Accelerometersterm·key were distributed between May 2008 and August 2009.

Ethical approval for the MCSterm·dom accelerometerterm·key study was granted by the Northern and Yorkshire Research Ethics Committee (REC number: 07/MRE03/32). The MCSterm·dom data for surveys 1 to 4 are currently available via the Economic and Social Data Service; the MCSterm·dom accelerometerterm·key data will be also be available shortly at the beginning of 2013.

Statistical Analyses Accelerometerterm·key data were downloaded using ActiLife Lifestyle Monitoring software (version 3.2.11) and processed using algorithms developed in R (version 2.14.1) [29] . Accelerometer non-wearterm was defined as any time period of consecutive zero-counts for a minimum of 20 minutes [8] . Data from all singleton childrenterm·key who returned an accelerometerterm·key with at least one hour of wear timeterm·key data (periods in which non-wear was not identified) were eligible for inclusion in our analyses ( n = 7,704). Twins and triplets were not included in the analyses because data were unintentionally not coded to allow the interview and accelerometerterm·key data for twins and triplets to be accurately linked.

All analyses were repeated using different samples depending on whether childrenterm·key met the varying threshold used to define wear timeterm·key based on the minimum daily wear timeterm (one to 10 hours) and the minimum number of wear days (one to seven days). Analyses were conducted for boys and girls combined, and separately. Analyses were also repeated separately for childrenterm·key that did, and did not, have at least one weekendterm·key wear day (of at least 10 hours wear). This wear timeterm·key period was chosen because it is most often used by large-scale studies in childrenterm·key to define the minimum daily wear period [3] , [4] , [5] , [8] , [9] .

The reliabilityterm of accelerometer-determined mean daily counts per minuteterm (cpmterm) was calculated using the Spearman-Brown prophecy formulaterm·key [30] , [31] based on the intraclass correlation coefficientterm (ICCterm) as a measure of reliabilityterm. The distribution of mean daily cpmterm was skewed so the Box-Cox family of transformations were used to account for non-normality [32] . The asymmetry parameter in this family was chosen by maximising the profile log-likelihood using the R function boxcox [33] . A linear mixed-effects (LME) model was fitted to the transformed cpmterm using the MCSterm·dom survey and non-response weights to account for the clustered sampling and attrition between contacts [34] . Single day ICCterm were calculated from the fitted LME models with the R function ICC1.lme [35] . The ICCterm describes how strongly units in the same group resemble each other, and is defined as the ratio of between-individual variance to the sum of the between- and within- individual variance [15] . The ICCterm is the most common way of summarizing the consistency of measurement across days [36] . An ICCterm value of 1.0 indicates that all the variation is between- rather than within-children, corresponding to perfect reliabilityterm or repeatability. An ICCterm value of 0.8 is commonly regarded as a marker of acceptable reliabilityterm.

Single-day ICCterm values were then used to calculate the influence of shortening or lengthening the monitoring period on the reliabilityterm of PA-measurement using the Spearman-Brown prophecy defined in the following equation: where: N = the number of days required, ICCsterm = single-day reliabilityterm [37] . We used heatmaps developed in R to produce graphical representations of reliabilityterm trends by minimum daily wear timeterm and minimum days of wear day time.

Results Total Sample A total of 13,681 singleton childrenterm·key were interviewed at age seven years in the MCSterm·dom: 12,872 (94.1%) of these parents/guardians gave consent for their child to wear an accelerometerterm·key. Accelerometersterm·key were sent to 12,303 (89.9%) consenting singletons; 27 (0.2%) singletons were not sent an accelerometerterm·key because the fieldwork team were unable to send it during the requested time period, and full contact details of the remaining 542 (4.2%) singletons were unavailable. A total of 9,772 singletons returned an accelerometerterm·key, of which 1,106 parents/guardians explicitly stated that the accelerometerterm·key had not been worn. A total of 7,704 (59.9% of consenting singletons) singleton childrenterm·key returned an accelerometerterm·key with at least one hour of wear timeterm·key data ( Table 1 ). There were 5,878 childrenterm·key with files that contained at least one hour of wear timeterm·key data for greater than seven days and who had presumably worn the accelerometerterm·key for longer than the wear timeterm·key period requested.

Number of childrenterm·key included in analyses according to minimum daily wear timeterm and minimum number of wear days for the total sample and by gender (boys, girls).

The reliabilityterm of PAterm measurement was influenced by the minimum daily wear timeterm and the minimum number of days of data required by each child for inclusion in analyses ( Table 2 ). Reliabilityterm coefficients increased as the minimum number of days required by each child for inclusion in analyses increased (between one to ten days) and also increased as the minimum daily wear timeterm increased from at least one hour per day up to, but no more than, at least eight hours per day.

Reliabilityterm coefficients derived according to minimum daily wear timeterm and minimum number of wear days (total sample).

Reliabilityterm was low when childrenterm·key with at least one day lasting between one to three hours were included in analyses (36%–40%). PAterm measurement was more reliable when childrenterm·key with at least two days or greater were included in analyses. Measurement reliabilityterm values of at least 90% were achieved when the following thresholds were used to define which childrenterm·key were included in analyses: at least three days lasting at least eight hours per day, at least four or five days lasting at least six hours per day, and at least six days lasting at least five hours per day (90%, 90%, 92%, and 90% respectively). As defined by the Spearman- Brown prophecy formulaterm·key, the most reliable measure of PAterm (97%) was achieved when childrenterm·key with at least nine or 10 days lasting from at least eight to at least 10 hours per day were included in analyses. A high reliabilityterm and sample size was achieved when childrenterm·key with at least two days lasting at least 10 hours per day were included in analyses ( n = 6,528; reliabilityterm 86%).

Gender When reliabilityterm coefficients were calculated separately for boys and girls the results followed a similar trend in both genders to that found for the total sample ( Figure 1 ). Reliabilities were again influenced by the minimum daily wear timeterm and the minimum number of wear days. The reliabilityterm of PAterm measurement exhibited a minimal gender-related trend: measurement was slightly more reliable in girls than boys for nearly all combinations of minimum daily wear timeterm and minimum number of wear days. The most reliable measure was achieved in boys when childrenterm·key with at least nine or 10 days lasting from at least seven to at least 11 hours per day (96%) were included in analyses and in girls with at least 10 days lasting from at least eight to at least 10 hours per day (97%).

Reliabilityterm (%) Heatmaps for boys and girls by minimum daily wear timeterm and wear days.

Inclusion of Weekendterm·key Days A total of 2,414 singleton childrenterm·key (31.3% of all singletons returning data) returned an accelerometerterm·key that contained at least one day of data (≥ one hour) but no weekendterm·key day data. At least one weekendterm·key days’ worth of data (≥10 hours) was obtained from 5,290 singleton childrenterm·key (68.7% of all singletons returning data).

Reliabilityterm coefficients increased as the minimum daily wear timeterm and the minimum number of wear days increased in both childrenterm·key with and without weekendterm·key data. Reliabilities were slightly higher when only childrenterm·key with weekendterm·key data were included in analyses compared to childrenterm·key with only weekday data when wear timeterm·key was defined as at least four hours per day up to 13 hours per day for all numbers of wear days ( Figure 2 ). For example, when childrenterm·key with at least two days lasting at least 10 hours per day were included in analyses reliabilityterm was high in both childrenterm·key with and without weekendterm·key data but reached 82% in childrenterm·key with only weekday data compared to 88% in childrenterm·key with weekendterm·key data.

Reliabilityterm Heatmaps (%) for childrenterm·key with and without weekendterm·key data by minimum daily wear timeterm and wear days.

Reliabilities calculated when including only childrenterm·key with weekendterm·key data available were similar to those calculated when not purposely sampling childrenterm·key based on whether or not they had weekendterm·key data for all combinations of minimum daily wear timesterm and number of wear days. For example, the reliabilityterm of mean daily cpmterm calculated from childrenterm·key with at least two days of data lasting at least 10 hours per day was 88% in childrenterm·key with at least one weekendterm·key day of data compared to 86% when not purposely sampling childrenterm·key based on weekendterm·key wear.

Discussion Summary of Findings A threshold of at least two days lasting at least 10 hours per day can be used to screen subjects who provide reliable estimates of PAterm in population studies of older primary school aged childrenterm·key. This threshold provided a high reliabilityterm and sample size ( n = 6,528; reliabilityterm 86%). The use of this threshold also increases our ability to compare our findings with other studies, as the most common threshold previously used to define minimum daily wear timeterm was at least 10 hours per day [3] – [5] , [8] . The 80% reliabilityterm threshold has also been used by previous studies exploring the influence of varying the wear timeterm·key threshold on the reliabilityterm of PAterm measurement [13] , [16] . Both the minimum daily wear timeterm and the minimum number of wear days required by each child for inclusion in analyses influenced the reliabilityterm of PAterm measurement. Reliabilities were similar for both genders, although measurement in girls was slightly more reliable than boys for nearly all combinations of minimum daily wear timeterm and minimum number of wear days.

Reliabilities were slightly higher for childrenterm·key with weekendterm·key data compared to childrenterm·key with only weekday data. However, the purposeful sampling of childrenterm·key with at least one weekendterm·key day of data resulted in similar reliabilities for all wear timeterm·key thresholds to those calculated when using the total sample independent of weekendterm·key wear. Therefore, our results suggest that population data should encourage the measurement of PAterm on weekendterm·key days, but the purposeful sampling of subjects which forces the inclusion of weekendterm·key data in all childrenterm·key is not necessary.

Comparisons with Existing Research Only two previous studies have explored the influence of varying the minimum daily wear timeterm and the minimum number of days of data required from each child to be included in analyses. In contrast to our study, Mattocks et al [15] and Penpraze et al [13] found that the minimum daily wear timeterm had less influence on reliabilityterm than the number of wear days. Penpraze et al [13] calculated the reliabilityterm of PAterm in 76 five to six year old Scottish childrenterm·key: measurement reliabilityterm remained relatively stable using at least three hours per day up to, but no more than, at least 10 hours per day. The authors reported lower reliabilities than those calculated in the present study [13] : defining wear timeterm·key as at least seven days lasting at least 10 hours per day produced the highest reliabilityterm (80%, 95% CI = 70%, 86%). Penpraze et al [13] used a small geographically clustered sample and only included childrenterm·key with seven complete days of PAterm monitoring which has the potential to introduce bias in results.

Mattocks et al [15] found that the reliabilities remained constant using varying daily wear lengths (between seven to 10 hours) but whilst the number of wear days required per child to be included in analyses remained constant. The Avon Longitudinal Study of Parents and Childrenterm·key (ALSPAC) also reported lower reliabilities than our study using the same thresholds to define wear timeterm·key: childrenterm·key with at least 12 days lasting at least seven hours per day of data produced the highest reliabilityterm (90%). However, the ALSPAC used a threshold of at least three days lasting at least 10 hours per day to define minimum wear timeterm·key despite the reliabilityterm of measurement reaching only 70% when using data from childrenterm·key meeting this threshold.

A number of articles have examined the reliabilityterm of PAterm measurement using varying numbers of wear days without considering the influence of varying the daily wear length. The findings of these studies vary greatly, and are dependent on the age of the childrenterm·key and the study design. Studies have found that reliabilities of 80% are achieved when including childrenterm·key with a greater number of days of accelerometerterm·key data than calculated by our study. For example, at least five [10] to seven [13] wear days were required from preschool childrenterm·key (aged two to five years), four [16] , [38] to seven [17] , [38] wear days were required from childrenterm·key aged six to twelve years, and five [18] to nine [16] wear days were required from adolescents (aged 13 to 18 years). Despite these findings, other large-scale accelerometerterm·key studies in childrenterm·key have analysed data in childrenterm·key providing one day of accelerometerterm·key data [8] , [18] . For example, the Trial Activity for Adolescent Girls study included childrenterm·key in analyses who provided at least one day of data lasting at least six hours [18] .

To our knowledge, there have been no published studies exploring the influence of gender on the reliabilityterm of accelerometer-determined PAterm. Only a few studies have explored the influence of the distribution of wear days on the reliabilityterm of PAterm measurement in childrenterm·key [13] . In agreement with this study, Penpraze et al [13] found that the purposeful inclusion of childrenterm·key in analyses with weekendterm·key days had little effect on reliabilityterm estimates: using data provided by childrenterm·key with four days wear including a weekendterm·key day compared to using data provided by childrenterm·key with four weekday wear days only reduced reliabilityterm estimates from 84% to 82%. Mattocks et al [15] did not explore the influence of purposely sampling childrenterm·key based on weekendterm·key wear, but in agreement with other studies [8] , [12] , [20] , they found that childrenterm·key’s PAterm differed on weekendterm·key days compared to weekdays.

Only one previous large-scale study has evaluated the influence of varying the number and distribution of accelerometerterm·key wear days on the reliabilityterm of population estimates of PAterm. In contrast to our study, McClain et al [23] found that stable estimates of population PAterm can be obtained from only one randomly selected day out of a possible sampled week in 2532 US adults (aged 20 years). However, in agreement with our study, they also found that the purposeful sampling of subjects which forces the inclusion of a weekendterm·key day is not necessary.

Strengths and Limitations This is the first study to explore the influence of varying the minimum daily wear timeterm and the minimum number of days of accelerometerterm·key data on the reliabilityterm of PAterm measurement in a large-scale UK-wide study of childrenterm·key. It has been suggested that previous thresholds of minimum wear timeterm·key may have been overestimated because of violations in the assumptions associated with the ICCterm formula [36] , [39] . However, we have shown that high reliabilityterm values can be attained from childrenterm·key with a relatively small number of days and hours of wearing time. We have proposed a threshold that maximises both reliabilityterm and sample size whilst also taking into our account our ability to compare our findings with other studies. We are also confident that the study design, accelerometerterm·key protocol, and analytical methodologies employed here enable us to define a robust definition of wear timeterm·key. Accelerometerterm·key data often follow a skew distribution, and it is important to account for this asymmetry to achieve the normality assumption required to correctly compute the ICCterm. Our ICCterm values also take into account the MCSterm·dom survey and non-response weights. Uniquely, we have also explored the influence of gender and the distribution of wear days on the reliabilityterm of PAterm measurement. In doing so, we used data from a large, contemporary, socially- and ethnically- diverse cohort of childrenterm·key from all four UK countries.

Our proposed wear timeterm·key threshold may not be applicable for use in different ages [16] . PAterm levels vary according to age, and childrenterm·key’s PAterm is very different to adult’s PAterm in many respects [3] ; it is therefore unlikely that without further research the findings of this study can be used in adolescent or adult populations. Reliabilityterm values may also be dependent on the derived PAterm outcome variable. It has been widely documented that the use of different thresholds to define activity intensities limits the ability for researchers to make reliable comparisons of moderate to vigorous PAterm levels between studies, and at present there is still no consensus on the best threshold to use [40] . However, studies have found similar reliabilityterm values for PAterm measurement when accelerometerterm·key data were expressed as cpmterm or as the percentage of time in different activity intensities [13] , [41] .

Recommendations for Study Practice and further Research It is important that researchers using accelerometerterm·key data only analyse data from childrenterm·key that meet a pre-defined wear timeterm·key threshold. Using the proposed threshold will enhance quality controlterm·key processes by ensuring that only childrenterm·key that provide enough data to reliably estimate weekly PAterm are used in analyses without compromising sample size. If population studies do not screen accelerometerterm·key data prior to processing this may lead to unreliable estimates of childrenterm·key’s habitual activity levels. The proposed threshold is appropriate for use in boys and girls, although studies using samples of only girls may be able to use a less stringent definition than studies including both genders. It is important that subjects are asked to wear their accelerometerterm·key over an entire day, and that both weekdays and weekendterm·key days are requested in the monitoring period. However, the purposeful inclusion of childrenterm·key with weekendterm·key data in analyses is not necessary.

Future research should be aimed at calculating whether the proposed threshold is applicable across different age groups and in studies deriving different PAterm outcome variables. Furthermore, this study suggests that the inclusion of data from childrenterm·key with at least two days of accelerometerterm·key data (at least 10 hours/day) out of a possible seven day monitoring period provides a reliable estimate of population-based estimates of PAterm. Further research is required to determine whether this is applicable in studies that ask childrenterm·key to wear their monitor for only two days. Bias may be introduced when data from childrenterm·key with only two wear days are included in analyses, especially if these days are not randomly sampled from a possible seven day week. Although beyond the scope of this study, the removal of childrenterm·key that do not meet the wear timeterm·key threshold may be dealt with through imputation methods [42] , and future research is needed to explore such approaches to adjust for potential bias introduced by removing unreliable data.

Conclusions It is important for population-based studies to integrate a core set of quality controlterm·key procedures prior to deriving activity outcome variables: this should include the screening of data using a wear timeterm·key threshold. Using a threshold of at least two days lasting at least 10 hours per day will enhance data quality. This threshold is applicable in 7–8 year olds and in population-based studies that monitor childrenterm·key over a full week including the weekendterm·key. It is unnecessary to only include childrenterm·key with weekendterm·key data in analyses.