Abstract
Background There is ambiguity about what measures to use to best identify physical activity and sedentary behavior, and agreement between methods for measuring physical activity and sedentary behavior in people with obstructive sleep apnea syndrome (OSAS) and obesity has not been evaluated.
Objective The objective of this study was to examine the level of agreement between an accelerometer and a self-report questionnaire (International Physical Activity Questionnaire [IPAQ]) or a logbook for measuring time spent on moderate to vigorous physical activity and time spent sedentary in people with OSAS and obesity.
Design This prospective study was a psychometric evaluation of agreement between measurement methods.
Methods Thirty-nine people who were obese (body mass index: X̅=36.1 kg/m2, SD=4.35) and had moderate to severe OSAS (apnea-hypopnea index of ≥15) were consecutively recruited from a sleep clinic in Sweden. All were treated with continuous positive airway pressure and were waiting for a follow-up sleep evaluation.
Results Agreement between the measurement methods was limited. For physical activity, the mean difference between the accelerometer and the IPAQ was 47 minutes, and the mean difference between the accelerometer and the logbook was 32 minutes. Agreement was limited for sedentary time as well; the mean difference between the accelerometer and the IPAQ was 114 minutes, and the mean difference between the accelerometer and the logbook was 86 minutes.
Limitations The small sample size may affect the interpretation and generalizability of the results.
Conclusions The results imply that the methods cannot be used interchangeably. A combination of an accelerometer and a daily logbook seems to provide a detailed description of physical activity and sedentary behavior.
Measuring physical activity and sedentary time is a challenge because everyday life includes diverse activities at different intensities and durations. There is ambiguity regarding what measures to use to capture physical activity and sedentary behavior. Questionnaires are easy to administer, but a well-known drawback of self-reports is that they are dependent on the reporting person's memory as well as the type of definition used for physical activity.1 Recall may be even more biased if the reporting person has trouble concentrating. In people with obstructive sleep apnea syndrome (OSAS), the prevalence of sleepiness and problems concentrating is substantial. The nocturnal apneas that repeatedly occlude the airflow result in desaturation, fragmented sleep, and, in turn, daytime symptoms such as sleepiness and difficulty concentrating.2 No study has been conducted to distinguish which method optimally identifies physical activity and sedentary time in people with OSAS.
Physical activity has been defined as “any bodily movement produced by skeletal muscles that results in energy expenditure,”3(p126) and 30 minutes of moderate-intensity physical activity 5 days per week (or a total of 150 min/wk) has been recommended to maintain health and prevent chronic diseases.4 An entity frequently used in the assessment of physical activity is metabolic equivalents (METs [1 MET=3.5 mL O2·kg-1·min-1]). Through measurement of the energy expended in different types of physical activity, various METs have been derived as multiples of the oxygen consumption at rest (1.0 MET). Moderate-intensity physical activity has been set at 3 to 6 METs, and higher METs indicate vigorous-intensity physical activity. Most assessment methods produce physical activity data in terms of such intensity levels or METs.
The most commonly used measurements are self-reports, in which a person recalls the duration, frequency, or both of physical activities at different intensity levels during a particular period, such as the preceding week. Some questionnaires even include time spent sitting as a representation of sedentary time. However, values obtained from self-reports differ greatly from those obtained with various types of activity monitors, such as accelerometers or pedometers. Physical activity values obtained from self-reports are generally higher than those measured by accelerometers.5 Most accelerometers count the number of steps taken and register movements around 1 or 2 axes to compute energy expenditure.6 Accelerometers are worn in various positions, such as the lower back or the upper arm, depending on the brand. Some devices may register body position, such as lying down, sitting, or standing. With these features, even inactivity or sedentary time may be determined.
The fact that algorithms and cutoff points used for data analysis vary among devices and among research projects has complicated comparisons of both accelerometers and studies.6,7 Some newer devices have proprietary software that sums and translates the raw data into information such as time spent at different METs. This procedure simplifies homogeneous interpretation and comparison of studies.6
Patients with OSAS may have daytime sleepiness and difficulty concentrating, which, in turn, may contribute to decreased physical activity and increased sedentary time. We, therefore, consider such patients to be clinically important from a behavioral sleep medicine perspective. However, these patients are not frequently targeted by clinical physical therapists, at least not with regard to strategies for enhanced physical activity and reduced sedentary time. Physical activity has positive effects for these patients, not only through its contribution to weight loss but also because regular physical activity has been reported to reduce the severity of OSAS symptoms, even without changes in weight or anthropometric measures.8–10
In most studies of people with OSAS, physical activity levels were measured with self-report questionnaires.11,12 Activity monitors were used to assess physical activity outcomes in 2 studies. In one study,13 the results were not presented in terms of type, duration, frequency, or intensity—each of which is an aspect of interest in interventions aiming to enhance physical activity. In the other study,14 only number of steps was presented. To our knowledge, for people with OSAS, no study has reported sedentary time, and the agreement between methods for measuring physical activity and sedentary time has not been evaluated.
The aim of this study was to examine the level of agreement between an accelerometer and 2 self-report methods (a questionnaire and a logbook) for measuring daily amounts of time spent on moderate to vigorous physical activity (MVPA) and time spent sedentary.
Method
Design
This prospective study was a psychometric evaluation of agreement between measurement methods. Data were collected from January 2009 to March 2010. Assessments were done over 6 to 10 consecutive days for each participant.
Participants
Participants were consecutively recruited from a sleep clinic in Uppsala, Sweden. Medical records were used to screen potential participants for body mass index, apnea-hypopnea index, and heart disease. People who were obese (body mass index of ≥30 kg/m2), had moderate or severe OSAS (apnea-hypopnea index of ≥15) treated with continuous positive airway pressure (CPAP), and were on a waiting list for a follow-up sleep evaluation after the prescription of CPAP were considered eligible to participate in the study. People with sequelae after stroke, known atrial fibrillation, severe heart failure, or other symptomatic heart disease were excluded.
A written invitation to the study along with an invitation to the follow-up sleep evaluation were sent to 211 eligible people (51 women, 160 men) who were on CPAP treatment for OSAS and were on a waiting list for follow-up sleep evaluation. Forty-three people were willing to participate in the study. Two of them were excluded before assessment because they had body mass indexes of less than 30 kg/m2, 1 participant did not attend, and another participant was excluded because of illness. Therefore, 39 people (7 women, 32 men) were enrolled in the study. Table 1 provides an overview of the characteristics of the participants.
Characteristics of the Participants (n=39)a
Measurements
Accelerometer.
A SenseWear Pro 3 Armband (SWA, firmware 8.02.28, Body Media, Pittsburgh, Pennsylvania) was used in combination with the proprietary software SenseWear Professional 6.1. The armband registers input from heat sensors, information on galvanic reactions, and movement and acceleration around 2 axes. Data are registered by the sensors every second but are stored as an average each minute. The software reports this information in terms of steps taken and time spent at different intensity levels. The armband has fairly good validity. It estimates energy expenditure within 16% of the values derived by indirect calorimetry,15 and the intraclass correlation coefficient, determined with doubly labelled water, has been reported to be .80.16 The device tends to underestimate energy expenditure during higher-intensity activities, such as tennis, road running, and track running,15,16 and has poor accuracy in adults who are obese during cycle ergometry, stair stepping, and treadmill walking.17 A test-retest study of the ability of the device to estimate energy expenditure during 2 separate and structured 13-hour observations produced an intraclass correlation coefficient of .97.18
Self-reports of time spent on physical activities and sitting.
The short version of the International Physical Activity Questionnaire (IPAQ) is a 7-day recall questionnaire about time spent sitting and number of days and time spent on activities at different intensity levels (walking activities, moderate-intensity activities, and vigorous-intensity activities).19 The respondent is instructed to record the daily average amount of time spent sitting and the number of days and the daily average amount of time spent on the 3 activity domains. Only activities lasting 10 minutes or more are to be included.19 Craig et al19 studied the test-retest reliability of this version at 5 different data collection sites, and it was found to be reliable for both physical activity (r=.66–.89) and sitting (r=.71–.95). The questionnaire was found to have acceptable criterion-related validity (r=.16 and r=.21 for sitting and MVPA, respectively) compared with the accelerometer in a Swedish population.20
In logbooks produced for this study, participants logged their daily time spent sitting and the type, time, and duration of physical activities in which they participated. They also rated their perceived exertion (according to Borg's Rating of Perceived Exertion, 6–2021) for every activity. Only activities lasting 10 minutes or more were to be included.
Demographic and medical data.
The participants answered questions about marital status, education, occupation, diseases, medications, time with OSAS, and time receiving CPAP treatment.
Procedure
All participants were contacted by a physical therapist (first author) and scheduled for 2 hospital visits. Each participant met the physical therapist on 2 occasions within 6 to 10 days. During the first visit, the participants gave informed consent and were measured for weight and height. Then they answered background questions and completed the IPAQ. Next, they were informed about the accelerometer, and the device was applied on the back of the right upper arm. The participants received information about the logbook, and time was allotted for the participants to try it.
During the following week, the participants wore the accelerometer and concurrently made daily notes in their logbooks regarding their physical activities and time spent sitting. The accelerometer was worn for 5 to 7 consecutive days and nights, including at least 1 weekend day. The device was removed for activities during which it could become wet or could be damaged. During the second hospital visit (6–10 days later), the participants returned the accelerometer and the logbook and again completed the IPAQ.
Data Analysis
All data analyses were done with the Statistical Package for the Social Sciences (SPSS), version 19.0 (SPSS Inc, Chicago, Illinois).
SWA.
For the descriptive aspects of the study, the following entities were used: daily number of steps, minutes of MVPA (lasting 10 minutes or more; ≥3.0 METs), and daily time spent sedentary (0–1.0 MET). To be considered valid, data must have met the following criteria: the SWA must have been worn for at least 90% of the waking hours, and data must have been collected for at least 5 days, including 1 weekend day. The proprietary software cannot be set to identify bouts of a certain duration. Therefore, a visual analysis of the time axis was made to define bouts of MVPA of 10 minutes or more. All MVPA for a total of at least 10 minutes within a 15-minute period was defined as a bout of MVPA. If the activity was interrupted by low-intensity activity for 2 minutes or more or by any sedentary behavior, that activity was not included as a bout of MVPA. From these bouts we calculated the group's mean daily average minutes of MVPA. The participants wore the SWA both day and night; the proprietary software has a feature for classifying sleep through the diminished input from the sensors. However, every time period during the night that was not considered to be sleep was classified by the software as sedentary time. Therefore, to delimit sedentary time to only time awake, an average of 7.5 hours of sleep was subtracted from the identified total amount of sedentary time.
IPAQ.
The IPAQ recommendations for data processing22 were followed. According to these recommendations, the 3 activity domains in the questionnaire should be considered equivalent to the following METs: walking activities as 3.3 METs, moderate-intensity activities as 4.0 METs, and vigorous-intensity activities as 8.0 METs. Data were collected from the SWA for 5 days, but according to the IPAQ recommendations, 7 days of data should be recalled. To enable comparisons between measurements, the data from the IPAQ were transformed into a daily average MVPA with the following equation: [walking activities (minutes × days) + moderate-intensity activities (minutes × days) + vigorous-intensity activities (minutes × days)]/7. From this daily average, a value for 5 days could be derived. All comparisons were based on the IPAQ from the second hospital visit. A t test of the 2 IPAQs was used to identify potential differences in the amounts of MVPA and sedentary time.
Study-specific logbooks.
The self-reported activities were classified into MET levels23 but were adjusted according to the participant's rated exertion. Metabolic equivalents of 1 to 2.9 were interpreted as low-intensity physical activity, METs of 3 to 6 were interpreted as moderate-intensity physical activity, and METs of greater than 6 were interpreted as high-intensity physical activity.24,25 After data processing, the following entities were used: daily minutes of MVPA and daily time spent sitting (sedentary). We used only logbook data from days corresponding to the days on which valid data were obtained from the SWA.
Agreement between methods.
We used Bland-Altman analysis and an additional graphic plot26,27 to study the agreement between methods and to determine whether the methods were interchangeable with regard to their capacity to identify MVPA and sedentary time. In a Bland-Altman analysis, the difference between a participant's mean value obtained by 1 measurement method and the mean value obtained by a different method is contrasted to the mean value obtained by both methods together. The difference between the values from the 2 measurement methods are plotted on the y-axis, and the average of the 2 methods together is plotted on the x-axis, allowing the reader to determine whether the observed differences are clinically important or not. If they are not, then the methods are interchangeable.26,27
Determination of the level of agreement between methods required 4 calculations, as shown in Table 2. The accelerometer was considered to be our reference method for several reasons. First, it registers continuously while it is being worn, enabling more accurate data to be obtained. In addition, it has the ability to capture both physical activity and sedentariness through the expression of energy expenditure.
Scheme for Calculation of the Level of Agreement Between Measurement Methods
The analysis was carried out in several steps. First, the difference between values measured by the 2 methods for each participant was calculated with a dependent t test. Next, an average of the values measured by both methods was derived for every participant. Both the difference and the average for all participants were then plotted in a scatterplot, with the between-method difference on the y-axis and the average of the values obtained by the 2 methods on the x-axis. The mean difference between the methods and the 95% confidence intervals were marked. The mean difference ± 1.96 × the standard deviation of the mean difference was added to the plot to derive the limits of agreement. These limits defined the interval in which differences between methods could be expected for 95% of future measurements in comparable people.27 The plot showed the amount of agreement or disagreement between the methods (via the differences) and illustrated how this agreement or disagreement was related to the magnitude of the measured values. A mean difference close to y=0 was a good indicator of agreement, as was a confidence interval encompassing y=0.
The association between the differences in the measured values and their average then had to be evaluated to confirm whether the differences fluctuated in any systematic manner over the range of measured values.26,27 For the methods to be considered interchangeable in this study, they were expected to differ by no more than 15 minutes for MVPA and 60 minutes for sedentary time. These a priori hypotheses were based on 2 research findings. For MVPA, it seems to be important to detect a clinical change of 15 minutes.28 At present, there are no recommendations about how much sedentary time is considered acceptable to maintain health and prevent illness. However, research has indicated that every 1-hour increase in sedentary time per day is linked to an increase in the prevalence of metabolic syndrome.29 Therefore, a difference of greater than 60 minutes would indicate incongruence between methods.
Role of the Funding Source
This study was supported by the Swedish Research Council, Stockholm, Sweden, and by Caring Sciences Funding at the Faculty of Medicine, Uppsala University, Uppsala, Sweden. The funding agencies did not contribute to the study's design or conduct or to analysis of the results.
Results
Internal Attrition
Accurate data about MVPA were available in the logbooks from all 39 participants. However, because of invalid data, information from the SWA for 3 participants and from the IPAQ for 2 participants was excluded. For 3 participants, logbook data about time spent sitting were invalid. To determine the level of agreement, we included in each calculation only values from participants for whom complete data were available from both measurement methods (Tab. 2).
Agreement Between Methods
According to the logbooks, the participants engaged in a spectrum of activities, but the most commonly reported activity was walking. There was a statistically significant difference between the 2 IPAQs (first day versus 6–10 days later) regarding time spent sitting (X̅=67 minutes, SD=175 minutes, t=−2.17, df=31, P=.038) but not regarding time spent on MVPA (X̅=23 minutes, SD=81 minutes, t=1.66, df=32, P=.107).
Time Spent on MVPA
According to the SWA, the IPAQ, and the logbooks, the daily average minutes spent on MVPA by the entire group were 37 (SD=44), 84 (SD=80), and 69 (SD=69), respectively. Figure 1 shows 2 Bland-Altman plots of the MVPA data. For each calculation, the boundaries of the confidence interval for the mean of the 2 methods were on the same side of y=0, indicating statistically significant differences between the SWA and the IPAQ of 47 minutes and between the SWA and the logbooks of 32 minutes. The limits of agreement were −204 to 110 minutes per day for the SWA and the IPAQ and −173 to 109 minutes per day for the SWA and the logbooks. The dispersion of the differences was wide, indicating less agreement. The greater the amount of time spent on MVPA, the larger the difference between methods (Fig. 1).
Bland-Altman plots of the difference between an accelerometer (SenseWear) and the International Physical Activity Questionnaire (IPAQ) (top) and the difference between the SenseWear and logbooks (bottom) for measuring moderate to vigorous intensity physical activity in people with obstructive sleep apnea and obesity. For the SenseWear, only bouts of 10 minutes or more were included in the analysis. The thick line indicates the mean difference (in minutes) between the 2 methods. The thin lines indicate the 95% confidence interval of the mean difference. The broken lines indicate the limits of agreement (1.96 × the standard deviation of the mean difference). METs=metabolic equivalents.
Time Spent Sitting
According to the SWA, the IPAQ, and the logbooks, the average time spent sitting was 9 hours (SD=3) 55 minutes (SD=10), 7 hours (SD=2) 51 minutes (SD=59), and 8 hours (SD=3) 21 minutes (SD=0), respectively. Figure 2 shows that there was a statistically significant difference between the SWA and the IPAQ because the boundaries of the confidence interval were on the same side of y=0. However, there was no statistically significant difference between the SWA and the logbooks because the boundaries of the confidence interval were on either side of y=0, even though the lower boundary of the confidence interval just barely passed y=0 (−1). The limits of agreement were −362 to 590 minutes per day for the SWA and the IPAQ and −396 to 568 minutes per day for the SWA and the logbooks, so the dispersion of the differences was wide. However, for time spent sitting, no trend regarding the level of agreement and the magnitude of the measured values was seen (Fig. 2).
Bland-Altman plots of the difference between an accelerometer (SenseWear) and the International Physical Activity Questionnaire (IPAQ) (top) and the difference between the SenseWear and logbooks (bottom) for measuring daily time spent sedentary in people with obstructive sleep apnea and obesity. The thick line indicates the mean difference (in minutes) between the 2 tests. The thin lines indicate the 95% confidence interval of the mean difference. The broken lines indicate the limits of agreement (1.96 × the standard deviation of the mean difference). METs=metabolic equivalents.
Discussion
The present study is the first to evaluate the levels of agreement of methods for measuring physical activity and sedentary time in people with OSAS and obesity. The results indicated that different values might be expected for both MVPA and sedentary time depending on the method used. Self-reported physical activity tended to be overestimated by both the IPAQ and the logbooks compared with the accelerometer, and the differences increased as the amount of MVPA increased. Because valid data were obtained from the SWA for a total of 5 days, only logbook data from days corresponding to the days on which valid data were obtained from the SWA were used. Thus, data obtained from the IPAQ had to be converted to 5 days instead of 7 days to enable an analysis of the agreement between measurement methods. Different days or certain activities cannot be identified with the IPAQ; therefore, physical activity reported in the IPAQ might not have been included in data obtained from the SWA or reported in the logbooks. These factors could explain the disagreement between the IPAQ and the SWA and the finding that the greater the physical activity reported in the IPAQ, the larger the difference between the methods. Even a daily increase of 15 minutes of MVPA is beneficial.28 The difference between the methods indicated that such an increase might be missed (eg, in an intervention trial) depending on the method used.
These results are in line with those of a validation study of the short version of the IPAQ in a Swedish general population, in which self-reported time spent on physical activity (both walking and MVPA) was significantly different from time measured by accelerometry.20 In addition, people with OSAS might have daytime sleepiness and a reduced ability to concentrate; these factors might increase recall bias further. With the IPAQ, there might have been differences in how the respondents classified “moderate” and “vigorous” physical activities. With the logbooks, the respondents did not have to make this distinction but could rate perceived exertion (according to Borg's Rating of Perceived Exertion, 6–20) for each of their activities. Such data provide more information about the particular physical activities performed and might allow the classification of a person's activities into predetermined intensity levels.
In an Australian study, a 2-day diary correlated moderately well with accelerometer data for MVPA (r=.45–.69).30 However, an earlier edition of the SWA was reported to be less valid for certain activities of a higher intensity, such as tennis, road running, and track running15; in addition, for people who were overweight or obese, the accuracy during cycle ergometry, stair stepping, and treadmill walking was poor.17 Hence, it is possible that the accelerometer failed to identify such activities. According to the logbooks, only 2 people used a treadmill, and neither cycle ergometry nor stair stepping was recorded. In addition, it is possible to conclude that very few vigorous activities were carried out. Therefore, the evaluation of vigorous activities by the SWA likely was valid in our participants, but the accuracy of the SWA for the intensities of vigorous activities in people who are obese must be studied further.
If a logbook is designed to record both type of activity and exertion, there may be a benefit to combining an accelerometer with a logbook to obtain a detailed description of a person's activity patterns. However, it is important to design a user-friendly logbook. Unambiguous instructions and a variety of examples of activities may make it easier for a person to record every type of physical activity.
Self-reported sedentary time (IPAQ and logbooks) seemed to be underestimated compared with the data obtained from the accelerometer. At present, there are no recommendations about how much sedentary time is considered acceptable for maintaining health and preventing illness, but there is at least research indicating a negative impact on health with every daily 1-hour increase in sedentary time.29 However, data obtained with both the IPAQ and the logbooks differed by more than 60 minutes from data obtained with the SWA; therefore, in the present study, these methods were not interchangeable in their ability to capture this variation. There was a statistically significant difference only between the SWA and the IPAQ, indicating slightly better agreement between the SWA and the logbooks. A decreased ability to concentrate because of sleep fragmentation and daytime sleepiness might accentuate recall bias, especially for 7-day recall (IPAQ) compared with daily notations (logbooks).
In the 2 self-report methods, the participants were instructed to record time spent sitting, whereas the SWA registered physical activity at a certain intensity level (0–1.0 MET). The latter might have included both sitting and lying down; this factor could be an additional explanation for the disagreement between the methods. In previous studies validating the IPAQ in Swedish populations, time spent sitting was only modestly correlated with accelerometer data (r=.1222 and r=.1620). In addition, in people with OSAS the higher prevalence of daytime sleepiness might contribute to the discrepancy between the methods. Perhaps there would be less of a difference between measurements if lying down were incorporated into the operationalization of sedentary time in the self-report methods.
In an Australian study, a diary correlated moderately well with accelerometer data for nonoccupational sedentary behavior (r=.57–.59).30 In the present study, the participants completed the IPAQ before and immediately after the periods during which they used the logbooks and the SWA. In a previous test-retest study of the IPAQ, high correlations regarding time spent sitting were found19; however, for time spent sitting in the present study, a statistically significant difference between the 2 IPAQ assessments was found. Perhaps some participants were unaware of their sitting behavior and using logbooks with daily registration raised their awareness about this behavior.
To help people identify periods of muscular inactivity (not just sitting), it seems to be important to assess all sedentary behaviors. For this purpose, an accelerometer has an advantage because it registers every occurrence of sedentary behavior. Therefore, a combination of data from accelerometers and specific self-reports about sitting seems to be worthwhile for surveying both the amount and the duration of sedentary time and for capturing specific situations in which sedentary time might be reduced.
Some details must be considered in the interpretation of the results of the present study. A Bland-Altman analysis should be used for methods with reasonable repeatability to minimize the impact of unstable methods on the results.26 Regarding the measurements used in the present study, both the SWA and the IPAQ have been found to be stable in test-retest studies; however, because the logbook was designed especially for the present study, it has not undergone such a psychometric evaluation. A logbook may be designed in various ways depending on the purpose of the assessment. This flexibility is offset by uncertain reliability. However, some authors have reported positive psychometrics for the use of diaries or logbooks30,31; in addition, it seems that this type of measurement method may offer additional information, such as perceived exertion, that cannot be obtained with either accelerometers or self-report questionnaires, such as the IPAQ.
The limited number of participants could have had an impact on the results, for example, through vulnerability to outliers. We examined the data for such values; however, no corrections seemed to be necessary because all of the extreme values originated from the same group of participants, indicating intraindividual consistency. In addition, depicting the participants' data with Bland-Altman plots provided an opportunity for a transparent interpretation of the impact of such variances. The small sample size also affected the generalizability of the present study. There was no information about physical activity and sedentary behavior or cognitive, disease-related, and social factors in people declining participation. The participants might have been more interested and motivated to engage in physical activity than people declining participation. It was not possible to rule out the effects of such factors on the external validity of the results. The present study addressed only people receiving CPAP treatment and attending a hospital sleep clinic. This means that people receiving treatment with oral appliances and attending other health care facilities were not represented in the present study. Our setting was deliberately chosen to represent a target for future interventions aiming to enhance physical activity and reduce sedentary behavior in people who are obese and have OSAS treated with CPAP.
In conclusion, in the present study we evaluated the level of agreement between methods for measuring MVPA and sedentary behavior in people with OSAS and obesity. The results indicated that the methods cannot be used interchangeably. Rather, a combination of an accelerometer and a daily logbook seems to provide a detailed description of physical activity and sedentary behavior patterns. This topic was not previously fully investigated; therefore, the results of the present study may contribute to both knowledge and tools for assessments in interventions aiming to enhance physical activity and reduce sedentary behavior in people with OSAS.
The Bottom Line
What do we already know about this topic?
Activity monitoring using accelerometry is a more reliable method of measuring physical activity and sedentary time compared with self-reports; however, agreement between measurement methods has not been studied in people with obstructive sleep apnea syndrome (OSAS). In addition, little is known regarding sedentary behavior in this population.
What new information does this study offer?
This study is the first to report sedentary behavior in people with OSAS and obesity, and it adds knowledge about physical activity in this group. The International Physical Activity Questionnaire (IPAQ) and daily logbook used in the study overestimated moderate to vigorous physical activity compared with an activity monitor and underestimated sedentary time. This discrepancy might be accentuated due to the symptoms of daytime sleepiness prevalent in this population, making recall harder.
If you're a patient or caregiver, what might these findings mean for you?
If you have OSAS and obesity, your physical therapist may recommend that you use a combination of an activity monitor and daily logbook to assess your physical activity and sedentary behavior.
Footnotes
-
All authors provided concept/idea/research design, writing, data analysis, and project management. Ms Igelström and Dr Lindberg provided data collection. Dr Åsenlöf provided fund procurement. Dr Lindberg provided study participants. Ms Emtner and Dr Lindberg provided facilities/equipment. Ms Emtner and Dr Åsenlöf provided institutional liaisons. Ms Emtner, Dr Lindberg, and Dr Åsenlöf provided consultation (including review of manuscript before submission).
-
This study was approved by the Regional Ethical Review Board of Uppsala University.
-
A poster of this research was given at the Annual Congress of the European Respiratory Society; September 24–28, 2011; Amsterdam, the Netherlands.
-
This study was supported by the Swedish Research Council, Stockholm, Sweden, and by Caring Sciences Funding at the Faculty of Medicine, Uppsala University, Uppsala, Sweden.
- Received March 23, 2012.
- Accepted September 4, 2012.
- © 2013 American Physical Therapy Association
References
Issue highlights
