Physical Activity, Sedentary Behaviour and Sleep, and Their Association with BMI in a Sample of Adolescent Females in New Zealand
Round 1
Reviewer 1 Report
Originality/Novelty:
The novelty of this work is related with the analysis of the activity performed during 24 hours by a total of 119 adolescent females. A description of their physical activity habits is performed and also associated with their BMI.
The objective of the study is well stated. However, this study present a notable limitation regarding the lack of detecting the behavior patterns of participants using an objective method. Therefore, it may have an important bias for detecting the daily behavior of participants in relation to the type of physical activity performed (Strath et al., 2013).
Significance:
Results need to be adjusted and corrected in order to clarify the results of this study, and therefore the conclusions reached by this research work.
In my opinion, there are limitations that should be fixed. On the one hand, data collected is done by using diaries written by the own adolescents (Strath et al., 2013). Besides, the fact that it is not possible to know how long the adolescents wear or not the accelerometer is a relevant limitation of the study. Researchers indicate that wearing at least 10h is considered to be valid for data processing, which is less than 50% of the 24h.
Researchers also suggest that each adolescent should wear the accelerometer at least during three days (a similar period of time than previous studies). However, authors indicate in the manuscript that at least four days should be recorded for indicating MVPA activity more precisely. In the same way, authors indicate the need to take into account if data is collected in a weekday versus a weekend day for comparing data from a population (Ricardo et al., 2020).
As a whole, the principal methodological problem of this study is related with data collection. The use of diaries filled by the adolescents may introduce an important inter-personal an intra-personal bias and thus may alter the conclusions of this research. What is the reliability and validity of the data collection process? Adolescents are who provide the sleep hours, the total hours that they are wearing accelerometer, the circumstances and reasons why they take accelerometers off, etc. Besides, adolescents should report the intensity of the physical activity carried by them when they are not wearing the accelerometers (which is based on a complex methodological guideless that is difficult to understand by non-profesionals) (Strath et al., 2013).
In the ‘Methods’ section, it is not explained if the accelerometer is worn by adolescents during bedtime. In the manuscript, authors cite an algorithm done by Sedah et al (1994), which was developed for adults using accelerometers with an epoch of one minute and wearing them on both wrists. Therefore, this methodology is very different from the one applied in the present work (epoch 15 seconds, accelerometer on the right hip and in adolescent girls). In the work carried out by Migueles et al (2019), a procedure is developed for collecting behavioral data of participants in an objective way by using the Actigraph GT3X accelerometer, avoiding to use the diaries as a key tool for confirming the behavior patterns of adolescents. In the Migueles and cols work, diaries are used as a complementary tool for collecting the physical activity habits of participants.
Quality of Presentation:
Data and results are presented in a structured and appropriate manner. However, supplementary material is needed for better understanding the results (what data is obtained from the diaries, raw data from the accelerometers, classification of the data per each day of the week, etc.). Besides, the methodology applied for implementing the diaries by the adolescent girls should be indicated. Do they receive any indication for filling in the diaries? Is the diary design based on previous studies? Is it previously validated? A model of the diary should be added as supplementary material.
Scientific Soundness:
The study has been carried out with tools or materials already used in previous researches.
However, more than 50% of the data used for reaching the conclusions of this work are obtained from questionnaires and are opinions of the adolescents, being perhaps unreliable. The procedure of data collection through the diaries may be validated. The procedure followed for obtaining data through questionnaires, as well as the instructions done to the adolescent girls for filling the diaries, should be added in supplementary material.
An intra-observer validation has been carried out to obtain the weight and height values of the adolescent girls (lines 85-87), but this procedure has not been applied for validating the way for obtaining information of adolescents’ behavior pattern.
Likewise, it would be desired to use a tool that would have allowed the evaluation of all types of physical activites (including water activities and bedtime) during the 24 hours (Fairclough et al., 2015; Migueles et al., 2017). Similarly, the study should justify why the accelerometer is placed on the hip, especially if the sleep algorithm used is based on a research work that analyses sleep behavior placing accelerometers on the wrists.
The statistical methods used are not described in the work. It is not indicated the statistical method used for comparing categorical variables, nor the type of regression carried out for correlating quantitative variables or the normality test (and its result) performed (Lines 119-131).
Interest to the Readers:
This work brings the important novelty of being able to structure the behavior patterns of adolescent females during 24 hours. Conclusions may provide a valuable information for knowing the behavior of adolescent girls, and for leading the authorities to develop specific actions in order to improve the health of adolescents (in particular) and society (in general).
However, these conclusions are based on subjective (and apparently biased) data. This work should be supported with questionnaires to confirm/validate data obtained by using an accelerometer (an objective tool), but diaries cannot be used as a principal source of data.
Overall Merit:
Several points should be addressed before to publish this work. Changes should be mainly oriented to justify in detail the validation and reliability of the data obtained via the diaries in order to complete the activity-related data collected by using accelerometers.
Special comments:
Line 90: Researchers indicate that accelerometer is placed on the right hip. The selection of this body area should be justified (should be argued in the discussion section). The placement of the accelerometer is a key point that is being investigated. Furthermore, the time of data collection and the frequency and fraction of time in which data is collected are important information that should be included when the device is presented. Technical aspects should be described in detail. For more information, please see the research works done by Migueles et al 2017, Fairclough et al 2015 and Strath et al 2013, among others. The limitations of not being able to collect data during the whole day (24 hours) should also be mentioned.
Line 144: In the results of “breaks in sedentary time”, the manuscript indicates 26 (54,68). It should be revised since 26 is lower than 54. Maybe is a typo error.
In the Discussion section: It is recommended to add a paragraph indicating the difficulty of obtaining valid and reliable data via personal diaries, and discussing the necessity to control the collection of data and the type of information collected.
Lines 217-226: Researchers list the limitations of using accelerometry devices. Knowing the existence of these limitations, several technical decisions should be made prior to the use of accelerometer devices. These considerations should be based on previous studies. For that reason, it is necessary that authors explain them in detail in the ‘Methods’ section and cite references used for decision making in this regard. The cut-off points used are those established by Freedson et al (1998), which use the y-axis as a threshold. There are subsequent publications that have analyzed the cut-off points with triaxial accelerometers at different age gropus, which can be a more up-to-date reference for making methodological decisions (Fairclough et al., 2015; Hibbing et al., 2020; Ricardo et al., 2020)
Additional references:
Fairclough, S.J., Noonan, R., Rowlands, A.V., van Hees, V., Knowles, Z., and Boddy, L.M. (2015). Wear Compliance and Activity in Children Wearing Wrist and Hip-Mounted Accelerometers. Med. Sci. Sports Exerc.
Hibbing, P.R., Bassett, D.R., and Crouter, S.E. (2020). Modifying Accelerometer Cut-Points Affects Criterion Validity in Simulated Free-Living for Adolescents and Adults. Res. Q. Exerc. Sport 1–11.
Migueles, J.H., Cadenas-Sanchez, C., Ekelund, U., Delisle Nyström, C., Mora-Gonzalez, J., Löf, M., Labayen, I., Ruiz, J.R., and Ortega, F.B. (2017). Accelerometer Data Collection and Processing Criteria to Assess Physical Activity and Other Outcomes: A Systematic Review and Practical Considerations. Sports Med. Auckl. NZ 47, 1821–1845.
Migueles, J.H., Cadenas-Sanchez, C., Tudor-Locke, C., Löf, M., Esteban-Cornejo, I., Molina-Garcia, P., Mora-Gonzalez, J., Rodriguez-Ayllon, M., Garcia-Marmol, E., Ekelund, U., et al. (2019). Comparability of published cut-points for the assessment of physical activity: Implications for data harmonization. Scand. J. Med. Sci. Sports 29, 566–574.
Ricardo, L.I.C., Wendt, A., Galliano, L.M., de Andrade Muller, W., Niño Cruz, G.I., Wehrmeister, F., Brage, S., Ekelund, U., and Crochemore M Silva, I. (2020). Number of days required to estimate physical activity constructs objectively measured in different age groups: Findings from three Brazilian (Pelotas) population-based birth cohorts. PloS One 15, e0216017.
Strath, S.J., Kaminsky, L.A., Ainsworth, B.E., Ekelund, U., Freedson, P.S., Gary, R.A., Richardson, C.R., Smith, D.T., Swartz, A.M., and on behalf of the American Heart Association Physical Activity Committee of the Council on Lifestyle and Cardiometabolic Health and Cardiovascular, E. (2013). Guide to the Assessment of Physical Activity: Clinical and Research Applications A Scientific Statement From the American Heart Association. Circulation 01.cir.0000435708.67487.da.
Author Response
Comment 1: The objective of the study is well stated. However, this study present a notable limitation regarding the lack of detecting the behavior patterns of participants using an objective method. Therefore, it may have an important bias for detecting the daily behavior of participants in relation to the type of physical activity performed (Strath et al., 2013).
Response: There seems to have been some misunderstanding about how the data was collected in this study. 24 h activity was determined by the use of Actigraph accelerometers worn on the hip, 24 h a day for seven days. Wear time diaries were only used for participants to record when they removed the accelerometers, if they performed activity when the accelerometer was removed, which only acconted for 0.5% of total wear time (e.g swimming). Wear time diaries were also used to record sleep and wake up times. To clarify this we have added more detail to the abstract (line 15) and the methods (line 99)
Comment 2 Results need to be adjusted and corrected in order to clarify the results of this study, and therefore the conclusions reached by this research work.
Response: please see response to comment 1
Comment 3: In my opinion, there are limitations that should be fixed. On the one hand, data collected is done by using diaries written by the own adolescents (Strath et al., 2013). Besides, the fact that it is not possible to know how long the adolescents wear or not the accelerometer is a relevant limitation of the study. Researchers indicate that wearing at least 10h is considered to be valid for data processing, which is less than 50% of the 24h.
Response: Please see response to comment 1. Additionally, a day was considered valid if wear time during waking hours was ≥ 10 h, if total wear time plus any imputed physical activity was ≥ 20 h AND the day contained ≥ 2 h of sleep. This information has been reworded at line 115 to provide further clarity to the reader.
Comment 4: Researchers also suggest that each adolescent should wear the accelerometer at least during three days (a similar period of time than previous studies). However, authors indicate in the manuscript that at least four days should be recorded for indicating MVPA activity more precisely. In the same way, authors indicate the need to take into account if data is collected in a weekday versus a weekend day for comparing data from a population (Ricardo et al., 2020).
Response: Our apologies, but we cannot find any mention of 4 days being required to indicate MVPA activity more precisely in this manuscript – however we do point out that while we required a minimum of 3 days of valid data (in line with previous studies) the median number of days of valid days in this sample was 6. The number of valid days of data we have for each day of the week is now indicated in the footnote for Figure
Comment 5: As a whole, the principal methodological problem of this study is related with data collection. The use of diaries filled by the adolescents may introduce an important inter-personal an intra-personal bias and thus may alter the conclusions of this research. What is the reliability and validity of the data collection process? Adolescents are who provide the sleep hours, the total hours that they are wearing accelerometer, the circumstances and reasons why they take accelerometers off, etc. Besides, adolescents should report the intensity of the physical activity carried by them when they are not wearing the accelerometers (which is based on a complex methodological guideless that is difficult to understand by non-profesionals) (Strath et al., 2013).
Response: See response to comment 1
Comment 6: In the ‘Methods’ section, it is not explained if the accelerometer is worn by adolescents during bedtime. In the manuscript, authors cite an algorithm done by Sedah et al (1994), which was developed for adults using accelerometers with an epoch of one minute and wearing them on both wrists. Therefore, this methodology is very different from the one applied in the present work (epoch 15 seconds, accelerometer on the right hip and in adolescent girls). In the work carried out by Migueles et al (2019), a procedure is developed for collecting behavioral data of participants in an objective way by using the Actigraph GT3X accelerometer, avoiding to use the diaries as a key tool for confirming the behavior patterns of adolescents. In the Migueles and cols work, diaries are used as a complementary tool for collecting the physical activity habits of participants.
Response: In line 101 of the manuscript states that accelerometers where worn 24 h a day for 7 days. The Sadeh algorithm has been shown to perform well in children when worn at the hip, particularly when constrained by bedtimes (see Kinder et al 2012). A statement to this effect has been added to the methods at line 125 The data was collapsed into 1 min epochs prior to applying the Sadeh alogrythm to the data. A statement to clarifying this has been added at line 121. Diaries in this study were used as a complementary tool in a very similar manner as described by Migueles et al.
Comment 7: However, more than 50% of the data used for reaching the conclusions of this work are obtained from questionnaires and are opinions of the adolescents, being perhaps unreliable. The procedure of data collection through the diaries may be validated. The procedure followed for obtaining data through questionnaires, as well as the instructions done to the adolescent girls for filling the diaries, should be added in supplementary material.
Response: Please see response to Comment 1
Comment 8: An intra-observer validation has been carried out to obtain the weight and height values of the adolescent girls (lines 85-87), but this procedure has not been applied for validating the way for obtaining information of adolescents’ behavior pattern.
Response: Please see response to comment 1
Comment 9: Likewise, it would be desired to use a tool that would have allowed the evaluation of all types of physical activites (including water activities and bedtime) during the 24 hours (Fairclough et al., 2015; Migueles et al., 2017). Similarly, the study should justify why the accelerometer is placed on the hip, especially if the sleep algorithm used is based on a research work that analyses sleep behavior placing accelerometers on the wrists.
Response: Please see response to Comments 1 and 6
Comment 10:The statistical methods used are not described in the work. It is not indicated the statistical method used for comparing categorical variables, nor the type of regression carried out for correlating quantitative variables or the normality test (and its result) performed (Lines 119-131).
Response: As described at line 141 mixed model regression was used to determine the differences in time spend in each component of the day by BMI z-score or by overweight or obese catergory. A statement indicating that the assumptions of the model (these includes normality and homoscedasticity of the residuals) were checked and found to be met was been added at line 143.
Comment 11:This work brings the important novelty of being able to structure the behavior patterns of adolescent females during 24 hours. Conclusions may provide a valuable information for knowing the behavior of adolescent girls, and for leading the authorities to develop specific actions in order to improve the health of adolescents (in particular) and society (in general).
However, these conclusions are based on subjective (and apparently biased) data. This work should be supported with questionnaires to confirm/validate data obtained by using an accelerometer (an objective tool), but diaries cannot be used as a principal source of data.
Response: See response to Comment 1.
Comment 12 Several points should be addressed before to publish this work. Changes should be mainly oriented to justify in detail the validation and reliability of the data obtained via the diaries in order to complete the activity-related data collected by using accelerometers.
Response: See response to comment 1
Comments 13: Line 90: Researchers indicate that accelerometer is placed on the right hip. The selection of this body area should be justified (should be argued in the discussion section). The placement of the accelerometer is a key point that is being investigated. Furthermore, the time of data collection and the frequency and fraction of time in which data is collected are important information that should be included when the device is presented. Technical aspects should be described in detail. For more information, please see the research works done by Migueles et al 2017, Fairclough et al 2015 and Strath et al 2013, among others. The limitations of not being able to collect data during the whole day (24 hours) should also be mentioned.
Response: Justifcation of placement at the hip has been added to the discussion at line 245. Sampling frequency and fitler use is now reported at line 100 epoch length in reported at line 120 and 121. Acceloerometers were worn for 24 h a day – this is described at line 101.
Comment 14: Line 144: In the results of “breaks in sedentary time”, the manuscript indicates 26 (54,68). It should be revised since 26 is lower than 54. Maybe is a typo error.
Response: Thank you for drawing this typo to out correction –this has now been appended to the correct number of breaks in sedentary time (n=62).
Comment 15 In the Discussion section: It is recommended to add a paragraph indicating the difficulty of obtaining valid and reliable data via personal diaries, and discussing the necessity to control the collection of data and the type of information collected.
Response: Only 0.5% of data was imputed from diaries so we don’t feel this warrants a paragraph in the discussion
Comment 16: Lines 217-226: Researchers list the limitations of using accelerometry devices. Knowing the existence of these limitations, several technical decisions should be made prior to the use of accelerometer devices. These considerations should be based on previous studies. For that reason, it is necessary that authors explain them in detail in the ‘Methods’ section and cite references used for decision making in this regard. The cut-off points used are those established by Freedson et al (1998), which use the y-axis as a threshold. There are subsequent publications that have analyzed the cut-off points with triaxial accelerometers at different age gropus, which can be a more up-to-date reference for making methodological decisions (Fairclough et al., 2015; Hibbing et al., 2020; Ricardo et al., 2020)
Response: see the response to comment 14
Reviewer 2 Report
Dear Authors
thank you for lettnig me review this article. It is interesting both as a research problem and as a paper and I am sure it can be a valuable contribution to the field of sport science.
Although the study is well-described (Methodology, procedure, results) and all necessery details are provided (for example if one wanted to duplicate the study), however, after reading this paper I had an impression of very narrow-approach to the ...frankly saying global problem.
The only methodological concern I has was inclusion of the participants with 3 day valid (fully recorded) measurements while in the Results section you provide results of the 7-days avarage.
but generally,
my suggestions would be - enhance the Intro - and place the issue you are going to study in a broader picture, for example looking into what is the "assessment of epidemiological obesity among youth and adults" where situation of various countries and continents of the world is presented, and show whether New Zeland's situation is any different?
Also, among published studies concerning measurment of daily PA of young adolescents there have been some showing the different patters of such behaviours in week and weekend days in different areas (look for example in "Patterns of daily Physical Activity during school days in children and adolescents or in "Mothers reframing physical activity: Family oriented politicism, transgression and contested expertise in Australia' to put some more light on why there is such a situation with young adolescents, specifically women. That would strenghten the rationale for the study.
In my opinion also a Discussion could be enhanced a little by discussing some other factors like culture/religion? or quality of PE? as maybe young adolescents do not need for example PA-related goals but efficency? (look into "Maybe it is not a goal that matters....')
I believe these enhancement would help the reader to related the findings of your study to their national situation and thus make the study message more universal.
Author Response
Thank you for your thoughtful comments on our manuscript please see below for our point by point response:
Comment 1: The only methodological concern I has was inclusion of the participants with 3 day valid (fully recorded) measurements while in the Results section you provide results of the 7-days avarage.
Response: the numbers of participants that had valid data for each of the days of the week has been added to Figure 2.
Comment 2: my suggestions would be - enhance the Intro - and place the issue you are going to study in a broader picture, for example looking into what is the "assessment of epidemiological obesity among youth and adults" where situation of various countries and continents of the world is presented, and show whether New Zeland's situation is any different?
Response: Thank you for this suggestion. Obesity rates in NZ and how they compare to other countries have now been added to the introduction at line 46
Comment 3: Also, among published studies concerning measurement of daily PA of young adolescents there have been some showing the different patters of such behaviours in week and weekend days in different areas (look for example in "Patterns of daily Physical Activity during school days in children and adolescents or in "Mothers reframing physical activity: Family oriented politicism, transgression and contested expertise in Australia' to put some more light on why there is such a situation with young adolescents, specifically women. That would strenghten the rationale for the study.
Response: thank you for this suggestion, a comment, highlighting our lack of understanding of how 24 h activity patterns may vary across the week has been added to the introduction at line 57.
Comment 4: In my opinion also a Discussion could be enhanced a little by discussing some other factors like culture/religion? or quality of PE? as maybe young adolescents do not need for example PA-related goals but efficency? (look into "Maybe it is not a goal that matters....')
Response: A paragraph discussing participation in PE and organised sport has been added to the discussion at line 232
Reviewer 3 Report
This a very well written study to assess 24-hour movement activity in NZ adolescent females. Interestingly LPA was positively associated with BMI and overweight/obese had worse sleep and MVPA.
The authors should be commended for this very well written and easy to follow paper. I only have a couple of minor comments:
1-Figure 1 – What is the ‘100 deleted due to low amounts’ referring too?
2-Table 2 results: is it possible to denote the significant results to make it easier to read? I also wonder whether the lack of significance with BMI was a result of very large standard deviations and skewed accelerometer data. Did the authors perform any normality tests, or try to log-transform the data? In my experience, accelerometer data is often very skewed.
3-Discussion-at the start of the discussion please remove ‘nationwide’ as this is not true of the relatively small sample included. It’s an overstatement.
Author Response
Thank you for your thoughtful comments on our manuscript. Please see below for a point by point response
Comment 1:-Figure 1 – What is the ‘100 deleted due to low amounts’ referring too?
Response: Apologies that this wasn’t clear. This statement has now been modified to read n=100 didn’t receive an accelerometer as a decide was not available in their location at the time of testing
Comment 2: Table 2 results: is it possible to denote the significant results to make it easier to read? I also wonder whether the lack of significance with BMI was a result of very large standard deviations and skewed accelerometer data. Did the authors perform any normality tests, or try to log-transform the data? In my experience, accelerometer data is often very skewed.
Response: Significant results have now been bolded in the table. A statement clarifying that the assumption of the regression models were tested and found to be met has been added to the statistical analysis section at line 143
Comment 3: Discussion-at the start of the discussion please remove ‘nationwide’ as this is not true of the relatively small sample included. It’s an overstatement.
Response: The word nationwide has been removed from line 186
Round 2
Reviewer 1 Report
All suggestions had been respond.
In my opinion, a paragraph in the discussion section about the lmitations to measure all of 24h with accelerometer and the interpretation of the information obtenined from diaries could do better the document.
There are limitations about the problems with the interpreted of measures from accelerometer (lines 260-270), so would be better if the authors indicate the limitations to measure the physical activity during 24h with the accelerometer actigraph on the hip too.
