International Study of 24-hour Movement Behaviors of Early Years (SUNRISE): A Pilot Study from Bangladesh

Mohammad Sorowar Hossain (  sorowar.hossain@brfbd.org ) Biomedical Research Foundation https://orcid.org/0000-0001-7143-2909 Iztiba M. Deeba Biomedical Research Foundation, Dhaka Mahmudul Hasan Biomedical Research Foundation, DHaka Katharina E. Kariippanon University of Wollongong Faculty of the Arts Social Sciences and Humanities Kar Hau Chong University of Wollongong Faculty of the Arts Social Sciences and Humanities Penny L. Cross University of Wollongong Faculty of the Arts Social Sciences and Humanities Shameema Ferdous Biomedical Research Foundation, Dhaka Anthony D. Okely University of Wollongong Faculty of the Arts Social Sciences and Humanities


Introduction
Childhood obesity poses an urgent public health challenge in the 21st century and is rising at an alarming pace in both high income and low-and middle-income countries (LMICs). In 2019, around 38 million children (age < 5 years) were reported as having overweight or obesity and according to a recent projection, 45% (approximately 4 billion) of the global population will experience obesity by 2035 [1][2][3].
Not only are immediate health and quality of life affected, but children with obesity are also at greater risk of developing several non-communicable diseases such as diabetes, cardiovascular diseases, hypertension, stroke, and cancer [2,4]. Obesity at a young age tends to persist into adolescence and adulthood and can give rise to chronic illness and premature death [3].
In Bangladesh, a lower middle-income country, childhood overweight and obesity have increased over the years, from 3.6% in 1998-2003 to 7.9% in 2010-2015. [4]Currently the numbers of children with overweight and obesity are overwhelmingly higher in low-and middle-income countries than in highincome countries [5,6]. In our recent study, we found around 14% children (4-7 years old) were overweight/obese at urban community setting in Bangladesh [7] . Movement behaviors such as low level of physical activity, prolonged sedentary activities and short duration of sleep are well established factors contributing to the development of overweight and obesity in childhood [8]. Though several intervention strategies have been implemented to overcome the obesity epidemic, they typically focused on individual movement behavior while excluding the others [9]. Since these strategies fail to recognize the health implications of the movement behaviours in a holistic way, the importance of developing an integrated approach including guidelines for all the movement behaviors was deemed to be of supreme importance. Thus, in 2019, the World Health Organization (WHO) released Global guidelines for physical activity, sedentary behavior, and sleep for children under ve years of age [10].
The primary goal of The International Study of Movement Behaviors in the Early Years (SUNRISEhttps://sunrise-study.com/) study is to assess the proportion of children adhering to the WHO Global guidelines from countries of different socioeconomic status. In addition to this, it aims to determine how 24-hour movement behaviours are associated with overweight and obesity, gross and ne motor skills and executive function in the early years and the variations among the countries [11]. Prior to conducting the study on a large scale, it was essential to assess the feasibility of the protocol in each participating country. Thus, the SUNRISE pilot study was designed and piloted in Bangladesh, a densely populated LMIC, in March 2019. In this paper we report the results of the Bangladesh SUNRISE pilot study.
The coronavirus (COVID-19) pandemic created an unprecedented situation impacting on every sphere of personal, social and economic life. To stop the transmission of the SARS-CoV-2, the causative of COVID-19, countries around the world implemented various public health measures (e.g lockdowns, school closures and movement restrictions). Bangladesh imposed a nationwide lockdown during the initial phase (23 March 2020 to 30 May 2020.) of the COVID-19 pandemic which caused widespread panic [12]. The changed environment of the COVID-19 pandemic created a unique opportunity to re-assess how the movement behaviours among children enrolled in the SUNRISE study had changed due to COVID-19.
The objectives of this study were to: (i) determine the proportion of preschool children (aged 3 and 4 years) in Bangladesh who met the WHO guidelines for physical activity, sedentary behaviour and sleep, (ii) examine the feasibility of the methods to be used in the SUNRISE main study and (iii) assess the impact of COVID-19 pandemic in the movement behaviors of Bangladeshi preschool children.

Study setting and participants
The SUNRISE pilot study was conducted in three schools in the greater metropolitan area of Dhaka: two schools were in an urban area (Dhaka Premier school and Akij Foundation School) and one (AkijBonoful Foundation School) in a rural-like setting (less-developed part of Dhaka). The principal of each school was contacted, informed and permission was granted for the study. Parents of eligible children were contacted through school class teachers to participate in the study. Once parents provided informed written consent, children aged between 3-5 years took part in the study. The ethical approval (Memo No: BRF/ERB/2019/03) of this study was obtained from the Institutional Review Board of Biomedical Research Foundation, Bangladesh and the University of Wollongong Human Ethics Research Committee.
The study was conducted between March 2019 and June 2019.On the day of data collection, child measures were completed in a single day. An interviewer-administered parent questionnaire (in local language) providing basic demographic information for both parent/caregiver and child and the movement behaviors of their child was completed on the same day. As a follow up study, parents who had participated in the SUNRISE pilot study were contacted again over the phone between 20 May2020 and 30 May 2020, during the COVID-19 pandemic to collect data on children's movement behaviours during the pandemic.

Anthropometrics
The height of the children was measured using a portable stadiometer (Leicester 214 Transportable Stadiometer; Seca, Germany) and weight was measured barefoot using a digital scale (SECA 878). All measurements were taken twice and average was used for analysis. Body mass index (BMI) and associated z-scores for BMI-for age (BAZ), height-for-age (HAZ) and weight-for-age (WAZ) were computed using the WHO AnthroPlus software (version 1.0.4; WHO, Geneva, Switzerland).Children were classi ed according to the reference standards of the WHO. [13,14] Accelerometry To assess the physical activity, sedentary time and sleep patterns of children,two types of accelerometers wereused: the ActiGraph wGT3x-BT was worn over the right hip and the activPAL 4 was attached on the right thigh.The children were askedto wear the accelerometers for four days. They could remove the ActiGraph only while bathing or swimming thus preserving the monitor from water.
The ActiLife software (version 6.1.2.1, ActiGraph Corporation) was used to initialize, download, and process the ActiGraph data. Accelerometers were programmed to capture data at a sampling rate of 30 Hz, which was integrated into 15-second epochs for the analysis using the low frequency lter. To be included in the analyses, a child needed to have at least 24 hours of valid data, with a minimum of 6 hours of valid wear time during waking hours. The identi cation of valid 24hr day(s) was done through visual inspection of the ActiGraph data to con rm if there were movement peaks throughout the entire day (de ned as midnight to midnight). If there were little to no peaks throughout the sleeping period, it was assumed that the child took off the monitor during the particular day(s) and was not included in the subsequent analyses. A pre-determined time lter (i.e., 7:20AM to 21:10PM) (based on the average parentreported wake-up and bedtime of the study population) was applied to all valid 24hr day(s) to exclude sleeping period from the analyses of PA and SB. Further, any waking periods with = > 20 consecutive minutes of zero counts were de ned as non-wear and were excluded from the analysis [15]. The cutpoints by were used to classify valid waking wear time as SB (< 800 counts per minute [CPM]), LPA (800-1679 CPM) or MVPA ( = > 1680 CPM) [16,17]. ActivPAL data were downloaded and analyzed using PALconnect and PALanalysissoftware v7. The feasibility of the activPAL was assessed, but only actiGraph data is presented, due to skin issues with the activPAL.

Sleep, screen-time and restrained sitting
Caregivers were asked to report their child's nighttime sleep and nap schedules. Sleep duration was calculated as the sum of nighttime sleep (difference between bedtime and wake-up time) and nap duration. Caregivers also reported time spent outside; screen time and use of screens before bedtime; time spent restrained (strapped in and unable to move) and time spent sitting.

Executive functions
Executive function tests were performed to assess the cognitive exibility (shifting), control of behavioral urges and impulses (inhibition) and visual-spatial working memory of children [18]. Three personalized iPad games (www.eytoolbox.com.au) including "Card Sort", the "Go/No-Go" and the "Mr. Ant".
[18]It took approximately 5 minutes to complete the "Card Sort"game and the total scores ranged between 0-12 points.
[18] This game measured the ability to exibly shift attention."Go/No-Go" was designed to assess inhibitory control. On the iPad screen, a sh or a shark appeared randomly and the children needed to tap the sh but not the shark before it quickly passed. The scores for this game ranged between 0 and 1, with1 representing the highest score.
[18]The score was produced by calculating the proportion of shes caught multiplied by the proportion of sharks spared by the children."Mr. Ant" assessed the child's working memory.In this game an ant appears on screen with a number of colored dots placed in different spatial locations on its body. The dots vanish after some time, and the child needs to recall and tap the locations of the dots. The score was provided after the completion of each level and ranged between 0-8 points. [18] Gross and ne motor skills The Ages and Stages Questionnaire-3 (48 months, ASQ3) was used to assess gross and ne motor skills. [19] Participating children got the opportunity to attempt tasks twice and their performance was categorized as "yes", "sometimes", or "not yet". If children were able to perform the task on a single attempt, they received 'yes'. If they initially failed, they got a second attempt and received a "sometimes" if successful or "Not yet" if they failed both attempts. For ne motor skills assessment, children had one chance to complete the tasks and were classi ed as either "yes" or "not yet" and scored according to the ASQ-3 tool. For each ASQ task the child was classi ed as i) child requires follow-up and further assessment/action; ii) child is developing on schedule but may bene t from extra practice in some of the skills; iii) child is developing on schedule. [19] COVID-19 questionnaire: The COVID-19 questionnaire consisted of 25 questions relating to the children's physical activity, sedentary behavior, sedentary screen time, and sleep at the height of COVID-19 restrictions in Bangladesh. The questions regarding time spent in various movement behaviors and time spent outdoors were identical to the previous questionnaire.

Data analysis
All statistical analyses were performed using SPSS Statistics for Window version 26.0 (IBM Corp, Armonk, NY). Descriptive statistics (mean and standard deviation or frequency and percentage) were computed for all study variables. Mann-Whitney U tests were conducted to examine differences in anthropometric indicators, movement behaviours, motor skills and executive function between the sexes and residential settings. Chi-square tests or Fisher-Exact tests were used to examine differences in the proportion of children meeting the 24-hour movement guidelines between the sexes. Wilcoxon-signed rank tests were used to examine changes in movement behaviours before and during COVID-19 period.
Statistical signi cance was set at p < 0.05.

Results
A total of 65 preschoolers participated with 48 from urban and 19 from less-developed areas. The mean age of 4.5 (± 0.5) years and 55.4 % of children were male. Table 1 shows anthropometric outcomes and the amount of time spent in the different movement behaviors, strati ed by sex and urban/ less-developed location. No signi cant differences were found between girls and boys for any of the anthropometric indicators. However, children from less-developed setting had signi cantly lower BMI (14.3 ± 1.0 vs. 15.8 ± 1.6) and BAZ (-0.69 ± 1.33 vs. 0.03 ± 1.17) scores than their urban counterparts. No signi cant differences between sexes were found for the amount of time spent in any of the movement behaviours. The proportions of children meeting the WHO guidelines are shown in Table 2. Only three children (4.7%) met the all components of the guidelines. The proportion of children who met the physical activity, sedentary screen time and sleep guidelines were 71.9% (n = 41), 17.5% (n = 11), and 59.7% (n = 37) respectively. There were no statistical differences between boys and girls in terms of meeting and not meeting guidelines individually or combined. e Differences between sexes were tested using Pearson chi-square tests or Fisher-Exact tests; N/A no statistical test was conducted due to the zero proportion identi ed for girls.
The results of the gross and ne motor skills tasks and executive functions were presented in Table 3. Overall, no difference between boys and girls or between children of urban and less-developed areas was observed for any motor skill tasks. The proportion of all children who were developmentally on-track for the gross and ne motor skills were 58.7% and 50.8%, respectively. No signi cant differences were observed between boys and girls and for any of the executive function tasks, however, children from lessdeveloped setting had signi cantly higher working memory scores than their urban counterparts (2.26 ± 0.86 vs. 1.58 ± 0.91). 74.4%) and sleep (n = 20; 51.3%), and nearly half felt they could not support their child to accrue a healthy balance of movement behaviors (n = 18; 46.2%). Table 4 presents parent-reported physical activity, time spent outside, and screen time pre-COVID-19 and during the pandemic. Most importantly, during the COVID-19 pandemic there was a signi cant decrease in Bangladeshi children's total physical activity (-193 min/day), and time spent outside on weekdays (-75 min/day) and weekend days (-131 min/day).
On the other hand, sedentary screen time of children increased signi cantly (+ 85 min/day) during the COVID-19 pandemic. Analysis of the protocol feasibility revealed the following regarding compliance with the 24hour accelerometer protocol: 89% (n = 57) of children wore the monitors for a minimum of one day (i.e. provided at least 24-h data; 47 [82.5%] provided 3 days of 24-h data); others (n = 7, 11%) removed the monitors during the night sleep time and/or did not wear the monitor for a signi cant portion of waking time. On the other hand, 37 of children (57.8%) also wore the activPAL devices simultaneously. However, 37.8% of them (14 children) developed itchy skin/rashes after wearing the devices. As a result, it was determined to suspend the collection of movement behaviour data using the activPAL monitor mid-way through the data collection period. Participation in the executive function tasks: 51 (79.7%) completed the inhibition task; 60 (93.7%) completed the working memory task; and 61 (95.3%) completed the shifting task.

Discussion
To the best of our knowledge, this was the rst study of its kind to investigate the 24-hour movement behaviors quantitatively irrespective of age groups of children in Bangladesh. Overall a low proportion of preschool children met the WHO Global guidelines. We found signi cant differences in movement behaviors and working memory between urban and children from less-developed setting. During the nationwide lockdown due to the COVID-19 pandemic physical activity was severely reduced, while sedentary screen time among Bangladeshi preschoolers. Implementation of the SUNRISE protocol were found to be feasible in Bangladesh with the exception of the activPAL.
Variations in time spent on physical activity guideline could be associated with methodical differences in the studies. For Bangladeshi children, both ActiGraph and ActivPAL acceleromoters were used while other studies used ActiGraph.
In Bangladesh, the proportion of children meeting the sedentary screen time recommendation was found to be 17.5%. Our result is similar to Australia but lower than Sweden (37.8%) and South Africa (48%). A lower proportion (59.7%) of Bangladeshi children met sleep recommendation as compared children in Sweden (62.5%), Canada (83.9%), Australia (88.7%) and South Africa (66%), while only 29.5% of children met the sleep recommendation [15,20,21,23].
The lower physical activity and higher screen time in Bangladesh could be attributed to lack of adequate playgrounds in schools and neighborhoods in Dhaka, which is a highly crowded (> 47,000/square km) megacity, with a population in excess of 20 million [24]. Bangladesh is undergoing rapid and unplanned urbanization, with playgrounds for children and greenspaces for outdoor play largely ignored in urban planning. Unlike in many HICs, most families in Dhaka city live in small apartments where children share abed with their parents or siblings. As expected, screen devices were present in majority of bedrooms of the children and most of them used screens just before sleep.

Feasibility of SUNRISE protocol in Bangladesh context
Considering the feasibility and willingness of children to wear the devices, the ActiGraph yielded better results compared to the activPAL. A high compliance for the ActiGraph (only 7 children refusing to wear the monitor during the sleep) and a low compliance for activPAL was observed (27 children refused to wear altogether). Further, since 37.8% (14) children developed itchy skin and rashes after wearing the activPAL, this device seems unsuitable for use in countries with hot and humid climates, such as Bangladesh. Thus, we recommend that the ActiGraph is used in the SUNRISE main study to increase compliance in hot and humid countries.
We found that the gross and ne motor skill assessments were feasible and acceptable in the Bangladesh context. Bangladeshi children performed poorly in gross and ne motor skills tasks (46.7 ± 13.1 and 40.0 ± 17.2 points) compared to the children from developed countries like Norway (54 ± 9 and 50 ± 13 points) and America (52 ± 10 and 44 ± 14 points). [25] Here, the gap between the scores might be due to the different methodology used in those studies. In the Norwegian and American studies, the gross and ne motor skills were reported by parents unlike the SUNRISE study where trained research personnel assessed the tasks. On the other hand, Bangladeshi children were quite ahead in points from the Swedish children (41.7 ± 15.6 and 35.5 ± 17.9 points) in terms of both gross and ne motor skills [21]. In Bangladesh, children from the less-developed area -demonstrated greater gross and ne motor skills which were signi cantly higher than the urban children-happened to engage more in physical activity in open spaces available there. It is well established that children's cognitive development is associated with ful lling the minimum requirement of physical activity and ndings from Bangladesh SUNRISE pilot study supported this theory [26,27].
The executive function protocol was feasible among the children. While assessing the executive functions, children started with "Mr. Ant" and then completed the "Go/No-Go". The "Card Sort" task was the last to be tested with a su cient break as most children could not hold the concentration for long.
Children from less-developed area played the "Card Sort" game in a separate day with a fresh start. Children irrespective of area they were from performed the best in the Mr. Ant task and the least well in the "Card Sort" task.

Impact of COVID-19 pandemic on 24-hour movement activities in Bangladesh
Pandemics such asCOVID-19 can in uence the movement behaviours signi cantly, due to the restrictions placed on movement and the closure of preschools. In our study, children's physical activity was reduced more than threefold, while screen time increased signi cantly. The opposite results for physical activity were observed in Swedish children during COVID-19 pandemic (ref). This was because of differences in public health policies set by the two countries. In Bangladesh, lockdown was strictly maintained during the early phase of the pandemic and people were so panicked that most people in Dhaka city were stayed home. Unlike the situation in many high-income countries, there is no backyards or parks in Dhaka city and therefore there was limited scope for physical activity for children in Bangladesh. In future potential infectious disease outbreaks, the issue of physical activity needs to be addressed by the policymakers of the country.

Strengths and limitations
The device-based measurement of 24-hour movement behaviours (physical activity, sleep), gross and ne motor skills and executive functions of individual preschool aged children in LMIC setting like Bangladesh is the major strength of this SUNRISE pilot study. On the other hand, convenient sampling and the small sample size are the major limitations of this study and therefore results from this study would not be generalizable to Bangladesh [28].

Conclusion
This SUNRISE Bangladesh pilot study provided initial ndings of 24-hour movement behaviors of preschool aged children in Bangladesh. Only 4.7% of Bangladeshi preschool aged children met the WHO Global guidelines. A low proportion of children met physical activity and screen time guidelines. During COVID-19 pandemic, the children's physical activity and time spent outside decreased but screen time was increased signi cantly in Bangladesh. Future studies with a representative sample size are needed to con rm these results and investigate associated factors. As the methods and devices used in this pilot study were proved to be feasible in Bangladesh setting, it has paved the way to conduct the main SUNRISE study in Bangladesh to depict the real scenario of children meeting WHO Global guidelines.

Declarations
Ethics approval and consent to participate