Movement Behaviors during COVID-19. A Survey among Children under Five Years of Latin American Origin or Descent from Chile, Mexico, and the US

Movement behaviors (physical activity, sedentary behavior (including screen time), and sleep) have been impacted by the COVID-19 pandemic. We aimed to report changes in and factors inuencing movement behaviors during COVID-19 among Latin American/Latino children aged 1 to 5 years in Chile, Mexico, and the USA. We conducted a cross-sectional study between April and August 2020. Caregivers reported changes in movement behaviors and provided information about family and household characteristics. In total, 4,136 children (mean age [SD], 3.1 [1.4] years; 51% boys). The proportion of children who met the WHO Guidelines decreased signicantly in all countries, with large declines in meeting the physical activity and screen time guidelines. Factors associated with changes were being an older child, unable to attend an early childhood education and care service, higher parental education levels, not having the opportunity to play with someone, and not having access to spaces to play. During COVID-19, Latino parents reported changes in physical activity, screen time, and sleep quality among their toddlers and preschoolers. The ndings highlight the need to minimize disparities faced by families by providing access to early childhood education and care and safe places for children to play. Our study conrms strong interrelationships among changes in PA, ST and sleep quality among preschool Latin American/Latino children from Chile, Mexico and the USA. 7,12,13 Several child, caregiver and household environment characteristics were consistently associated with unhealthy changes in all movement behaviors and across sites. For the child, these included being older, not being able to attend an ECEC. Higher parental education levels, not having the opportunity to play with someone, and not having access to spaces to play were important caregiver and household environment factors that negatively impacted movement behaviors among Latin American/Latino preschool children during the pandemic. .


Introduction
The COVID-19 pandemic has changed how children, families, and communities conduct their daily lives. Latin America and the USA are an epicenter for the pandemic due to many factors, including widespread poverty and socioeconomic inequalities in the region. Despite comprising only 18% of the US population, Latino populations represent 30% of COVID-19 cases. 1,2 Similarly, Latin America has accounted for around 30% of the total deaths globally due to COVID-19. 3 Consequently, Latin Americans/Latinos face a disproportionate burden that is likely to last well beyond the post-vaccination period.
In 2019 the World Health Organization (WHO) released the rst global guidelines for physical activity (PA), sedentary behavior, and sleep for children under ve. 4 These guidelines acknowledge the interrelationships between these three movement behaviors and reinforce that "the whole day matters" when promoting healthy levels of these behaviors. A recent systematic review found that only 10% of toddlers and 15% of preschoolers met the global guidelines for PA, sedentary behavior, and sleep. 5 Studies have documented the impact of the pandemic on movement behaviors in adults 6 and school-age children and youth 7 in North America, but there is little empirical evidence of the impact on children under the age of 5, especially among Latin America/Latino populations. The closure of early childhood education and care services During both periods, sleep quality was assessed with the question "How would you rate your child's sleep quality?", with Likert scale response options (1 to 7, a higher score indicates better quality). Children with a sleep quality ≥4 were considered as with good sleep quality. We calculated the proportion of children with good sleep quality in both periods and the difference in sleep quality between them.

Statistical analysis
On each study site, means and standard deviations, and frequencies and percentages were used to describe the samples. For each country, total time in movement behaviors and movement behavior guidelines compliance before and during COVID-19 were compared using t-tests and proportion tests, respectively. Multiple linear regressions were used to assess the association between factors and changes in movement behaviors for each country, adjusting by all the factors. Logistic regressions were used to explore the association between factors and changes in movement behavior compliance (supplementary le). All data preparation and analysis were conducted with Stata 15.0 (College Station, TX: StataCorp LLC). The level of signi cance was p<0.05, two-tailed.

Results
In total, 4,136 participants were included in the nal analysis, 3,045 in Chile, 632 in Mexico, and 459 in the USA. Around half the sample was female, and the average age was 3.1 (1.4) years. A larger proportion of caregivers in Chile were younger. In Mexico, a larger proportion of caregivers had a university degree, households were bigger and from a high-income level. In the USA, fewer Latino children were enrolled in ECEC, more lived in apartments, from lower-income and located in rural areas (see Table 1).  b Measured with a Likert scale where "1" indicates di culty to fall asleep, waking up several times throughout the night for long period and being restless (coughing and turning, moving and kicking the bedclothes off), whereas "7" indicates falling asleep easily and within few minutes, sleeping well through the night, breathing normally and sleeping deeply. Those answering 4-7 were classi ed as with "good sleep".

Factors associated with changes in physical activity
In all countries, PA decreases were consistently associated with reductions in sleep quality, inability to play with someone else, and not having available space to play (Table 3). In Mexico and Chile, PA decreases were also associated with increases in ST, older children, living in an apartment or condominium, and not having an available backyard. In Chile and USA Latinos, decreases in this outcome were associated with having more educated caregivers. As for country-speci c results, in Chile, decreases in PA were associated with being girl, being enrolled in an ECEC, having older caregivers, living with fewer children, and living in urban areas. In Mexico, PA reductions were more likely in children living in non-traditional residential structures (e.g., huts, motor homes, ranches) and those living with fewer adults at home.
Changes in other movement behaviors Physical activity (min/day) Factors associated with changes in screen-time Across countries, ST increases were associated with reductions in sleep quality, being older, having great access to electronic devices, no parental-imposed restrictions in using such devices, and no available space to play (Table 3).
In Chile and Mexico, increases in ST were also associated with reductions in PA, being older, being previously enrolled in ECEC, unable to play with someone else, and living in an apartment or condominium. In Chile, high parental education and residing in non-traditional homes (e.g., huts, motor homes, ranches) was associated with increased ST during COVID-19.

Factors associated with sleep duration and quality
Across sites, decreases in sleep duration were more likely among those living in apartments or condominiums (Table 4). In Mexico and Chile, decreases in sleep duration were also associated with reductions in sleep quality. In Chile, older children, those who were not previously enrolled in ECEC, and more educated caregivers had higher decreases in sleep duration. Being a boy was associated with decline in sleep duration in Mexico.    In terms of sleep quality, increases in ST and decreases in PA were associated with decreases in this outcome across all sites (Table 4). In Chile and Mexico, increases in sleep quality were also associated with increases in sleep time and having available space to play at home. In Chile alone, male children and those in high SES families had better sleep quality. Meanwhile, in the USA, having electronic devices in the child's room was associated with worse sleep quality during the pandemic.

Discussion
Our study con rms strong interrelationships among changes in PA, ST and sleep quality among preschool Latin American/Latino children from Chile, Mexico and the USA. 7,12,13 Several child, caregiver and household environment characteristics were consistently associated with unhealthy changes in all movement behaviors and across sites.
For the child, these included being older, not being able to attend an ECEC. Higher parental education levels, not having the opportunity to play with someone, and not having access to spaces to play were important caregiver and household environment factors that negatively impacted movement behaviors among Latin American/Latino preschool children during the pandemic.
Some of our results are consistent with reports from other countries focused on children (versus preschool children). This includes a decline in PA 7,12 and an increase in ST during 13 This is likely due to children not being able to go outdoors, not attending their ECEC in-person (and often, switching to online educational activities), 14 and having increased access to electronic media devices during COVID-19. It has been shown that children are more active and less sedentary the more time they spend outdoors, [15][16][17] and are more active at ECEC than at home 18 . Outdoor play is also important in maximizing sleep quality, which can be explained by the effect of sunlight exposure on sleep and circadian rhythms. 19 In our study, increases in ST and decreases in PA were much greater among preschoolers than toddlers, which is likely because children aged 3-5 had greater changes to their daily routines given the closure of ECEC services, and a greater proportion of them are enrolled than toddlers. Parents working from home may also be allowing their children greater access to electronic media devices to keep them from distracting them while the parent works. This could help explain why higher parental education was associated with more negative changes in movement behaviors in our study, as higher educated adults were most likely to have jobs allowing them to work from home during  This hypothesis warrants further exploration.
We found strong and consistent relationships across the three countries among PA, ST and sleep quality, but not sleep duration. As expected, PA increases were associated with positive changes in sleep quality and decreases in ST and vice versa. [21][22][23] This reinforces that these behaviors are co-dependent and should be promoted in terms of what is a healthy movement pattern over 24 hours, as suggested by the WHO global guidelines. 4 The null association for sleep duration with changes in PA and ST can be explained by the lack of change in sleep duration, unlike the difference in sleep quality, which was more profound and likely due to children going to bed and waking up later, a pattern of sleeping associated with poorer health outcomes. 24 Although most of our ndings are not unexpected, documenting the adverse effects of COVID-19 on Latin American/Latino preschool children has added value beyond the description of temporary changes in movement behavior patterns during COVID-19. The impacts that COVID-19 might have on the health and wellbeing of people are expected to be long-lasting. [25][26][27] Before the pandemic, Latin American countries already faced a disproportionate double burden of disease, 28,29 which resulted from fast demographic and health transitions (including several large waves of migration to the USA), 30 in an environment with pervasive social and economic inequalities. 31,32 Understanding if and how COVID-19 has contributed to widening the health disparities gap and identifying factors associated with worse or better outcomes can help provide critical information to develop tailored programs for promoting resilience and healthy movement behaviors during and after the pandemic. Our results suggest that such strategies must employ a multilevel, equity-driven approach.
One key nding of our study was the role of social and environmental opportunities to play in facilitating healthy movement behaviors during the pandemic. In line with other studies. 15,16 opportunities for the child to play with other children or adults was an important factor associated with healthier levels of all three movement behaviors.
Similarly, having a space for children to play was associated with more favorable changes in movement behaviors.
It is possible that low-income families or those living in highly dense urban centers may be less likely to have designated "play spaces" in their private residences for their children to stay active and avoid sitting and screentime. Usually, the lack of private space for play can be addressed by using public open recreation spaces such as parks, playgrounds or public squares. However, during COVID-19, many local governments have either limited or fully restricted access to these public facilities -a situation which has likely contributed to widening the gap of access to health-promoting spaces. Furthermore, the social isolation that results from lockdowns or stay-at-home orders is likely to impact Latin American/Latino children and families, especially given their strong collectivistic (community-oriented) identity. In fact, there is evidence that for adults, use and access to places that facilitate social interaction is a key driver of PA behaviors 33,34 .
In contrast, we found that the level of access to electronic devices was associated with ST across all countries. A challenge faced by many caregivers is the ubiquity of screen devices. The mean number of connected devices owned by a USA household was 10.4 in 2020. 35 This provides children with more opportunities for ST, making it more challenging for caregivers to monitor the amount of ST for their child. However, we were encouraged that around two-thirds of caregivers set limits on the amount of time their child could use electronic devices and that setting limits effectively reduced their child's ST. This strategy has been shown to be highly effective in lowering ST among children 36,37 with the additional bene ts of higher social interaction levels, mostly verbally, 38 which is important in language development. 39,40 The results of this study should be interpreted with consideration of its limitations. Although we asked caregivers to retrospectively report on their children's patterns of movement behaviors before and during the pandemic, the crosssectional survey design precludes us from inferring causality. Because all of our measures are based on self-report, and caregivers are reporting on behalf of their children, there may be some degree of information bias. Further, we recruited our sample using online social media outlets. Although the income and education distribution of participants in our sample broadly resembles those of the general population in Chile and among USA Latinos, there is potential for selection bias as respondents are most likely to be those with access to and frequent use of online media.
Our study also had many strengths. We used standardized methods to collect comparable data across three countries. This is the rst study documenting movement behavior change patterns during COVID-19 among preschool children of Latin American origin or descent. We achieved robust sample sizes across all sites. Our results highlight consistent ndings across countries, suggesting broad applicability of our results across Latin America, with potential for developing general joint-cross national recommendations. At the same time, we uncovered some country-speci c results, which provide important information for developing locally tailored plans and actions to narrow the gap of healthy living opportunities during and after a major global crisis like COVID-19.

Conclusions
In this study of Latin American and Latino children during the rst wave of the COVID-19 pandemic, fewer met the WHO global guidelines for PA and ST and sleep quality declined compared with pre-COVID-19. Parents with higher education levels and those with older children were more likely to report unhealthy movement behavior changes. We found that not having access to an ECEC or to a space to play and not having the opportunity to play with someone were negatively associated with movement behaviors during the pandemic. This research provides further evidence on the impact the closure of pre-schools has had on children's physical and social health and some challenges parents found having their child at home during this pandemic stage, especially with managing a healthy level of screen time.