Multilevel Factors Associated With Obesity Among South Korean Adolescents Before and During the COVID-19 Pandemic

Abstract

Background:

Coronavirus disease 2019 (COVID-19)-necessitated lockdowns and school closures have limited social interactions among adolescents, which result in unhealthy behaviors. This study compared the multilevel factors associated with obesity among adolescents in South Korea before and during the pandemic.

Method:

We applied the social-ecological model and analyzed the 2019 and 2020 Korea Youth Risk Behavior Web-based Survey (KYRBS), including middle and high school students. We considered factors at the individual, family, and community levels in the logistic regression. Age and gender-adjusted obesity (body mass index ≥ the 95th percentile) was the dependent variable.

Result:

The prevalence of obesity slightly increased from 11.31% before the pandemic to 12.48% during it; the odds of obesity were 1.12 times (95% confidence interval [CI] = 1.08–1.16) as high during the pandemic compared to before it. Several individual factors (e.g., age, gender, fast-food consumption) and perceived family economic status were commonly associated with obesity before and during the COVID-19 pandemic. Physical activity and city type were related to obesity only before the pandemic. Downgraded economic status owing to COVID-19 was adversely related to obesity during the pandemic.

Conclusion:

Inclusive multilevel support is needed to combat obesity in adolescence during the pandemic and those economically impacted require additional support.

Introduction

Lockdowns and school closures during the coronavirus disease 2019 (COVID-19) pandemic have interrupted the daily routines and restricted the social interaction and school life of adolescents, which results in their unhealthy lifestyle behaviors (e.g., limited physical activity [PA], increased screen time, and delayed bedtime) that are reported as a cause of the increased prevalence of obesity in Europe,^1–3 the United States,⁴ and China.⁵ Given that the pandemic has limited social interaction, accounting for changes in adolescents' weight status before and during the pandemic is required. Furthermore, although adolescents show resilience and the ability to cope during the pandemic,^6,7 they are relatively incapable of coping with disasters or stress compared with adults and are generally passive to the impacts of crises and consequent social changes.^8,9 Thus, the pandemic, which may have exacerbated adolescents' health risks and accompanied by negative changes in their environment, has become a critical public health agenda.

In South Korea (hereinafter Korea), despite an initially less strict distancing policy than the United States or Europe, schools moved to an online modality. This change inevitably caused the reduction in adolescents' PA levels.^10,11 Furthermore, the pandemic has caused a decrease in Korea's GDP growth rate and the loss of approximately 460,000 jobs.¹² This study investigates not only individual factors but also the social and environmental aspects associated with the prevalence of obesity among Korean adolescents before and during the COVID-19 pandemic. Particularly, we examine the association between changes in economic conditions caused by the pandemic and the prevalence of obesity among Korean adolescents by applying the social–ecological model (SEM).

Conceptual Framework: SEM

SEM has been used to explain health behavior and develop health promotion interventions, encompassing factors at the individual, relationship, community, and societal levels (Fig. 1, left).^13–15 With the proposed SEM of layers influencing weight status, obesity is not just a matter of individual lifestyle but is also influenced by the overall social, economic, and environmental context.^16–19

Figure 1.

The SEM suggested by CDC and the adapted SEM. SEM, social–ecological model.

At the individual level, food intake and PA are significant factors associated with obesity.⁹ Family-level factors (e.g., socioeconomic status [SES])^17,18 and social environment factors such as social networks¹⁹ also affect obesity in children and adolescents. We applied the adapted SEM with the individual (age, gender, food intake, and PA), family (perceived familial SES and experience of downgraded SES owing to COVID-19), school (food education at school), and community (city type) levels (Fig. 1, right).

Methods

Data Sources

We analyzed the 2019 and 2020 Korea Youth Risk Behavior Web-based Survey (KYRBS). The KYRBS provides nationally representative cross-sectional data on various health statuses and behaviors among adolescents in middle and high schools, aged 12–18 years.²⁰ The KYRBS included 400 middle and 400 high schools from 17 provinces. There were sample allocations based on province and school characteristics (general vs. specific; coed or not).

The data comprised 14%–15% of middle and high schools and about 2% of all middle and high school students. The 2019 KYRBS was conducted between June and July and the 2020 KYRBS between August and November. The response rates were 95.3% in 2019 and 94.9% in 2020. We excluded cases with missing values for height and weight, thus resulting in a total sample of 55,749 for 2019 and 53,534 for 2020.

Ethical Approval

This study was reviewed and received an exemption determination from the institutional review board of the university to which the corresponding author is affiliated (No.: 2021-09-029).

Measures

The dependent variable was obesity, defined as age- and gender-based BMI percentile²¹ based on self-reported weight and height ≥the 95th percentile.

We used multilevel factors as independent variables. First, individual-level factors were age, gender, weekly fast food intake frequency, and PA (days of moderate exercise per week). Second, family-level factors were perceived familial SES (1 = high; 5 = low) and downgraded SES owing to COVID-19 (1 = totally agree, 4 = totally disagree). Third, the school-level factor was whether the participants received food education at school (no vs. yes). Finally, the community-level factor was the city type (small- and mid-sized cities vs. big cities).

Analytic Strategy

We used a chi-squared test to see whether the odds of being obese differed before and after the COVID-19 pandemic. Logistic regression was used to examine factors associated with obesity. We used STATA 17, and the significance level was set at p < 0.05.

Results

The proportion of adolescents with obesity slightly increased from 11.31% before COVID-19 to 12.48% during the pandemic (Table 1). The odds of obesity from an unadjusted model were 1.12 times (95% confidence interval [CI] = 1.08–1.16) as high during the pandemic compared with before it (χ² = 35.374; degree of freedom = 1; p < 0.05).

Table 1.

Characteristics of Survey Participants, n (%)

	Before the COVID-19 pandemic (n = 55,748)	During the COVID-19 pandemic (n = 53,534)
Weight categories based on age and gender-adjusted BMI
Not obesity	49,443 (88.69)	46,855 (87.52)
Obesity (≥95th percentile)	6305 (11.31)	6679 (12.48)
Individual level
Age, mean (SD)	14.96 (1.77)	15.09 (1.75)
Gender
Male	29,059 (52.13)	27,687 (51.72)
Female	26,689 (47.87)	25,847 (48.28)
Fast food consumption frequency in a week
None	10,203 (18.30)	9721 (18.16)
Several times a week	44,544 (79.90)	42,892 (80.12)
Every day or more often	1001 (1.80)	921 (1.72)
Moderate exercise in a week, mean (SD)	2.02 (2.12)	1.90 (2.12)
Family level
Perceived family economic status
High	6104 (10.95)	5862 (10.95)
Mid-high	15,794 (28.33)	15,033 (28.08)
Mid	26,856 (48.17)	25,738 (48.08)
Mid-low	5811 (10.42)	5724 (10.69)
Low	1183 (2.12)	1177 (2.20)
Experienced downgraded economic status due to COVID-19
Disagree at all	—	15,923 (29.74)
Somewhat disagree	—	21,350 (39.88)
Somewhat agree	—	13,189 (24.64)
Totally agree	—	3072 (5.74)
School level
Received food education at school
No	28,390 (50.93)	27,899 (52.11)
Yes	27,358 (49.07)	25,635 (47.89)
Community level
City type
Small- and medium-sized city	31,090 (55.77)	30,476 (56.93)
Metropolitan city	24,658 (44.23)	23,058 (43.07)

Before and during the pandemic, the participants were 15 years old on average with a standard deviation of 1.77 (before) and 1.75 (during); about half of them were girls, received food education at school, and lived in a metropolitan city. Although fast food consumption between the two periods was very similar, the mean hours of moderate exercise decreased from 2.02 to 1.9 hours. During the pandemic, approximately 5.74% strongly agreed and about a quarter (24.64%) somewhat agreed that they experienced downgraded SES.

Several factors were associated with obesity both before and during the pandemic (Table 2). For example, for both periods, at the individual level, the likelihood of obesity among older and male adolescents was higher than among younger and female adolescents. The possibility of obesity in groups consuming fast food several times a week or daily was lower than among those who did not consume fast food. However, moderate exercise, negatively associated with obesity before the pandemic, was not significant during the pandemic.

Table 2.

Logistic Regression Predicting Factors Associated With Adolescent Obesity

	Before the COVID-19 pandemic (n = 55,748)	During the COVID-19 pandemic (n = 53,534)
Individual level
Age	1.10^*** (1.08–1.12)	1.05^*** (1.03–1.06)
Gender (ref.: male)
Female	0.54^*** (0.51–0.57)	0.49^*** (0.47–0.52)
Frequency of fast food consumption in a week (ref.: none)
Several times a week	0.84^*** (0.78–0.90)	0.89^** (0.83–0.95)
Every day or more often	0.76^* (0.62–0.94)	0.88 (0.72–1.08)
Moderate exercise in a week	0.97^*** (0.96–0.98)	0.99 (0.98–1.00)
Family level
Perceived family economic status (ref.: high)
Mid-high	0.99 (0.90–1.09)	0.96 (0.88–1.06)
Mid	0.98 (0.90–1.08)	1.050 (0.96–1.15)
Mid-low	1.32^*** (1.19–1.48)	1.33^*** (1.18–1.48)
Low	1.63^*** (1.37–1.93)	1.46^** (1.23–1.74)
Perceived downgraded economic status due to COVID-19 (ref.: totally disagree)
Somewhat disagree	—	1.05 (0.98–1.12)
Somewhat agree	—	1.12^*** (1.13–1.31)
Totally agree	—	1.22^*** (1.09–1.37)
School level
Received food education at school (ref.: no)
Yes	0.99 (0.94–1.04)	0.98 (0.93–1.03)
Community level
City type (ref.: Small- and medium-sized city)
Metropolitan city	0.94^* (0.89–0.99)	0.99 (0.94–1.05)
Constant	0.05^*** (0.04–0.06)	0.09^*** (0.07–0.12)
Adjusted R²	0.019	0.022

Odds ratios, 95% confidence intervals in parentheses.

***

p < 0.001; ^**p < 0.01; ^*p < 0.05.

At the family level, the likelihood of obesity among adolescents who perceived familial SES as mid-low and low was higher than among those who perceived it as high both before and during the pandemic. In addition, during the pandemic, adolescents who perceived downgraded SES showed higher ORs than those who did not. Specifically, the likelihood of obesity among those who somewhat agreed (OR = 1.12; 95% CI = 1.13–1.31) and totally agreed that their familial SES had been downgraded (OR = 1.22; 95% CI = 1.09–1.37) was higher than of those who totally disagreed.

School-level factor was not significantly related to the likelihood of obesity for both periods. At the community level, the likelihood of obesity among adolescents who lived in metropolitan areas was lower than that among those who lived in small and medium cities only before the pandemic. However, the impact of the city type was not significant anymore during the COVID-19 pandemic.

Discussion

We found that the likelihood of obesity among adolescents was higher during the pandemic than before it. The SEM also demonstrated that lower level factors (i.e., individual and family level) were more critical than higher level factors (i.e., school and community level) with regard to obesity among Korean adolescents before and during the pandemic. These results are consistent with those of previous studies that applied SEM to investigate obesity in Korean children and adolescents.^22,23

Focusing on the differences in correlates of the prevalence of obesity before and during COVID-19, moderate PA was associated with a lower risk of obesity only prepandemic. Although Korea did not implement a robust social distancing policy during the pandemic, there were cases wherein school gym classes were limited, or private sport facilities were restricted.²⁴ Considering that the amount of PA at home in Korea is low,^25,26 such restrictions at school can be considered the cause of the decline in PA and the reason that its predictive power was not significant during the pandemic. Thus, despite the slight difference in PA time before and during the pandemic, it is possible that PA may inevitably be limited. Therefore, it is necessary to provide more opportunities for physical education even in remote classes and support exercise in their homes.

Similar to previous studies,^17,27 the likelihood of obesity among students who perceived their familial SES as low was higher than that in their counterparts. In addition, perceived downgraded SES owing to COVID-19 was negatively related to obesity during the pandemic, which contributed to the limited literature on obesity and the COVID-19 pandemic. At the family level, more policies are needed to support adolescents from low-income families or those experiencing a deep economic downturn directly resulting from COVID-19.

In Korea, the exclusion of public education was minimized by rapidly shifting to online classes without closing schools completely during the pandemic, but this transition was not easy for PA.²⁴ Given the negative impact of COVID-19 on school-based obesity intervention²⁸ and that PA in Korea is mainly centered on public education activities,^24,25 it is essential to support vulnerable groups that cannot afford to participate in obesity interventions through private activities.

Although the data analyzed in this study include a representative sample of Korean adolescents every year, there is a limitation in that they do not reveal causality. Owing to the constraints of secondary data, this study could not examine parental and community characteristics related to obesity among adolescents, although family and community factors had a stronger relationship with childhood obesity than did individual factors in the United States.¹⁴ Future research should comprehensively address the various multilevel factors that influence obesity among adolescents according to the duration of COVID-19.

Footnotes

Authors' Contributions

Conceptualization of the study was done by S.H.J. and H.H. Data curation was carried out by S.H.J. Formal analysis was taken care by S.H.J. Methodology of the study was taken care by S.H.J. and H.H. Project administration was done by H.H. Visualization was taken care by S.H.J. Writing—original draft—was done by S.H.J. and H.H. Writing—review and editing—was done by S.H.J. and H.H. All authors were involved in writing the article and approved the submitted version.

Funding Information

No funding was received for this article.

Author Disclosure Statement

No competing financial interests exist.

References

Medrano

, Cadenas-Sanchez

, Oses

, et al. Changes in lifestyle behaviours during the COVID-19 confinement in Spanish children: A longitudinal analysis from the MUGI project. Pediatric Obes, 2021; 16:e12731.

Ten Velde

, Lubrecht

, Arayess

, et al. Physical activity behaviour and screen time in Dutch children during the COVID-19 pandemic: Pre-, during-and post-school closures. Pediatric Obes, 2021; 16:e12779.

Pietrobelli

, Pecoraro

, Ferruzzi

, et al. Effects of COVID-19 lockdown on lifestyle behaviors in children with obesity living in Verona, Italy: A longitudinal study. Obesity, 2020; 28:1382–1385.

Burkart

, Parker

, Weaver

, et al. Impact of the COVID-19 pandemic on elementary schoolers' physical activity, sleep, screen time and diet: A quasi-experimental interrupted time series study. Pediatric Obes, 2022; 17:e12846.

Xiang

, Zhang

, Kuwahara

. Impact of COVID-19 pandemic on children and adolescents' lifestyle behavior larger than expected. Prog Cardiovasc Dis, 2020; 63:531.

Kuhlman

, Straka

, Mousavi

, et al. Predictors of adolescent resilience during the COVID-19 pandemic: Cognitive reappraisal and humor. J Adolesc Health, 2021; 69:729–736.

Ellis

, Dumas

, Forbes

. Physically isolated but socially connected: Psychological adjustment and stress among adolescents during the initial COVID-19 crisis. Can J Behav Sci, 2020; 52:177–187.

Kusumaningrum

, Siagian

, Beazley

Children during the COVID-19 pandemic: Children and young people's vulnerability and wellbeing in Indonesia. Child Geogr, 2021:1–11. [Epub ahead of print]. doi: 10.1080/14733285.2021.1900544.

Kousky

Impacts of natural disasters on children. Future Child, 2016; 26:73–92.

10.

Kim

, Kwon

, Choe

, et al. COVID-19-related school closing aggravate obesity and glucose intolerance in pediatric patients with obesity. Sci Rep, 2021; 11:5494.

11.

Dighe

, Cattarino

, Cuomo-Dannenburg

, et al. Response to COVID-19 in South Korea and implications for lifting stringent interventions. BMC Med, 2020; 18:321.

12.

Kang

, Min

S-W

, Park

. The impact of the COVID-19 pandemic on the Korean economy and industry: An interim assessment of one year after outbreak. KIET Ind Econ Rev, 2021; 26:6–17.

13.

Stokols

Establishing and maintaining healthy environments: Toward a social ecology of health promotion. Am Psychol, 1992; 47:6–22.

14.

Ohri-Vachaspati

, DeLia

, DeWeese

, et al. The relative contribution of layers of the Social Ecological Model to childhood obesity. Public Health Nutr, 2015; 18:2055–2066.

15.

, Rillamas-Sun

, Bishop

, et al. Together We STRIDE: A quasi-experimental trial testing the effectiveness of a multi-level obesity intervention for Hispanic children in rural communities. Contemp Clin Trials, 2018; 67:81–86.

16.

Office of Disease Prevention and Health Promotion. Physical Activity, Nutrition, and Obesity [cited 2022 Jan 8]. https://www.healthypeople.gov/2020/leading-health-indicators/infographic/nutrition-physical-activity-and-obesity Last accessed May 20, 2022.

17.

Lindberg

, Persson

, Danielsson

, et al. Obesity in childhood, socioeconomic status, and completion of 12 or more school years: A prospective cohort study. BMJ open, 2021; 11:e040432.

18.

Zeller

, Reiter-Purtill

, Modi

, et al. Controlled study of critical parent and family factors in the obesigenic environment. Obesity, 2007; 15:126–136.

19.

Koehly

, Loscalzo

. Adolescent obesity and social networks. Prev Chronic Dis, 2009; 6:A99.

20.

Kim

, Choi

, Chun

, et al. Data resource profile: The Korea youth risk behavior web-based survey (KYRBS). Int J Epidemiol, 2016; 45:1076–1076e.

21.

Korea Disease Control and Prevention Agency (KDCA). Growth Chart for Children and Adolescents. 2017. https://knhanes.kdca.go.kr/knhanes/sub08/sub08_01.do Last accessed May 20, 2022.

22.

Park

SA.

Multilevel model analysis of factors associated with adolescent obesity. Health Soc Welfare Rev, 2019; 39:314–347.

23.

, Jeong

Y-H

. Risk factors for overweight and obesity of Korean preschool children: Based on the ecological model. J Korean Public Health Nurs, 2019; 33:175–187.

24.

Lee

The necessity of youth physical activity and role and task of school physical education in COVID 19. Korean J Spt Pedagog, 2021; 28:175–198.

25.

Lee

, Kang

. An analysis of gender difference on physically active lifestyle among Korean middle school students. Korean J Spt Pedagog, 2015; 22:53–70.

26.

Jang

J-H

, Lee

, Park

J-Y

. Analysis of daily physical activity of male and female middle school students according to periods. Korean J Sport Sci, 2020; 31:581–592.

27.

, Lee

, Kim

, et al. Eating habits, physical activity, nutrition knowledge, and self-efficacy by obesity status in upper-grade elementary school students. Nutr Res Pract, 2016; 10:597–605.

28.

Camp-Spivey

, Newman

, Stevens

, et al. “We've Had to Build the Plane as We Flew It.”: Impacts of the COVID-19 pandemic on school-based weight management interventions. Child Obes, 2021; 17:1–10.