Changes in substance use among adolescents before and during the COVID-19 pandemic in Guatemala

Objectives Due to the COVID-19 pandemic, on March 16th, schools had to be closed in Guatemala and went to online teaching. We sought to analyze the change in substance use among high school students in Guatemala associated with the lockdown. Methods Data from two surveys (2019, n=2096, and 2020, n=1606) of a student cohort in private high schools in Guatemala City was used. Logistic models for past 30-day cigarette, e-cigarette, marijuana, and alcohol (including binge drinking) were used, regressing these on survey wave, while adjusting for sex, scholastic performance, high school year of student, parental education, substance use, and household member tobacco use. Results Prevalence declined for smoking (10% to 3%, p<0.001), e-cigarette (31% to 14%, p<0.001), marijuana (4.3% to 1.9%, p<0.001), and alcohol use (47% to 38.5%, p<0.001), and binge drinking (24% to 13%, p<0.001). Adjusted models showed wave 2 associated with lower odds of using cigarettes (AOR=0.44, 95%CI=0.32-0.62), e-cigarettes (AOR=0.41, 95% CI=0.35-0.49, p<0.001), and binge drinking (AOR=0.73, 95%CI=0.59-0.89; p=0.002) Conclusion Among Guatemalan adolescents, COVID-19 restrictions were associated with a significant decrease in smoking, e-cigarette use, and binge drinking.


Introduction
Adolescence is a critical development stage of life that is accompanied by physical, cognitive, emotional, social, and behavioral changes (1).It is the peak time for substance use initiation, including smoking, alcohol, and other illicit drugs (2).Substance use might disrupt cognitive and emotional development and can also lead to substance use disorders, dependence, and poorer physical and mental health (1,3) and has been reported as a leading risk factor for premature death globally (4).Substance use prevalence varies across the globe, for example, people from lowand middle-income countries (LMICs) as defined by the World Bank (5), were reported to have higher smoking prevalence than those in high-income countries (HIC) (6) due, in part, because of the rapid pace of globalization which facilitated the expansion of tobacco markets across borders (7,8).
Electronic Nicotine Delivery Systems (ENDS) were first introduced in the market in 2007 (9,10) and first documented in Guatemala in 2014 (11,12) and now include Heated Tobacco Products (HTPs).Since then, consumption has increased, particularly in HIC.In the U.S., for example, e-cigarette use has been the most commonly used tobacco product among youths since 2014, and in 2020, 13.1% of middle and high school students were current users with 80% of them using flavored e-cigarettes (13).This increase is consistent with other countries as well (14)(15)(16)(17)(18)(19) and in LMICs and upper-middle-income countries, like Guatemala (20).Even though Guatemala signed and ratified the Framework Convention on Tobacco Control (FCTC) (21), it lacks strong tobacco control strategies and ENDS remain unregulated.In 2014, 15.7% of adolescents were smoking compared to 5.6% using e-cigarettes (22).However, in 2019, we found a 27.7% e-cigarette and 8.7% cigarette use prevalence among adolescents (23).Therefore, it appears that consumption patterns shifted towards e-cigarettes in Guatemala City over time.
In Guatemala, since the first COVID-19 case was detected in March of 2020, curfews, lockdowns, and building occupancy restrictions based on epidemiological indicators (45, 46) were implemented (47,48).Schools were shut down on March 22 nd , 2020, and turned to a fully virtual mode throughout that year until 2021.Regarding tobacco control, Guatemala has only a weakly enforced smoke-free environment law and consequently a frail tobacco control program (49).In addition, e-cigarettes are readily available and heavily advertised at the point of sale (50).Therefore, we sought to assess whether the COVID-19 school shutdown influenced adolescent cigarette, e-cigarette, marijuana, and alcohol (including binge drinking) use in a sample of high school students in Guatemala City.

Methods
We analysed data from an ongoing, open cohort of students attending private schools in Guatemala City (51).Paper surveys were conducted in person from May to September 2019 (wave 1) and online from June to November 2020 (wave 2).Of the 2666 students, 1036 students were surveyed at both waves, and 1630 were surveyed once, due to maturation (out of and into high school) or absence during survey administration (n=3702 observations including those surveyed 2 times plus those surveyed only once).Passive parental consent was obtained by asking parents to respond to a consent letter only if they wished their kid to be removed from the study.After that, students provided active assent as described elsewhere (23,52).Study protocols and data collection instruments were approved by the Institutional Review Board of Central America and Panama Nutrition Institute (INCAP).
The questionnaire focused on smoking and e-cigarette use and their risk factors, including alcohol and marijuana use.It was adapted from the Population Assessment of Tobacco and Health (PATH) study (53) and previously implemented in Mexico (54) and Guatemala (23).Current use was defined as using any product (cigarettes, e-cigarettes, marijuana, and alcohol) in the past 30 days.
Participants were asked about their marijuana used as "during the past 12 months, have you smoked marijuana?"For the analysis, it was coded as a dichotomic variable (yes/no).We assessed binge drinking by asking "Have you ever had about 4 or more drinks on one occasion over the prior 30 days?".They also reported on smoking and vaping among household members and their five closest friends dichotomized variables as follows: (1) At least one household member smokes; (2) at least one household member uses e-cigarettes; (3) at least one friend smokes; (4) at least one friend uses e-cigarettes.Sociodemographic variables included gender, grade (7 th to 11 th ), scholastic performance (from 1-100), and parental highest educational attainment (None, middle school and high school, some college and bachelor's degree or more).
For the descriptive analysis, we compared sample characteristics across survey waves using omnibus chi-square statistics.We used logistic Generalized Estimating Equations (GEE) to estimate the influence of the COVID-19 lockdown over the use of cigarettes, ecigarettes, marijuana, alcohol, and binge drinking.Models were adjusted for repeated measures and clustered within individuals using an exchangeable correlation structure.Models were adjusted for school performance, friends, and household member smoking and/or vaping, and sociodemographic characteristics.In total, we fitted 5 logistic GEE, one for each outcome assessed, all models were also adjusted for the use of the other substances that were not modeled as the outcome.All analyses were conducted in Stata version 17 (StataCorp LLC, College Station, Texas).

Discussion
To our knowledge, this is the first study assessing adolescent substance use before and during the COVID-19 restrictions in an upper-middle-income country in Latin America.We found that the reported smoking, vaping, marijuana and alcohol use, and binge drinking significantly decreased during COVID-19 restrictions.These declines remained significant in the adjusted models for all substances except alcohol and marijuana use.Although reports in other countries Similarly, in Ontario, Canada, data among 14-18-year-old adolescents before and after the emergency stay-at-home order in Ontario, yield a significant decrease (approximately 5%) in substance use after the shutdown (except for alcohol) (3).Even though these findings agree with what we found, these countries have significantly different tobacco control environments compared to Guatemala.We also found that the reported friends' and household members' substance use was significantly associated with teenagers' substance use, and they all significantly decreased in wave 2 (Table 1).This may be associated with a change in substance use during COVID-19 restrictions and is consistent with studies that suggest social networking as a significant means for initiation and continuation of substance use (60,61) and particularly e-cigarettes.It also supports the idea that ENDS are generally obtained through peer networks and not from family members or direct purchasing as suggested elsewhere (62, 63).The report of cigarette and e-cigarette use significantly declined in our sample, this makes sense during the lockdowns given that adolescent substance use is a highly social behaviour (57) and is consistent with another publication from Guatemala (64) a Canadian study in teens who provided information 3 weeks before and 3 weeks after the start of the COVID-19 lockdown (3).
Regarding alcohol use, binge drinking, and marijuana use, we also found a consistent decrease although in the adjusted models binge drinking decrease was not statistically significant.Other studies show mixed results, for instance, the Canadian study reported that fewer teens reported drinking, or using marijuana, however, alcohol use remained unchanged (3).Interestingly, among teens who did use substances, the mean number of days they used alcohol (0.76-0.96) and marijuana (0.94-1.1 days) were significantly higher after COVID-19 compared with before (3).Finally, another study among 12 th graders reported that adolescent marijuana and alcohol use had no significant changes during the pandemic despite decreases in the substances' perceived availability (55).
These mixed results underscore the importance of fully characterizing the effects of the pandemic on adolescent substance use patterns in the context of well-known risk factors for substance use such as drug availability, association with peers who use drugs, a lack of parental supervision, boredom, and coping with negative affect (44,57) and be better prepared to address them.
Possible reasons that could explain our findings include that societal restrictions during the pandemic may have contributed to a decrease in the time youth spent with their peers, which ultimately has an impact on adolescent substance use though studies on this effect were done before the pandemic (65,66).Another study suggested decreased commercial availability and access to vape devices which may be associated with decreased rates of vaping (39).Other studies however, attribute the increased amount of time spent within the household as a risk factor for familiar substance use exposure (67).These findings warrant future research on how significant policy changes affecting social behaviour (i.e.lockdowns) can affect substance use while also pointing to the fact that different substances may respond to its various determinants in a particular way and thus public policies will not function as a "one size fits all" for each substance.

Limitations
This study should be interpreted considering some limitations.First, data from both waves is self-reported.However, this method has proved to be reliable to assess substance use in adolescents (68) and therefore internal validity should not be a source of bias.Also, we have reports that in 2019 and 2020, the Centers for Disease Control in the United States reported various outbreaks of Lung Injury Associated with the Use of E-cigarettes (EVALI), and thus we need to considerate that this could have also affected the e-cigarette use in adolescents in our sample, however, no reports of EVALI cases in Guatemala have been documented.
Furthermore, we changed survey implementation across waves, from in-person to online self-administration due to COVID-19 restrictions.It is yet unclear whether any resulting bias would have under-or over-estimated the changes we found.Finally, cigarettes and e-cigarettes utilize high-income status and exclusivity as part of its advertising and people with low-income status are less likely to afford these products and given that our sample was drawn from private schools in Guatemala City, our results might not be generalizable to public schools, lower-income status populations and rural Guatemala.Regardless, our findings provide evidence of the potential impact that the COVID-19 pandemic had in the substance use from our sample.

Conclusion
Our results suggest that during the COVID-19 restrictions, substance use among Guatemalan adolescents declined, as well as smoking and e-cigarette use among peers and household members.However, declines in marijuana use and alcohol consumption appear independent of these influences and merit further investigation.

a
Measured on a scale from 1-100 b At least one parent attended at least high school c Responded at least frequently to the question: "in the last 30 days how frequently have you seen an e-cigarette ad?" d Smoking at least once in the past 30 days e Responded at least once to the question "In the last 30 days, how many days did you drink an alcoholic beverage?"f Using an e-cigarette at least once in the past 30 days g Had 4 or more drinks in one unique occasion in the past 30 days.h E-cigarette use at least once in the past 30 days Monzon et al. 10.3389/fpsyt.2024.1331962Frontiers in Psychiatry frontiersin.org

TABLE 1
Demographic characteristics and substance use in Wave 1 and Wave 2 of adolescents' cohort in Guatemala City.

TABLE 2
Adjusted Odd Ratios of past 30-day substance among adolescents.