Changes in body mass index and behavioral health among adolescents in military families during the COVID-19 pandemic: a retrospective cohort study | BMC Public Health

Major findings:

We identified more than 160,037 adolescents from military families, ages 13 to 15 years in FY 2017–2018, who were followed during FY 2020 to June 2021. The cohort was assessed pre-pandemic for body composition (obese, overweight, normal weight, or underweight) and for behavioral diagnoses (mood disorders, anxiety disorders, conduct disorders, and ADD/ADHD), at which time, rates of these conditions were similar to those in the greater U.S. population of adolescents [6] As described in the Introduction, however, children of military families have three potential advantages over their peers: they are universally insured; their military parents have guaranteed employment through the end of their contracts; and they tend to choose military service for themselves at higher rates than do adolescents in the greater U.S. population, raising the possibility of mental and physical fitness as cultural or family expectations. Therefore, the follow-up of this cohort during the height of the pandemic aimed to determine whether military service of at least one parent was protective against increases in overweight, obesity, and behavioral health disorders.

Overall, a majority (> 50%) of adolescents in the Underweight, Healthy Weight, and Obese categories in 2017–2018 remained in those same categories during the follow-up period. Of those in the Overweight category, the largest proportion (36.3%) remained in that category, with approximately 35% decreasing to Healthy Weight. Of the adolescents who were overweight in 2017–2018, 28.5% progressed to obesity in the follow-up period, and a small number (0.27%) progressed to underweight. Of adolescents who were healthy weight in 2017–2018, 4.0% progressed to obesity, 3.7% progressed to underweight, and 8.2% to overweight during the follow-up period. This appears to match data in the general population which documents a rise in child and adolescent obesity, and in the rate of BMI increases, during the COVID-19 pandemic [11, 12]. The Centers for Disease Control and Prevention (CDC) acknowledges that pandemic-related school closures likely reduced children’s access to “structured physical activity” and to healthy foods [12]. Nutritional access could influence both overweight, which is robustly studied in literature, and underweight, which is far less studied in COVID-19 despite being of concern. A 2020 survey by Sharma, et al. [13] revealed food insecurity among 93.5% of respondents during the first year of the pandemic; however, comparative literature associates this more strongly with overweight and obesity than with underweight [14]. The overall rise in BMI among adolescents in military families also mirrors that of the active duty service members themselves, with the Army, Navy, Air Force, and Marine Corps all reporting increases in BMI during the COVID-19 pandemic [15]. Notably, the DoD also moved to a largely remote work platform during this time, with approximately 88% of military and federal civilian respondents reporting a full or partial transition to telework [16]. Taken together, this suggests that neither universal insurance nor military culture was sufficient to eliminate adverse changes in BMI among adolescents or their military parents during the COVID-19 pandemic.

The results for behavioral health were similarly mixed. Prevalence of conduct disorders and ADD/ADHD decreased during the follow-up period; however, prevalence of mood and anxiety disorders, and suicide ideation and suicide attempt, increased during the follow-up period. Conduct disorders and ADD/ADHD co-occur frequently [17], as do anxiety and mood disorders, and therefore each pair would be expected to trend in the same direction. It may be that these pairs are oppositely affected by school closures, with conduct disorders and ADD/ADHD responding positively to a more constrained and predictable home environment, while anxiety and mood disorders responded negatively to the reduced opportunity for social interaction [18]. Alternately, the decrease could be due to reduced access to psychological and primary care during the pandemic, with ADD/ADHD diagnoses especially subject to bias as the manifestations are subtler than for conduct disorders. Effects of the pandemic, access to care, and particularly school closures on youth with conduct disorder and ADD/ADHD are understudied in literature and therefore worthy of future research.

The most concerning findings were the increase in suicide ideation and attempts among adolescents and young adults in this study. The CDC reports an approximately twofold increase in emergency department (ED) visits for suspected suicide attempts in adolescents aged 12–17 during the period of Spring 2020 to Winter 2021, particularly among females [19]. Among young adults aged 18–24, inclusive of the top age range in our study, the CDC reports a decrease in number, but slight (1.1–1.4-fold) increase in rates of ED visits for suspected suicide attempts. The 77% increase in suicide attempts seen in our study falls within the 1.1–twofold increase reported by the CDC. However, our study reports a 43% increase in suicide ideation, in contrast to CDC-reported figures of 25.5% of adolescents aged 18–24 having considered suicide within the last 30 days [20]. This difference may be due to multiple factors including differing lengths of study period, the use of self-reported surveys in the CDC study and the smaller sample size (5470). Further research is needed to address any other factors which may be driving suicide ideation among military adolescents.

While pandemic-associated behavioral health effects are understudied in the active duty population, early work compares the stress of the pandemic to that associated with humanitarian disaster missions, and anticipates similar effects on mental health [21]. The same article also recognizes the common stressors associated with pandemic-related social distancing, which affected military and civilian populations alike [21]. While universal insurance may enable greater access to behavioral health resources, neither universal insurance nor military culture was sufficient to eliminate adverse changes in mental health status among adolescents or their military parents during the COVID-19 pandemic.

Racial disparities

As described in Results, the largest percentages of adolescents who were underweight (61.8%) or obese (49.4%) during the follow-up period were of White race, which is expected given that White race accounted for 64% of study population (Table 1). Black and Other race adolescents had the greatest percent changes (51.1%) in underweight BMI and Native American/ Alaska Native adolescents had the greatest percent change (20.3%) in obesity, compared to the previous study period. The percent change in underweight among Black adolescents is almost double that for White adolescents (26.6% change) and is suggestive of disparity. These issues could include reduced access to food, especially if the non-military parent lost income due to the pandemic; increased physical activity; or eating disorders. These latter, particularly bulimia, may occur in Black adolescent women at a rate 50% higher than in White adolescent women, often as a maladaptive response to stress, while disordered eating is approximately 1/3 as likely to be diagnosed in Black vs. White adolescent women with the same symptoms [22]. Discussion of BMI in U.S. adolescents focuses strongly on overweight and obesity rather than underweight, and comparisons across racial groups are rare. Therefore, these findings represent an important contribution to the literature.

Differences between racial groups were also observed in behavioral health. Black adolescents had notably greater percent increases in anxiety disorders (104.1%) and suicide attempt (90.0%), compared to White adolescents (83.8%, 72.9% respectively). The increase in suicide attempt is particularly concerning, as suicide in 2018 became the third leading cause of death among Black teens aged 15–19 [23]. The National Institutes of Mental Health identifies several factors including differential access to care, and mistrust or negative perception of providers leading to lower completion rates for depression treatment programs, as potential contributors to increased suicide rates in Black adolescents. In our study, this population also showed greater negative percent changes in diagnoses of conduct disorders (-20.1%) and ADHD (-19.4%) compared to White adolescents (-14.7% and -6.1%, respectively). Prior research shows that Black adolescents in the MHS are more likely to be diagnosed with conduct vs. mood disorders, compared to White adolescents with similar symptoms [24]. It is possible that the decrease in conduct disorders reflects a decrease in triggering symptoms outside of a home environment as described earlier, or a decrease in school-based referrals for symptoms which are viewed as problem behaviors. Further research is needed to answer these questions.

Racial differences were also noted for Asian/Pacific Islander adolescents, who had a higher percent increase in anxiety, suicide ideation, and suicide attempt; lower percent increase in mood disorders; and greater percent decrease in ADD/ADHD and conduct disorders, compared to White adolescents. Comparison to published literature shows mixed results, with a 2007 study showing reduced rates of conduct disorders among Asian young adults but increased rates among Native Hawaiian/Pacific Islander young adults, compared to their White counterparts [25], while a 2018 study shows a greater prevalence of anxiety disorders among second but not first-generation Asian young adults [26]. Both prior studies indicate cultural differences as drivers of mental health and mental health care, but these factors could not be assessed from our dataset. The percent increases in suicide ideation (51.6%) and suicide attempt (76.1%) are particularly concerning in Asian and Pacific Islander young adults, as the CDC reports suicide as the second most common cause of death in this population as of 2020 [27]. Further research is needed to identify the risk factors for suicide ideation and attempt in this population and to determine effective, culturally-sensitive interventions.

Cultural implications of military service

In 2018, approximately 1% of U.S. adults were expected ever to serve in the military [28], and approximately 25% of new recruits are drawn from military families [7]. This significant overrepresentation demonstrates the cultural expectation of military service for young adults in military families, and indicates that this expectation should be considered when designing interventions.

Over 9000 young adults, or approximately 23% of the cohort, were categorized during the follow-up period as obese, which is a barrier to enlistment. An additional 13% were overweight, which is not necessarily a barrier depending on exact BMI and the standards of the Service into which the person is enlisting. For instance, the Army standards for a male recruit of 5 feet, 10 inches tall include a maximum weight of 189 for those 17–20 years of age and 192 for those 21–27 years of age [29], corresponding to an overweight BMI of 27.1 and 27.5 respectively. However, weights above those values would be disqualifying for enlistment even though the BMIs are below the point of obesity. Lifestyle interventions remain the treatment of choice for reducing weight in adolescents, though drugs such as liraglutide and semaglutide are gaining acceptance for those with obesity [30, 31]. Bariatric surgery may be recommended for those with severe obesity [30, 32], but long-term effects on adolescents are not known [30] and the surgery itself is disqualifying for enlistment [8].

Many mental health issues including depression and anxiety are also disqualifying, if care was received for 12 cumulative months or within 36 months prior to accession [33]. Our data shows 23% of young adults having at least one mental health or behavioral health diagnosis during the follow-up period, representing nearly 88,000 people ineligible for service. Many military programs focus on family-level strategies, particularly for those with school-aged children [34, 35], which may serve a protective role against poor mental and behavioral health and thereby preserve fitness to serve. However, while some of the programs offered may be evidence-based, the most widely available are not externally validated and may not may not meet the needs of young adults vs. school-age children. Further research is needed to address this issue.

Next steps and recommendations

As described in our previous work, children in military families have access to a number of resources designed to preserve mental and physical fitness, including both school-based and military-provided programs. However, schools and universities were closed or operating on virtual status during much of 2020 and early 2021 due to the COVID pandemic, limiting access to many programs. This is particularly important given the increased rate of suicide attempts seen in our cohort. The Department of Defense (DoD) should invest in a multi-layered health infrastructure as recommended by the CDC [19], and should ensure sure that interventions are easily accessible even during times of disruption, are evidence based, and are validated. Further research should also be performed to determine whether those at risk for suicide are able to access care through the MHS and whether they are alternatively seeking care outside the MHS. Similarly, the DoD should also prepare to implement validated health and fitness interventions where and when these are not accessible through the schools or communities.

Limitations

As with all studies that rely on claims data, this one is subject to potential errors in coding, such as misclassification and underreporting, and the loss of clinical nuance as notes were not captured in this data set. Of the total cohort, we were only able to identify 11% with a BMI recorded in both time periods, which could potentially lead to significantly underestimating the changes observed in MHS adolescents; however, our data reflects similar findings observed in national statistics during the pandemic. In-person healthcare services were significantly reduced during the pandemic as well, which could have impacted our behavioral health prevalence and BMI estimates. As noted above, combination of demographic categories such as Asian with Pacific Islander, and small numbers in some populations such as Native American/Alaska Native, hamper full analysis of factors affecting mental and behavioral health. This study does not clearly separate the effects of age from effects of the pandemic on the rates of behavioral health diagnoses; for instance, in ADD/ADHD which is commonly diagnosed in children and for which rates are reasonably expected to decline during adolescence, or for anxiety and depression which are known to increase during the teen years. Although findings in the greater U.S. population support the idea of the pandemic having negative effects on adolescent behavioral health, further research would be needed to separate conclusively the effects of progression through adolescence from the effects of the pandemic. Finally, this study does not capture care paid for outside of the MHS, such as peer or spiritual counseling for suicide ideation. This is particularly important to consider in light of fitness to serve, for which thoughts of and attempts at suicide are considered disqualifying conditions [9].