Healthy Behaviors and Service Member Mental Health

Did you know the practice of many of the same Performance Triad behaviors that support Service member physical health and readiness may also benefit Service member mental health and readiness?

Results of a study published in the March 2021 edition of the Journal of Affective Disorders provide strong support for the importance of healthy behaviors, including adequate sleep, physical activity, abstaining from smoking, and limiting alcohol intake, for helping ensure Service member mental health and well-being. Based on their study results, the authors suggest the practice of these healthy behaviors, when combined with professional mental health treatment, might also be beneficial for managing symptoms of depression, anxiety, and post-traumatic stress disorder (PTSD) in Service members when they do arise.

In this study, researchers examined how behaviors such as sleep, physical activity, and substance use might be related to symptoms of depression, generalized anxiety disorder (GAD), or PTSD in active-duty U.S. Service members. To investigate this, researchers leveraged information about U.S. Service members’ sleep, physical activity, smoking, and alcohol use collected as part of the 2015 Health Related Behaviors Survey.

The researchers reported independent associations of sleep, physical activity, and substance use with one or more psychological disorders. Moreover, some associations appeared to be bi-directional, as depression, GAD, and PTSD were in some cases associated with being less likely to practice beneficial health behaviors. For example, longer nightly sleep was consistently beneficial for Service members in terms of mental health outcomes, while the presence of any psychological disorder significantly lowered the amount of nightly sleep reported by Service members. Together, these findings support the importance of sleep and activity for mental health.

If you are unfamiliar with research-specific terms like “independent association” and “bi-directional,” you can see a quick explanation by expanding the following section: “Association” refers to the presence of a relationship between two factors, such as the relationship between inadequate sleep and blood pressure. Often, studies show that study participants who sleep less have higher blood pressure. These two factors are said to be related or associated. When two factors remain related, even when other factors that could possibly explain that relationship are accounted for, they are considered independently associated. For instance, older people are often at greater risk for high blood pressure. However, if we account for age in the sleep and blood pressure example by investigating both older and younger participants, and both older and younger participants who sleep less still have higher blood pressure than those who sleep more, sleep and high blood pressure are independently associated.

“Bi-directional” refers to a relationship between two factors that works in both directions. Using the example just discussed, if high blood pressure also caused participants to sleep more or sleep less, then the relationship between the two factors would be said to be bi-directional in nature. This type of relationship must be determined using special statistical analyses.

Now that you have a general idea of the study's results, let's take a closer look at how the study was conducted and the specific results leading to researchers' conclusions.

WHO: 12,708 active-duty military of the Army, Navy, Air Force, Marine Corps, and Coast Guard

28.2% between ages 17 and 24; 41.7% between ages 25 and 34; 23.0% between ages 35 and 44; 7.1% age 41 and older
14.7% female
39.2% Minority race/ethnic group

38.0% Army; 22.9% Navy; 21.6% Air Force; 14.1% Marine Corps; 3.3% Coast Guard
Primarily enlisted (82.0%)

HOW: This study was observational, meaning study participants were only observed; they were not given any interventions such as medical treatments or medications. This study was also cross-sectional, meaning researchers only collected information about participants at a single time point.

To answer questions about how Service members' health-related behaviors might be related to their likelihood of having depression, GAD, or PTSD, researchers used U.S. Service members' anonymous responses to questions in the 2015 Health Related Behaviors Survey. Participants answered questions pertaining to social and demographic factors such as age, sex, and marital status, body measurements, and their health-related behaviors including sleep, physical activity, smoking, and alcohol consumption. Participants also answered questions about mental health symptoms related to depression, GAD, and PTSD. To see more detailed information about the survey questions, expand the following section:

Sociodemographic factors:

Age
Race/ethnicity
Service branch
Pay grade
Marital status
Education level (high school diploma or equivalent, some college, 2-year college degree, 4-year college degree, post-graduate school)

Anthropometrics (body measurements)

Height
Weight
Body mass index (calculated from participants' self-reported height and weight)

Health-related behaviors

Normal, overweight, or obese BMI
Frequency and duration of:

Moderate physical activity
Vigorous physical activity
Strength training

Cigarette smoking
Alcohol intake:

Number of drinks per day
Binge drinking
Problem drinking (binge drinking on more than one occasion or high score on AUDIT-C questionnaire, which assesses alcohol use and dependency)

Illicit drug use
Prescription drug misuse
Sleep hours in a typical 24-hour period on weekdays and weekends

Psychological disorder symptoms (through use of specialized questionnaires)

Generalized anxiety disorder
Depression
PTSD

Whether they had ever been informed by a medical professional of the following adverse health conditions (summed for total number of conditions):

High blood pressure
High blood sugar or diabetes
High cholesterol
Respiratory problems
Arthritis
Heart disease or other heart condition
Ulcer
Skin cancer
Other cancer

The researchers then used a method of statistical analysis called regression to examine possible relationships between health-related behaviors and participants' mental health. To account for the contribution of all factors potentially related to the likelihood of depression, GAD, or PTSD, researchers considered all factors simultaneously in their regression analysis, referred to as a fully adjusted analysis. This means in addition to including all health-related behaviors in the analyses, factors such as sex, military service branch, and pay grade were also included. For a more detailed explanation of relationships the researchers examined, expand each section below:

Did any of the health-related behaviors reported by Service members predict any of the psychological disorders examined in the study? This analysis was done to better understand if any of the health-related behaviors predicted the likelihood of depression, GAD, or PTSD. For example, did sleeping more hours each night make it more or less likely that Service members would report depression symptoms?
Did having a particular psychological disorder influence the likelihood of having a second disorder, known as co-morbidity? Because mental health is complex and having one psychological disorder might increase the likelihood of having another disorder, this analysis was done to better understand this relationship. For example, were Service members with PTSD more or less likely to have depression?
Did co-morbid psychological disorders change any of the associations observed in question one? This analysis was done to better understand how co-morbid psychological disorders might affect associations previously observed between health-related behaviors and single-psychological disorders. For example, although participation in moderate aerobic activity might have lowered the likelihood of reporting PTSD, was participation in moderate aerobic activity while having depression still associated with lower likelihood of reporting PTSD?
Did any of the psychological disorders self-reported by Service members predict any of the health-related behaviors in the study? This is the reverse of question one and was performed to examine whether there were any bi-directional relationships between health-related behaviors and psychological disorders. In other words, did having a psychological disorder make it more or less likely that a health-related behavior would be practiced? For example, were Service members who reported depression symptoms more or less likely to sleep fewer hours each night?
Did the presence of co-morbid psychological disorders change any of the associations observed in question four? Similar to question three, this analysis was done to determine if the presence of co-morbid psychological disorders would affect the likelihood of health-related factors being practiced. For example, were Service members who had depression more or less likely to report current smoking when they also had PTSD?

Man on bike performing vigorous activity

Main Findings: Sleep was most consistently associated with all three psychological disorders. Specifically, short sleep increased the likelihood of depression, GAD, and PTSD, even when other psychological disorders were simultaneously present. Additionally, higher levels of weekly exercise lowered the likelihood of most disorders, although this varied depending on exercise type. In most cases, the presence of a psychological disorder decreased the likelihood of beneficial health-related behaviors, including less physical activity participation and fewer hours of sleep each night, and increased the likelihood of adverse health behaviors, including smoking and greater alcohol intake. The presence of a single psychological disorder often increased the likelihood of psychological co-morbidity.

This was a complex study that sought to answer numerous questions. Results pertaining to each of the five questions will be explained briefly below. Additionally, there is the option to expand each section for a table detailing the study's findings.

1. Many of the health-related behaviors, including sleep, physical activity, and substance use were significant determinants of psychological disorders.

Factor		Outcomes of Interest
Moderate aerobic exercise		Participation in 150–299 minutes per week was associated with lower likelihood of: Depression Generalized anxiety disorder PTSD Participation in more than 300 minutes per week was associated with lower likelihood of: Generalized anxiety disorder PTSD
Vigorous aerobic exercise		Participation in more than 150 minutes per week was associated with lower likelihood of: Depression Generalized anxiety disorder PTSD
Strength Training		Participation 1 to 2 days per week was associated with lower likelihood of: Depression Generalized anxiety disorder
Sleep		Each additional hour of sleep each night was associated with lower likelihood of: Depression Generalized anxiety disorder PTSD
Alcohol Intake		Each additional drink per day was associated with higher likelihood of: Depression Generalized anxiety disorder PTSD
Current Smoking		Associated with higher likelihood of: Generalized anxiety disorder PTSD
Illicit Drug Use (in past year)		Associated with higher likelihood of: Depression Generalized anxiety disorder PTSD
Prescription Drug Misuse		Associated with higher likelihood of: Depression

2. Self-reporting any of the psychological disorders was also generally associated with higher likelihood of self-reporting at least one other psychological disorder.

Factor		Outcomes of Interest
Depression		Associated with higher likelihood of self-reporting co-morbid: Generalized anxiety disorder PTSD
Generalized anxiety disorder		Associated with higher likelihood of self-reporting co-morbid: Depression PTSD
PTSD		Associated with higher likelihood of self-reporting co-morbid: Depression Generalized anxiety disorder

3. The presence of additional psychological disorders often changed the relationship between health-related behaviors and the psychological disorder of interest described in question one analyses. In some cases, the beneficial effect of healthy behaviors, namely moderate and vigorous exercise, appeared to be lost. This suggests that, in cases of co-morbid psychological disorders, the influence of some health-related behaviors may be reduced. Notably, however, the beneficial influence of strength training one to two days each week and longer nightly sleep on mental health outcomes both remained, even after accounting for potential co-morbid psychological disorders.

Factor		Changes in outcomes of interest when co-morbid psychological disorders considered
Moderate Aerobic Exercise		Participation in 150 to 299 minutes per week was no longer associated with any psychological disorder Participation in more than 300 minutes per week was no longer associated with any psychological disorder
Vigorous Aerobic Exercise		Participation more than 150 minutes per week now only associated with lower likelihood of: Generalized anxiety disorder
Strength Training		Participation in strength training 1 to 2 days per week remained associated with lower odds for: Depression Generalized anxiety disorder
Sleep		Each additional hour of nightly sleep remained associated with lower likelihood of: Depression Generalized anxiety disorder PTSD
Alcohol Intake		Each additional drink per day now only associated with higher likelihood of: Depression
Current Smoking		Now only associated with higher likelihood of: Generalized anxiety disorder
Illicit Drug Use (in past year)		No longer associated with any psychological disorder
Prescription Drug Misuse		Remained associated with higher likelihood of: Depression

4. Generally speaking, all three psychological disorders predicted unfavorable health behaviors, including less participation in strength training, fewer hours of sleep, higher likelihood of smoking and obese BMI. These results have been combined with the results of question five, below.

5. Like the results for question three, some relationships between individual psychological disorders and health-related behaviors changed when co-morbid psychological disorders were examined. However, even when considering co-morbid psychological disorders, all disorders remained associated with fewer hours of nightly sleep. This suggests that poor sleep may be a common outcome of all three disorders. The results of questions four and five have been combined into a single table, which can be viewed by expanding the section below.

Factor	Outcomes of Interest	Outcome of interest when co-morbid psychological disorders considered
Depression	Associated with higher likelihood of: Low participation in moderate aerobic exercise Low participation in vigorous aerobic exercise Low participation in strength training Obesity Fewer hours of nightly sleep	Now associated with higher likelihood of: Low participation in strength training Obesity Fewer hours of nightly sleep
Generalized Anxiety Disorder	Associated with higher likelihood of: Low participation in moderate aerobic exercise Low participation in vigorous aerobic exercise Low participation in strength training Obesity Current smoking Fewer hours of nightly sleep Greater number of alcoholic drinks each day	Now associated with higher likelihood of: Obesity Current smoking Fewer hours of nightly sleep
PTSD	Associated with higher likelihood of: Low participation in moderate aerobic exercise Low participation in vigorous aerobic exercise Low participation in strength training Fewer hours of nightly sleep	Now associated with higher likelihood of: Fewer hours of sleep

These findings suggest healthier behaviors, including obtaining more sleep each night, greater participation in physical activity, avoidance of smoking, and drinking less alcohol may be beneficial for mental health. Of these behaviors, sleep was most consistently associated with better mental health, highlighting its importance in mental readiness. Unfortunately, sleep also appeared to be most consistently and negatively influenced by the presence of any of the three psychological disorders investigated. This bi-directional relationship demonstrates the importance of mental health treatment: untreated depression, GAD, or PTSD could give rise to sleep loss, which in turn might exacerbate the psychological disorder, resulting in a vicious cycle or escalating disorder. Exercise also emerged as a potentially beneficial factor, as both moderate aerobic exercise and strength training were often associated with better mental health outcomes. However, like sleep, exercise participation was often negatively influenced by the presence of psychological disorders, suggesting that individuals diagnosed with disorders such as depression, GAD, or PTSD may find themselves less inclined or capable of physical activity. The association of obesity as an outcome of depression and GAD may be related to the association between mental health and physical activity observed in this study. However, the researchers suggest the association between mental disorders and obesity may be complex and beyond the scope of this study, proposing that factors such as genetics may also play an important role. Nevertheless, exercise as means to improve mental health symptoms is supported by this study.

Together, these findings suggest healthy behaviors, including practice of good sleep, maintenance of healthy weight, and participation in weekly physical activity, may be important factors contributing to Service member mental health and well-being. Encouraging Service members diagnosed with mental health disorders to take part in these behaviors as a means of managing symptoms may be particularly important as they may be less likely to take part.

Most research studies have limitations or conditions in the study that were not ideal. While such limitations generally do not reduce the value of the study, it is always important to consider a study’s limitations and how they might have affected study results. Limitations of this study include use of older data, cross-sectional study design, use of self-reported participant information, the complex nature of psychological disorders, and low survey response rate. To see a more detailed explanation of each limitation and potential effects on study results, expand the section below.

Data for this study was collected in 2015. When using older information, it is possible that application to current populations may be limited. Various military and social factors have changed in the U.S. military since 2015, including differences in the demographic makeup of the Army and a change from a deployed U.S. military to a garrisoned military. Such differences may affect how study results translate to the current-day military population.
The cross-sectional nature of the study. Cross-sectional study design prevents determination of causality since it is impossible to know definitively whether one factor caused another. Since information was not collected at more than one timepoint, the definitive conclusion that health-related behaviors were present before a psychological disorder, or that health-related behaviors definitively prevented a disorder, cannot be made. While it is possible that participants practicing healthy behaviors prevented symptoms of depression, GAD, or PTSD, it is equally plausible that some participants may have started practicing healthy behaviors to manage mental health symptoms. This is called reverse causality. However, researchers recognized this possibility and included analysis to investigate bi-directional relationships as part of the research question, a strength of this study.
Health-related behaviors and psychological symptoms were all self-reported. With self-reported data, participants' recall of practicing health-related behaviors may have been inaccurate, which could lead to flawed study results. There is also the risk of “social desirability response bias" with self-reported health data. This type of bias occurs when participants provide answers based on what they feel may cast them in a better light, such as inaccurately reporting better sleep habits or fewer mental health symptoms. This could be especially true of this military population since healthier behaviors are generally expected of Service members as part of their occupational requirements.
Psychological disorders were based on self-reported symptoms. Because of the complex nature of psychological disorders, they are ideally diagnosed though interaction with a mental healthcare professional and not based solely on responses to a single questionnaire. Participants that screened positive for any of the psychological disorders may or may not have been diagnosed with the disorder when seen by a healthcare professional, which could affect study results. Furthermore, psychological disorders represent complex behavioral, social, physiological, psychological, and even genetic factors difficult to accurately capture via questionnaire. It is possible that the associations and interactions described in this study do not adequately address all factors, and relationships in this study may be more complex than what could be assessed with available information.
Low survey response rate. The study used information from a survey originally sent to many more potential participants than completed the survey; 201,990 Service members were mailed the survey, but only 16,699 responded. Furthermore, of the 16,699 participants, 3,991 participants did not answer all questions necessary for this study and their surveys were not used. In both cases, potential participants may have been worried that their responses would not truly be anonymous; they might have felt the questionnaire was too long or they may have had other reservations. These reasons for not completing the survey could have resulted in a systematic bias, meaning that some unknown factor might have made the people who did not answer the questionnaire different from the people who did answer the questionnaire. For example, less healthy Service members or those worried about stigma of mental health disorders may have been less likely to complete the survey. However, because it is impossible to know why nonrespondents did not complete the survey, it is impossible to know how these individuals compare with those who did respond. For this reason, it is possible that the study participants may not truly reflect the U.S. military.

To conclude, despite its limitations, this study provides strong support for the importance of healthy behaviors, including adequate sleep, physical activity, abstaining from smoking, and limiting alcohol intake, to help ensure Warfighter mental health and well-being. Moreover, when combined with professional mental health treatment, the practice of positive behaviors might also be beneficial for managing symptoms of depression, anxiety, and PTSD in Warfighters when and if they do arise.

Citation:

Hruby A, H Lieberman, and TJ Smith. 2021. “Symptoms of depression, anxiety, and post-traumatic stress disorder and their relationship to health-related behaviors in over 12,000 U.S. military personnel: Bi-directional associations." J Affect Disord 283:84–93.

Disclaimer: The mention of any non-federal entity and/or its products is for informational purposes only, and is not to be construed or interpreted, in any manner, as federal endorsement of that non-federal entity or its products.