Insomnia: Sex differences and age of onset in active duty Army soldiers

Abstract

Background:

The prevalence of insomnia symptoms, insomnia diagnostic status, and age of onset compared by sex is understudied within the military population.

Method:

Data were examined from the All Army Study to Assess Risk and Resilience in Service members (N = 21,294; 18-61 years; 11.7% female and 87.6% male). Participants were given a self-administered version of the Composite International Diagnostic Interview Screening Scales to assess psychopathology and cognitive deficits, and the Brief Insomnia Questionnaire for insomnia disorder assessment. Participants identified the age they first experienced sleep problems for 1 month or longer as part of the self-administered questionnaire.

Results:

Among this sample of Army soldiers, 22.8% met insomnia diagnostic status (22.0% of males and 28.4% of females). A binary logistic regression model revealed that insomnia diagnostic status was associated with female soldiers (OR = 1.26, P< .001, 95% CI = 1.13-1.41) compared to male soldiers, even when accounting for sociodemographic variables and mental health disorders. No significant sex differences emerged at insomnia symptom level or sleep problem age of onset.

Conclusion:

This study highlights the importance of examining insomnia by sex in active-duty populations. Results suggest that female active duty soldiers experience insomnia at a higher rate than their male counterparts, which may be driven by reports of daytime impairment. Given this information, more specific clinical recommendations on assessing and treating insomnia could be provided, especially when considering readiness for military duty.

Introduction

While 10%-15% of the general adult population will experience chronic insomnia at some point in their lifetime,¹ the prevalence of insomnia is even higher in particular segments of the population. Two groups with a greater degree of sleep disturbance are women and military service members. A meta-analysis by Zhang and Wing found that women are 1.4 times more likely than men to develop insomnia across all ages, worldwide.² Military-service status may compound this risk of insomnia, although data on sex differences in this population are mixed and sparse.

In a clinical sample of military personnel who returned from deployment and were referred for polysomnography, Mysliwiec et al. (2013) found from their review of 761 military electronic medical records that insomnia was the second most common primary sleep disorder and female soldiers were more likely than male soldiers to have insomnia.³ In an additional clinical sample of military personnel using review of sleep medicine evaluations and in-laboratory sleep studies, Foster et al. (2017) found that, among a cohort of 209 individuals, 72 females, as compared to 41 males (P < .001), met criteria for insomnia after adjusting for age and body weight, with women also exhibiting greater likelihood of depression, pain, and anxiety diagnoses.⁴ Due to both sleep disorders and greater behavioral health comorbidities, the authors suggest that female military personnel may have greater disease severity.

In contrast to the above findings, a predeployment assessment of insomnia and other psychological variables in a large nonclinical sample of active-duty Army soldiers (N = 4,101) found that approximately 20% of the sample had insomnia, but there were no significant sex differences in insomnia status.⁵ The difference from civilian samples may be driven by women soldiers being more physically healthy compared to their civilian counterparts and reporting physical symptoms equal to those of military men.⁶ It could be speculated that this discrepancy across studies with military samples also may be associated with stages of military career and timing of deployment, in that women prior to deployment may be at lower risk for insomnia than those following deployment. Here, the results continue to be mixed. Utilizing the All Army Study, which includes representative Army-wide samples of soldiers across different service phases, Klingaman et al. (2018) explored rates and correlates of insomnia. The authors found that insomnia is prevalent in this military population (22.76%) and, in addition to other identified risk factors, there were significant, yet small, sex differences.⁷ Caldwell et al.⁸ utilized a population of US Army soldiers from 1997 to 2011 to complete a retrospective cohort investigation, studying the relationship between sleep disorders and deployment. Findings indicated deployment increased risk of insomnia; however, the authors did not find differential effects of deployment on insomnia when comparing females to males. More recently, a prospective, longitudinal investigation of insomnia patterns across deployment in Army soldiers revealed that women soldiers were more likely to belong to the incident insomnia group, those who developed probable insomnia during deployment or soon after that did not immediately remit upon return, compared to those who never developed insomnia, or only developed it during deployment.⁹ Overall, more data could be helpful to examine sex differences from additional representative, nonclinical military samples and at varying stages of service, in order to potentially better understand future needs of this vulnerable population.

The present study aimed to explore these sex differences in insomnia prevalence among active-duty soldiers in more detail, by examining sex differences in both insomnia diagnostic status and at symptom level in this representative sample of Army soldiers. Given that mental health disorders are importantly recognized factors associated with insomnia,¹⁰ this study also sought to examine sex differences in insomnia rates, independent of the presence of these conditions. Finally, as there may be significant delays in help-seeking for psychological conditions, including insomnia,¹¹ we sought to understand whether there were sex differences in the age of insomnia onset, which could serve as a proxy of the underlying vulnerability to experience insomnia, as well as chronicity of the symptoms among these soldiers. If there are significant sex differences in age of onset, it may reflect differences in underlying vulnerability and/or a more chronic course of insomnia. Early interventions could enhance military readiness and the ability to successfully perform military operations. Together, the findings from the present study could contribute to identifying soldiers who would benefit from allocation to sleep interventions (eg, cognitive behavioral therapy for insomnia; CBT-I) or sleep-resiliency training prior to deployment.

Participants and methods

Participants of the All Army Study

The All Army Study (AAS) of the Army Study to Assess Risk and Resilience in Service Members (STARRS) served as the data source for these analyses. Study procedures and methodology for the AAS are outlined by Kessler et al.¹² In brief, the AAS was conducted between 2011 and 2013 as a self-administered questionnaire (SAQ) completed by established active duty US Army soldiers, recruited in quarterly samples. Participants of the study were asked to provide informed consent to: complete the SAQ, allow connection of their SAQ responses to the associated and appropriate Army and Department of Defense administrative records, and finally, to be contacted in the future for potential data collection. Study procedures were approved by the Human Subjects Committees of all collaborating organizations.

Measures

For the context of this study, the DSM-5 criteria were applied in combination with the Brief Insomnia Questionnaire (BIQ)¹³ measure to ascertain current insomnia disorder. The BIQ measures frequency of insomnia symptoms within the past 30 days (e.g., how many nights did it take more than 30 minutes to fall asleep at night [sleep onset latency], waking up 3 or more times during a single night [frequent night awakenings], waking up at night and taking more than 30 minutes to go back to sleep [wake after sleep onset], and waking up more than 30 minutes too early in the morning [early morning awakening]), along with the impact insomnia symptoms had on daytime functioning (e.g., daytime fatigue, somatic symptoms, mood, and impaired performance at work/school), utilizing a Likert scale (1 = extremely, 2 = a lot, 3 = some, 4 = a little, 5 = not at all). Current insomnia disorder diagnosis was based on the following criteria: (1) reported insomnia symptoms 3 or more nights per week (individual sleep items were dichotomized into <3 nights per week or >=3 nights per week); (2) sleep problems that interfered with functioning at least “a lot”; and (3) endorsed sleep problems for at least 3 months within the past year. To identify the approximate age of onset of a first insomnia episode, participants also were asked to identify the age when they first experienced sleep problems for 1 month or longer.

Participants also completed a self-administered, computerized version of the Composite International Diagnostic Interview Screening Scales (CIDI-SC), as well as a screening version of the PTSD Checklist (PCL).^14,15 Current (past 30 days) presence of 4 internalizing disorders were examined and explored for the context of this study: major depressive episode (MDE), posttraumatic stress disorder (PTSD), panic disorder (PD), and generalized anxiety disorder (GAD), along with one externalizing disorder: substance abuse or dependence (SUB/D). Diagnoses of the aforementioned disorders were determined in absence of DSM-IV diagnostic hierarchy or organic exclusion rules. These diagnoses also were made without the use of any sleep-related items (eg, insomnia symptoms, nightmares), reducing potential construct overlap with the insomnia diagnosis. Previous validation studies have found high concordance between diagnosis based on the CIDI-SC combined with modified PCL diagnosis and structured clinical interview.^14,16

Data procedures and analysis

All analyses were conducted using IBM SPSS, version 24. First, descriptive statistics were explored in order to find the prevalence rates of insomnia and psychiatric disorders representative in the total sample. Additionally, age of onset of insomnia symptoms lasting for 1 month or longer was documented categorically within the data set (age 17 and younger; age 18 and older). Odds ratios were calculated using binary logistic regressions, with sex as the dependent variable for all factors to (1) understand the relationship between current insomnia diagnosis and sex when including the presence of sociodemographic variables (age, ethnicity, and race), 4 internalizing disorders (MDE, PTSD, PD, and GAD) and one externalizing disorder (SUB/D); (2) determine the relationship between specific insomnia symptomatology and sex, and (3) determine the relationship between reported age of onset (age 17 and younger vs age 18 and older) and sex.

Over 20,000 soldiers (N = 21,449) completed the survey, of which 99.3% (n = 21,294) had some data available for the analyses in the present sample. Of note, the sample sizes vary across analyses because some questions were only given to individuals who endorsed a prior question in the survey, or the sample accounts for a differing number of individuals who did not respond to each item. The analytic sample for exploring sex differences of insomnia symptom and diagnostic status included 20,572 soldiers. Regarding age of onset analyses, only soldiers reporting sleep problems that interfered in their daytime functioning at least some in the past 30 days and had complete data on the other variables were used (n = 9,314).

Results

Participants of this sample were military soldiers between the ages of 18 and 61 years (M = 28.7, SD = 7.4). A majority of the sample were men (n = 18,790, 88%) and identified as White (n = 14,909, 69.5%). See Table 1 for additional sample characteristics.

Table 1.

Sample characteristics by sex.

	Total N = 21,294	Males n = 18,790	Females n = 2504
	M (SD) or%	M (SD) or%	M (SD) or%
Sociodemographic
Age in years	28.7 (7.4)	28.7 (7.4)	28.3 (7.4)
Race
Hispanic	15.2%	14.9%	17.6%
Ethnicity
White	69.5%	72.5%	58.1%
Black	16.6%	15.3%	29.4%
Asian	4.1%	4.1%	4.4%
American Indian/Alaskan Native	2.9%	2.7%	4.2%
Pacific Islander	1.3%	1.3%	2.0%
Other	8.4%	8.4%	9.5%
Marital status
Married	56.9%	58.7%	43.7%
Never married	30.9%	30.3%	34.9%
Divorced, separated, or widowed	12.2%	11.0%	21.4%
Education (≥ associate degree)	28.2%	52.9%	37.9%
Lifetime mental health factors
PTSD diagnosis	7.2%	6.8%	9.9%
MDE diagnosis	6.8%	6.5%	9.9%
GAD diagnosis	5.8%	5.4%	8.8%
SUB/D diagnosis	4.3%	4.5%	3.0%
PD diagnosis	3.5%	3.3%	5.4%

Note. GAD, generalized anxiety disorder; MDE, major depressive episode; PD, panic disorder; PTSD, posttraumatic stress disorder; SUB/D, substance use disorder, n’s and percentages may vary across items due to non-response on those items.

Insomnia disorder and symptoms by sex

Approximately 22.8% soldiers (22.0% of males and 28.4% of females) met the selected insomnia criteria. With a model that included sociodemographic variables and the mental health conditions, insomnia disorder was significantly associated with female soldiers (OR = 1.26, CI = 1.13-1.41, P < .001). No statistically significant sex differences emerged at the insomnia symptom level (sleep onset latency, frequent night awakenings, wake after sleep onset, and early morning awakenings; ORs = 1.00).

Age of onset of insomnia by sex

Subjects reported age of onset before age 18 14.6% of the time, with 85.4% indicated adult onset. With a model that included socio-demographic variables and the mental health conditions, no statistically significant sex differences emerged for age of sleep problem onset (P = .55). Findings indicated that the most common reported age of onset of insomnia symptoms for military soldiers was the age of 18 or older, regardless of sex.

Discussion

In light of Klingaman et al.’s⁷ results that suggest insomnia rates are common in the military population, this study aimed to explore in more detail the prevalence of current insomnia disorder and symptoms by sex in active duty Army soldiers, beyond the impact of co-occurring psychiatric disorders. Findings indicate that female soldiers endorsed past month insomnia disorder at a significantly higher rate than males. These results were maintained after considering the influence of comorbid conditions, indicating that the greater prevalence of insomnia disorder in females is not only secondary to higher rates of other disorders. The present study also intended to expand Klingaman et al.⁷ previous findings, by exploring potential sex differences at the insomnia symptom level and reported age of onset of these symptoms. No statistically significant sex differences emerged in these domains.

Our results corroborate previous findings of sex differences in insomnia diagnostic status that cannot be attributed solely to mental health conditions. While not specifically explored in this study, it is worthwhile to note the adjacent comparison of sex and insomnia for military soldiers vs the civilian population, in order to understand potential implications. Findings indicate that women, in general, tend to have higher prevalence rates. The American Insomnia Survey reported 27% insomnia prevalence in civilian women vs 19% prevalence in civilian men (7,428 participants).¹ Focusing directly on the female population, Rissling et al.¹⁷ also demonstrated that women veterans have similar insomnia prevalence rates as compared to civilian women (30.5% vs 30.8%, respectively). Finally, most related to this study, Foster et al.⁴ utilized a similar active duty population; albeit smaller sample; they found that 72 females, as compared to 41 males (P < .001) met criteria for insomnia. Of this population, females exhibited more symptoms related to depression and anxiety diagnoses.

Interestingly, when examining sex differences at the insomnia symptom level, there were no differences in the present study. This finding suggests that male and female active-duty soldiers may experience insomnia symptoms at a similar frequency and chronicity, as there were no age of onset differences; however, their perception of the sleep difficulties interfering on daytime functioning, a criterion for insomnia diagnostic status, may not be shared. A previous study examining perpetuating factors and experiences in a civilian sample with chronic insomnia found that, compared to men, women reported greater perception of insomnia severity and reported higher levels of daytime fatigue.¹⁸ Women often have competing demands (e.g., work and gender-based roles) that may make sleep problems more noticeable, and these authors also noted that socio-cultural perspectives that encourage women to express emotional and bodily distress could account for the difference. While not explicitly assessed, similar factors could be contributing to the differences seen in the present study. Male soldiers may not experience as much of a daytime impairment or may be less willing to endorse that their duties are being impacted by sleep disruption compared to their female counterparts. Additionally, female soldiers may have additional sleep disruption, such as sleep apnea, that leads to the difference in daytime impairment. One study indicates that female service members may have a high rate of obstructive sleep apnea (OSA) and those with a combined insomnia and OSA diagnosis had greater prevalence of additional physical and mental health comorbidities.¹⁹ Overall, it is important to note that this study’s findings suggest that discrepancies in sex differences between insomnia disorder vs insomnia symptoms can be accounted for by the fact that the insomnia disorder variable is a composite of the individual symptoms. So while there may not be significant sex differences in sleep latency, nocturnal wakefulness, or early morning awakenings, there differences could remain in whether individuals have any 1 of 3 symptoms.

These results point to important assessment and treatment implications. A comprehensive evaluation, that includes subjective and objective sleep and performance measurements, may be needed to identify those most in need of treatment, regardless of sex. Male soldiers may not seek help for their sleep disturbance, even when chronic, and might not receive treatment when needed. Targeted education on how sleep disturbances may indeed impact daily duties and long-term military readiness might be particularly beneficial for male soldiers. Additionally, female soldiers’ health and sleep experiences should not be ignored. These results point to the importance of considering how insomnia may affect female soldiers differently than male soldiers. An integrated sex-specific approach is needed to investigate variables of reproductive health, hormonal changes across the lifespan, and environmental factors that are contributing to the observed sex-differences.²⁰

Limitations

While this study is strengthened by its large sample of soldiers, several study limitations should be considered. Due to the cross-sectional and self-report nature of the assessment, this study may be biased by participant recall. Additionally, biological differences between men and women that may influence the presence of insomnia symptoms were not assessed. For examples, higher prevalence of insomnia in women is suggested to relate to menopause, late-luteal phase of the menstrual cycle, and third trimester of pregnancy, according to Krystal et al.²¹; all variables were not assessed in this study. This study also did not account for the impact of relationship status, parental identity, separation from family, and type of trauma experienced by the soldiers at the time of assessment. Similarly, given the variability within the military environment, the unaccounted contributions of the soldiers’ current sleeping conditions cannot be ignored. The age of onset question referred to sleep problems broadly; therefore, we cannot ascertain that responses were only about insomnia. Lastly, the presence of other sleep-disorder diagnoses, whether these symptoms were treated, as well as any potential impact due to comorbid physical health comorbidities, such as various cardiovascular, respiratory, gastrointestinal, urinary, neurological, pain, and diabetes, were not known. Future research should emphasize longitudinal measurement and consider the influence of these variables. This information could help determine the relationship between these aforementioned variables and sleep quality throughout a lifespan. The findings of such studies could potentially aid in more nuanced prevention and treatment of insomnia symptomatology and physical health comorbidities over time.

Conclusions

Overall, this study corroborates past research that identifies a high prevalence rate of insomnia in military soldiers. Among a sample of active-duty Army soldiers at varying phases of their service, the results found that insomnia diagnostic status was significantly associated with female soldiers compared to male soldiers. It is suggested that future research include longitudinal studies in order to track insomnia symptoms in male and female soldiers over time, with and without the presence of comorbid psychiatric disorders. Moreover, further investigating report of daytime functioning could be helpful to ascertain how these symptoms influence insomnia severity. Subsequently it could be helpful to explore barriers to treatment for female and male soldiers that experience insomnia symptoms, such as exploring the differences in sex, in relationship to help seeking behaviors. Additionally, researchers could potentially expand this work by exploring if the number of deployments experienced by female and male military soldiers influences the severity of insomnia symptoms. Next, expanding on the literature regarding the influence of biological differences between males and females, across a lifespan, on the progression of insomnia symptomatology over time, could be helpful to those treating sleep complaints. Lastly, given the results of this study and past similar research, it is important to support and develop treatment and research programs specifically targeted for women with sleep difficulty, with and without comorbid mental health conditions such as anxiety and depression, and to make available resources explicitly directed toward soldiers with sleep complaints. Overall, this study hopes to emphasize the need for women’s health advocacy, to increase awareness for elevated prevalence of insomnia in female soldiers, and to encourage more longitudinal studies in order to further examine properties associated with age of onset of insomnia symptoms.