Incidence and burden of endometriosis among U.S. active component service women, 2017-2024

The study population consisted of all ACSW ages 17–62 years in any branch of service of the U.S. Armed Forces, excluding the Coast Guard, from January 1, 2017 through December 31, 2024. Demographic, deployment, and inpatient and ambulatory care medical encounter records were obtained from the Defense Medical Surveillance System (DMSS); deployment data, only available through December 2022, were analyzed to determine numbers and lengths of deployments at any time prior to incident dates. Demographic variables included age, service branch, racial or ethnic group, rank, marital status, BMI, and occupation. BMI was obtained through the Defense Centers for Public Health–Portsmouth (DCPH-P) MHS Data Repository (MDR) and Periodic Health Assessments (PHAs). BMI records were excluded if height was less than or equal to 1 meter (m); height greater than or equal to 2.5 m; weight less than or equal to 20 kilograms (kg); weight greater than or equal to 180 kg; or obtained during pregnancy. The BMI record closest to the incident date was used for cases, while the BMI record closest to the start of a service record was used for the remaining population. International Classification of Diseases, 9th and 10th revisions, Clinical Modification (ICD-9-CM / ICD-10-CM) codes were used to determine endometriosis diagnoses and co-occurring gynecological conditions, including menorrhagia, PCOS, uterine fibroids, and infertility (Table 1) . In addition to ICD-9-CM / ICD-10-CM codes, Procedure Coding System (ICD-9-PCS / ICD-10-PCS) codes and Current Procedural Terminology (CPT) codes were used to identify prior parity and current contraceptive use (Table 1) . Description of Diagnostic Criteria for Outcomes of Interest for Endometriosis, U.S. Active Component Service Women, 2017–2024 Abbreviations: ICD-9-CM, International Classification of Diseases, 9th Revision, Clinical Modification; ICD-10-CM, International Classification of Diseases, 10th Revision, Clinical Modification; CPT, Current Procedural Terminology; ICD-9-PCS, International Classification of Diseases, 9th Revision, Procedure Coding System; ICD-10-PCS, International Classification of Diseases, 10th Revision, Procedure Coding System; PCOS, polycystic ovarian syndrome; IUD, intrauterine device. Indicates all codes following the parent code. Associated symptoms of uterine fibroids. Individuals can be diagnosed with uterine fibroids by a combination of association symptoms and uterine fibroid ICD codes. Parity was defined as a delivery-related code (Table 1) in any diagnostic position prior to the incident date. Delivery events were counted once every 280 days and recorded as a binary variable (‘yes’ or ‘no’) and as a categorical variable representing the number of births (0–3+). Current contraceptive use was defined as use of at least 1 contraceptive type: implant, injection, IUD, oral birth control, patch, vaginal ring, or miscellaneous type (i.e., unspecified or not already listed). Service women were counted once per category. Current use for long-acting contraceptives was determined within 5 years preceding the incident date for IUDs and within 3 years preceding the incident date for implants; all other contraceptive types were determined as current use within 12 months preceding the incident date. Pharmaceutical data were also utilized to analyze contraceptive use for implants, injections, IUDs, oral birth control, patches, and vaginal rings (Table 1) . A case of endometriosis was defined as an individual with 1 inpatient encounter with a case-defining code in any diagnostic position or 2 ambulatory encounters within 180 days with a case-defining code in any diagnostic position. 8 Individuals were counted as an incident case once per lifetime. 8 Menorrhagia was defined as an individual with 1 inpatient encounter with a case-defining code in the primary diagnostic position or 2 ambulatory encounters within 180-day period with a case-defining code in any diagnostic position. 19 Menorrhagia cases were counted once every 365 days. 19 PCOS was defined as an individual with 1 inpatient encounter with a case-defining code in the primary or secondary diagnostic position, or 2 ambulatory encounters in any diagnostic position. 20 Uterine fibroids were defined as 1 inpatient or ambulatory encounter with a case-defining code in the primary diagnostic position, or 1 inpatient or ambulatory encounter with a case-defining code in the secondary diagnostic position and at least 1 associated symptom (Table 1) in the primary diagnostic position. 21 Infertility was defined as 1 inpatient encounter with a case-defining code in the primary diagnostic position or 2 ambulatory encounters with a case-defining code in the primary or secondary diagnostic position. 22 PCOS and uterine fibroids counted once per lifetime, while infertility was counted once per surveillance period. 20 - 22 Crude IRs for demographic variables and case year were calculated per 10,000 p-yrs. Parity and deployment were stratified by count, while contraceptive use was categorized by type, to assess trends. Incident rate ratios (IRRs) and 95% confidence intervals (CIs) were then calculated for baseline characteristics. Prior parity and deployment may include person-time occurring outside the surveillance period; therefore, overall person-time was used to calculate crude IRs. IRRs were not calculated for prior characteristics, as using the overall person-time produced non-comparable rate contrast. To estimate the health care burden of endometriosis, medical encounters with a case-defining code in the primary diagnostic position were examined to evaluate the total numbers of medical encounters, individuals affected, and hospital bed days. All analyses were conducted using SAS ® Enterprise Guide ® software (version 8.3, SAS Inst., Inc., Cary, NC).

During the 8-year surveillance period, 5,733 ACSW, or 1.3% of all eligible service women during the period, were diagnosed with an incident case of endometriosis, at an overall rate of 32.8 cases per 10,000 p-yrs (Table 2) . Baseline Characteristics, Incident Endometriosis Diagnoses, U.S. Active Component Service Women, 2017–2024 Abbreviations: IRR, incidence rate ratio; CI, confidence interval; LL, lower limit; UL, upper limit; y , years; BMI, body mass index. Rate per 10,000 person-years. Overall, non-Hispanic Black women (IRR 1.1, 95% CI 1.1, 1.2) and women in health care occupations (IRR 1.7, 95% CI 1.5, 2.1) (Table 2) were more likely to be diagnosed with incident endometriosis than their counterparts. Additionally, women with a marital status of married (IRR 2.1, 95% CI 1.9, 2.2) or other (IRR 2.3, 95% CI 2.1, 2.5) were twice as likely to be diagnosed with incident endometriosis. Rates of incident endometriosis increased with age, with women ages 40 years or older demonstrating the highest IR overall (69.8 cases per 10,000 p-yrs) (Table 2) . This trend was generally observed throughout the surveillance period (data not shown). Overall, compared to ACSW of normal BMI, overweight women were 31.0% more likely to be diagnosed with incident endometriosis, while underweight women were 57.0% more likely, and obese women were 97.0% more likely to be diagnosed with endometriosis (Table 2) . Among women with incident endometriosis, 23.1% had co-occurring menorrhagia, while 21.1% had co-occurring infertility, 9.8% had co-occurring uterine fibroids, and 7.3 had co-occurring PCOS (data not shown). Endometriosis cases with no prior deployments had higher IRs (19.1 cases per 10,000 p-yrs) compared to women with prior deployments (13.7 cases per 10,000 p-yrs) (Table 3) . Among women with prior deployments, women with 1 deployment had the highest IR (6.3 cases per 10,000 p-yrs) (Table 3) . On average, prior deployments lasted approximately 6 months and occurred about 9 years prior to the incident endometriosis diagnosis (Table 3) . Additionally, nulliparous women had a higher IR (22.3 cases per 10,000 p-yrs) compared to uniparous and multiparous women (10.5 cases per 10,000 p-yrs) (Table 3) . With each delivery, IRs of endometriosis decreased (Table 3) . Prior Deployment and Parity, Incident Endometriosis Diagnoses, U.S. Active Component Service Women, 2017–2024 Average duration of deployment: 178 days Average duration from deployment to incident endometriosis diagnosis: 8.9 years Abbreviation: No., number; n , number. Rate per 10,000 person-years. Among women with incident endometriosis, 24.0% were not currently using any form of contraceptive, while 76.0% were currently using some form of contraceptive (Figure 2) . Oral birth control was the most common (23.1%) type of contraceptive used by ACSW, followed by miscellaneous type (22.2%) and IUDs (16.7%) (Figure 2) . Incidence of Endometriosis, U.S. Active Component Service Women, 2017–2024 Percentage of Current Contraceptive Use, Incident Endometriosis Diagnoses, U.S. Active Component Service Women, 2017–2024 Figure 3 presents the burden of endometriosis among ACSW. The majority of 2017–2024 medical encounters for endometriosis were ambulatory care encounters (Figure 3) . The number of medical encounters remained relatively stable before 2021, when a continued annual increase began (Figure 3) . The number of individuals with medical encounters decreased in 2018, but since 2020 counts have increased, along with medical encounters (Figure 3) . The number of hospital bed days was at its lowest in 2021, while only 2 years later, in 2023, the highest number of hospital bed days was recorded. Counts more than doubled in 2023 compared to the previous year, but this is attributed to a small number of individuals rather than a reflection of the entire population (Figure 3) . Burden of Endometriosis, U.S. Active Component Service Women, 2017–2024

This study analyzed incidence of diagnosis rates of endometriosis, and this report describes the distributions of prior deployment, parity, and BMI on IRs. Co-occurring gynecological conditions, current contraceptive use, and health care burden were examined as well. Compared to the prior MSMR report on endometriosis, 8 overall crude incidence of endometriosis has increased from 30.8 cases per 10,000 p-yrs during 2012-2016 to 32.8 cases per 10,000 p-yrs in 2017-2024. During the surveillance period, IRs increased nearly 42.0% from 2017 to 2024. The current findings suggest an upward trend of newly diagnosed endometriosis among ACSW. When compared to civilian women, ACSW have greater accessibility to health care and diagnosis, provided by MHS, which may explain the increase in endometriosis diagnosis, rather than reflect a true increase in cases. Consistent with the prior reporting of endometriosis among ACSW, 8 service women who were older, non-Hispanic Black race or ethnicity, and in health care occupations had higher rates of endometriosis; similar findings for age at diagnosis were reported for the general population. 11 Civilian women ages 36-45 years were found to have higher IRs of endometriosis. 11 Delayed diagnosis from symptom onset could explain why IRs are higher among women ages 35 years or older compared to younger service women. Differences between racial and ethnic groups appear to be unique to the military population, when compared to the general population. Several studies have found lower incident endometriosis among non-Hispanic Black women compared to non-Hispanic White women, or no significant difference at all. 11 , 23 , 24 Disparities in civilian health care and need for surgical diagnosis may explain differences among military and civilian rates. 23 Further exploration of higher IRs of endometriosis among non-Hispanic Black women compared to civilian populations may be warranted. Service women with no deployment history prior to endometriosis diagnosis had higher IRs of endometriosis compared to those with prior deployments, while women with 1 deployment had higher IRs than women with multiple deployments. These findings are somewhat unexpected, given the epidemiological associations between deployment and adverse reproductive and mental health outcomes. 9 , 15 The IRs are crude, however, and therefore the observed difference may be due to unadjusted confounding variables rather than a true statistical difference in risk. Combat trauma could negatively influence mental health, leading to riskier sexual behaviors and avoidance of reproductive health care 11 ; these factors lead to poorer reproductive health outcomes. 13 Determining the effects of deployment on endometriosis are difficult, however, due to the timing of disease onset and disease diagnosis. In this study, deployment occurred approximately 9 years, on average, prior to endometriosis diagnosis. Additionally, women with endometriosis demonstrate poor physical performance compared to women without the condition, 25 which may disqualify women from deployment, possibly inferring the ‘healthy warrior effect’, with healthier ACSW more likely to deploy. Furthermore, less than 4% of incident cases deployed following diagnosis (data not shown). These findings suggest that endometriosis diagnosis may inhibit deployment of ACSW and require greater medical management of ACSW to maintain force readiness. BMI is reported to have an inverse relationship with endometriosis, 2 , 6 and in this study underweight women did have overall higher crude IRs than overweight women; this finding was not evidenced throughout the surveillance period, however (data not shown). Obese service women were observed to have the highest incidence of endometriosis overall, and throughout the study. Previous research found that this inverse relationship of BMI and endometriosis was not evident among women ages 30 years or older. 26 When this study examined BMI by age group, underweight ACSW ages 25-34 years had the highest IRs, while in all other age groups, obese ACSW had the highest rates (data not shown). These findings support the previous research, 26 suggesting a greater association of underweight BMI with younger ACSW and endometriosis diagnosis. Lower parity was associated with higher rates of endometriosis. Given the association between infertility and endometriosis, 17 these findings may be unsurprising, however. In this population, 21.1% of women with incident endometriosis also had co-occurring infertility. Additionally, the majority of endometriosis cases in this population were currently using contraceptives prior to diagnosis. Pregnancy prevention is not the only indication for contraceptive use, as oral contraceptives are a primary or ‘first-line’ treatment for endometriosis and endometriosis-associated symptoms. 27 Oral birth control was the most common contraceptive type among women in this population. Several limitations are important to consider when interpreting these findings. The delay in obtaining endometriosis diagnosis 5 creates a challenge for determining disease onset and how reproductive health, demographic, and service-related factors affect the condition. The IRs for prior parity and prior deployment were calculated using overall person-time due to an inability to calculate person-time prior to diagnosis for the population at risk, as not all ACSW at risk in the population were diagnosed with endometriosis. Additionally, the ‘healthy warrior effect’ may explain higher IRs among ACSW with no deployment history compared to those with deployment history. The crude IRs for prior parity and prior deployment should be evaluated as preliminary and with caution. Lack of a standardized case definition for endometriosis in the literature presents comparison challenges. 3 , 4 Additionally, the difficulty of obtaining a diagnosis may obscure true incidence. Furthermore, results related to parity and contraceptive use are observational. Reasons why women are nulliparous or contraceptive users are numerous and unknown in this study. Finally, deployment data were only available through 2022, under-estimating rates of prior deployment among ACSW. Endometriosis is associated with a multitude of symptoms 2 that can affect military readiness and quality of life. Future studies should evaluate endometriosis severity to understand its effects on force readiness and health care provision. A more comprehensive cohort study of symptomology, mental health, deployment, and demographics, from accession to end-of-service contract, may better explain the effects of military service on endometriosis.