Menstrual and Oral Contraceptive Use Patterns Among Deployed Military Women by Race and Ethnicity

Abstract

Menstrual cycle patterns and concerns and oral contraceptive use in the combat environment were examined in Caucasian, Asian, Hispanic, and African American women to guide the development of educational resources for women soldiers. An anonymous, questionnaire was completed by 455 U.S. Army women—Caucasian (CA: n = 141); Asian (AS: n = 67); Hispanic (HIS: n = 67); and African American (AA: n = 184) to compare menstrual patterns and concerns, dysmenorrhea, and oral contraceptive patterns. Total menstrual concerns were significantly lower among African Americans relative to Caucasians, Asians, or Hispanics; Asians and Hispanics reported the greatest concern. Overall, secondary amenorrhea was noted by 14.9% of women. Severe dysmenorrhea rates were significantly lower in African American (6.1%) compared to Caucasian (11.6%), Asian (20.9%) and Hispanic (19.7%) women. Asian women reported missing less work—only 9.3% with moderate to severe dysmenorrhea missed work compared to 25.1% of all other women. Only 9.2% of women with mild, compared to 25.8% with moderate to severe (OR = 3.44; p ≤ 0.0001) dysmenorrhea sought health care. Less than 50% of women took oral contraceptive, and less than half of those women took oral contraceptive continuously. African Americans seemed to experience menstruation as less bothersome than others, despite no difference in the proportion with menstrual irregularities, mean duration of menses, and/or mean time between cycles.

KEYWORDS:

• dysmenorrhea
• breast tenderness
• health care
• missed work
• oral contraceptive agents
• race/ethnicity

INTRODUCTION

Historically, menstruation was perceived as a limiting factor for women in many occupational endeavors and cited as the source of women's inferiority as workers (Harlow, 1986). Although it has been made clear now that menstruation does not negatively affect performance (Whitehead, 1934; Grow & Armstrong, 1942; Gamberale, 1985; Kishali et al., 2006), it did and remains a logistical consideration during military deployments. Menstrual disturbances, in particular dysmenorrhea, have been reported in adolescent females, college women, and other women of childbearing age, with rates ranging from 40% to 83% (Banikarim, Chacko, & Kelder, 2000; O'Connell, Davis, & Westhoff, 2006; Chang and Chen, 2009). The severity of these menstrual disturbances has led some adolescents and women to seek medical assistance and/or miss time from school or work (Banikarim, Chacko, & Kelder, 2000; Dell, 2004; Wong & Khoo, 2009). The introduction of prescription oral contraceptives (OC) for birth control freed many women from the constraints of menstruation, and reduced the number of menstrual-related complaints (Edelman et al., 2005; O'Connell, Davis, & Westhoff, 2006; Sillem et al., 2003). Moreover, OC use re-established menstrual regularity in many women chronically exposed to environmental stressors (Preston, 1978).

Only a few studies have reported on menstrual patterns and dysmenorrhea as a function of race/ethnicity (Wong and Khoo, 2009; Gordley et al., 2000; Janiger, Riffenburgh, & Kersh, 1972; Harlow & Campbell, 1996; Harlow et al., 1997; Gold et al., 2007). Wong and Khoo (2009) noted ethnic differences in the prevalence, impact, and management of dysmenorrhea in a multiethnic population of adolescent Asian girls, whereas Janiger, Riffenburgh, and Kersh (1972) compared menstrual symptoms in women from various countries. Only one group has compared multiple ethnicities within an American population (Gold et al., 2007), who reported significant racial/ethnic differences in menstrual symptoms.

This study examined differences in menstrual patterns and the proportion with dysmenorrhea in a population of racially and ethnically diverse, military women who had been deployed in the austere and resource limited, inconvenient, and logistically challenging overseas combat environment. Specifically, menstrual cycle patterns and concerns in the combat environment, including lost duty days, time missed for medical evaluation, and patterns of OC use were assessed in a sample of Caucasian, Asian, Hispanic, and African American women. The results were intended to identify unmet educational needs and guide the development of educational resources for our women soldiers.

MATERIALS AND METHODS

Participants and Recruitment

The Tripler Army Medical Center's (TAMC) Institutional Review Board (IRB) through the Department of Clinical Investigation approved the study. Active Duty, Reserve, and National Guard women between the ages of 18 and 45 years who had returned from overseas combat operations within the last 30 days were eligible. Participants were recruited from the 25th Infantry Division's post-deployment health assessment center via a poster, which described the study, and by family physicians, who were seated at tables in the center. Women were able to pick up a questionnaire from the table or receive it from a third party, independent of their chain of command. They completed it while waiting for rotations through the other stations. Recruitment occurred between January 2005 and April 2005 and in October 2007 over the course of 20 days. Women did not sign an informed consent document as the IRB agreed that voluntary completion of the anonymous questionnaire implied consent and was sufficient.

Recruitment days were predetermined based on projected high attendance at the Post-Deployment Health Assessment Center. A total of 508 questionnaires were collected, eight of which were rejected due to age exclusion; another 12 were excluded from analyses of menstrual concerns due to use of depo-medroxy progesterone by the respondent, a potential confounder. Because the study sought to focus on racial/ethnic differences, another 33 questionnaires were omitted from the analysis because of the small sample size of the racial/ethnic classification: 10 Native Americans and 23 other racial/ethnic groups. The daily response/participation rates of women who came to the center and were asked to complete the questionnaire varied from 86% to 100%. Approximately 2,289 females served in the Division during 2007; thus the sample of 508 represents 22% of all military women who had been deployed during the time periods.

Questionnaire Development

The questionnaire contained 44 questions: seven pertained to demographic questions, 22 to menstrual history, patterns, and concerns, and 15 to OC knowledge, attitudes, experiences, and compliance (Appendix 1). The demographic, menstrual history, and OC knowledge and attitudes questions had been used in a previous study (Powell–Dunford et al., 2003). Menstrual concern was assessed through nine questions specific to menstrual symptoms and experiences during deployment: they were formatted as a 10-point Likert-type scale from least to greatest degree of symptomatology. Six of the nine questions were derived from the Menstrual Distress Questionnaire (Whitehead, 1934; Grow & Armstrong, 1942), an instrument that has been substantiated as internally consistent and reliable, and widely used for measuring menstrual related symptoms (Whitehead, 1934; Grow & Armstrong, 1942). This study targeted five menstrual-related symptoms: nausea, cramping, bloating, headache, and breast tenderness. In addition, participants were asked about exercising or performing strenuous physical labor while menstruating/wearing sanitary products, unexpected menstrual cycles, unexpected spotting and if they were bothered by having to change sanitary products during deployment. Strenuous physical labor was defined as exerting at near maximal effort with heavy breathing. Each participant was asked to rate how often she was bothered during the deployment by the various menstrual symptoms and activities on a scale of 1 to 10, with 1 being “never” and 10 being “very often.” Because psychosomatic factors, such as anxiety and tension, were potentially confounded by combat experience, they were not incorporated into the questionnaire as menstrual symptoms.

Terminology and Classification

Total menstrual concern was calculated by summing the responses to the nine “perceived bother” questions to achieve a single estimate of menstrual concern (highest possible score was 90). Individual and total concern scores were then categorized depending on how they rated perceived bother: low (1–3), moderate (4–7), and high (8–10). Likewise, a score for dysmenorrhea was created by summing the scores reported for cramps, headache, bloating, nausea, and breast tenderness. A woman whose total score was between 16–35 (highest possible = 50) was defined as moderate and a score of 36 or greater was considered severe dysmenorrhea. Also, a woman was classified as a conventional OC user if she took three weeks of hormonally active pills before a week of placebo pills to induce menstruation whereas a woman taking hormonally active OCs for three or more months in a row was categorized as a continuous OC user. A compliant OC user reported missing “no pills per week on average,” and a non-compliant OC user was someone who “missed at least 1 pill per week on average.”

Data Analysis

Descriptive statistics were used to summarize demographic data. Multivariate analysis of variance was used to determine differences in reporting of concerns by the racial/ethnic groups after adjusting for potential confounders (age, rank, OC use, length of deployment, and frequency of strenuous duties), and then non-parametric Kruskal Wallis tests were used to examine differences for each specific concern. When statistical significance was noted, a Bonferroni correction was used for post hoc analysis of racial/ethnic differences. Chi Square was used to test differences after individual results were re-categorized into two or three groups as described above under classification. Cronbach's alpha and principal component analysis were used to assess the reliability of the questionnaire. The Cronbach's alpha for the nine-item concern scale was 0.916 for the study population. A principal component analysis revealed a single factor solution that accounted for 60% of the test score variation. Each menstrual concern item correlated significantly with the total concern score (0.63 < r < 0.82).

RESULTS

The average age of participants was 26.7 ± 5.9 years, but the Asian (AS) women were the oldest (34.8% ≤ 25 yrs) and Hispanic (HIS) women the youngest (76.9% ≤ 25 years) among the four racial/ethnic groups (Table 1). The primary deployments were to Iraq (45%) and Afghanistan (53%), but 2% of respondents reported being deployed to other areas of operation; one African American (AA) woman had been deployed to both Afghanistan and Iraq. The mean length of deployment was 13.1 ± 2.0 months, with 98% deployed between 12 and 15 months; the deployment location conditions were likely very similar, as both were rugged, remote locations with similar hardships. As expected, the majority of the women were enlisted. In terms of strenuous physical labor, over 75% of all women reported engaging in strenuous activity three or more days per week.

TABLE 1 General Characteristics of Women by Race/Ethnicity

	Caucasian (n = 141)	Asian (n = 67)	Hispanic (n = 67)	African American (n = 184)
Racial/Ethnic Distribution (%)	30.9	14.6	14.6	40.4
Age (yrs), Mean (SD)	26.7 (± 0.5)	28.4 (± 0.7)*	24.4 (± 0.6)	26.8 (± 0.5)
Enlisted (%)	92.3%	91.9%	94.0%	75.0%*
Deployed ≥ 12 months (%)	96.5	97.0	97.0	97.8
Deployed to Iraq (%)	55.6	67.2*	46.3	46.2
Deployed to Afghanistan (%)	41.0	25.4*	57.2	48.9
Strenuous Duties ≥ 5 d/wk (%)	28.8	51.6*	38.8	36.4

Menstrual Patterns

No significant differences were noted by Chi Square analyses across racial/ethnic groups in the proportion reporting different menstrual patterns (Table 2). However, a trend toward fewer days of bleeding for HIS and AA and greater pain for HIS women was noted. Two definitions were used to assess the prevalence of amenorrhea in women not taking OC during deployments: less than 7 periods per year.

Menstrual Concerns and Dysmenorrhea

Data for menstrual concerns and dysmenorrhea were adjusted for potential confounders, to include age, rank, use of OC, time deployed, and strenuous physical activity. Overall after adjusting for confounders, AA women were significantly less likely to report high total concerns and being bothered by menstruation than Caucasian (CA), AS, or HIS women (Table 3). Likewise, among AA women the score for dysmenorrhea was significantly lower than for the other groups (p < 0.01). No to minimal total concern was reported by 66% of all AA women compared to 50% for CA, 39.4% for AS, and 37.9% for HIS women. AA women were minimally bothered by breast tenderness, bloating, and performing physical activity during menstruation, whereas others viewed them as moderately to very bothersome. Significantly more AA reported no to minimal concerns for menstrual cramping, breast tenderness, and participating in strenuous physical exercise than the other groups (Figure 1). Additionally, significantly fewer AA women reported high concern scores for headaches, bloating, nausea, and unexpected spotting relative to other groups. AA reported less symptomatology, with only 6.1% reporting severe dysmenorrhea compared to 11.6% of CA, 20.9% of AS, and 19.7% of HIS women (Figure 2). From all perspectives, AA women experienced menstruation as less bothersome than others, despite no real divergence in menstrual irregularities, duration of menses, and/or time between cycles.

TABLE 2 Characteristics and Patterns of Menstrual Cycles by Race/Ethnicity

	Caucasian (n = 141)	Asian (n = 67)	Hispanic (n = 67)	African American (n = 184)
Cycles every 22–28 days (%)	51.8	40.3	54.4	49.7
Short Cycles* (%)	17.0	17.1	11.4	14.4
Long Cycles**	2.3	7.3	3.9	3.0
Amenorrhea*** (%)	26.1	19.5	23.5	18.2
Amenorrhea**** (%)	18.2	9.8	17.6	13.1
Break-Thru Bleeding ^# (%)	31.9	40.3	36.9	34.3
Menstrual bleeding ≥7 d (%)	14.7	19.4	9.2	9.4
Reported Moderate to Severe Menstrual Pain (%)	54.4	46.8	65.0	51.9
Sought Med Treat ≥ once (%)	21.9	18.2	16.7	16.3
Missed ≥ 1 day of work (%)	16.9	6.1*	15.4	12.2

TABLE 3 Total and Individual Menstrual Concerns and Dysmenorrhea Scores (Mean ± SEM) by Race/Ethnicity After Adjusting for Age, Rank, Use of OC, Time Deployed, and Strenuous Activity

Menstrual concern scores ^1	Caucasian (n = 141)	Asian (n = 67)	Hispanic (n = 67)	African American (n = 184)
Total Concerns	35.4 ± 1.7 ^a	41.1 ± 2.5 ^a	40.8 ± 2.5 ^a	28.1 ± 1.5 ^b
Physical training ^2	4.2 ± 0.2 ^a	4.7 ± 0.4 ^a	4.1 ± 0.3 ^a	2.9 ± 0.2 ^b
Change	4.4 ± 0.3 ^a ^b	5.4 ± 0.4 ^b	5.3 ± 0.4 ^b	3.5 ± 0.2 ^a
Cramps	4.6 ± 0.3 ^a ^b	4.4 ± 0.4 ^a ^b	5.2 ± 0.4 ^b	3.6 ± 0.2 ^a
Headache	5.0 ± 0.3 ^a ^b	5.5 ± 0.4 ^b	5.8 ± 0.4 ^b	4.3 ± 0.2 ^a
Bloating	3.9 ± 0.3 ^a	4.7 ± 0.4 ^a	5.0 ± 0.4 ^a	3.0 ± 0.2 ^b
Nausea	3.6 ± 0.2 ^a ^b	4.2 ± 0.4 ^b ^c	4.6 ± 0.4 ^b	2.8 ± 0.2 ^a
Breast tenderness	3.4 ± 0.2 ^a	3.8 ± 0.3 ^a	3.5 ± 0.3 ^a	2.4 ± 0.2 ^b
Unexpected cycle	3.3 ± 0.3 ^a	4.7 ± 0.4 ^b	3.7 ± 0.4 ^a ^b	2.9 ± 0.2 ^a
Spotting	3.0 ± 0.2 ^a ^b	3.8 ± 0.4 ^b	3.5 ± 0.4 ^a ^b	2.6 ± 0.2 ^a
Dysmenorrhea	20.5 ± 1.0 ^a	22.5 ± 1.5 ^a	24.1 ± 1.5 ^a	16.1 ± 0.9 ^b

FIGURE 1 The percent of CA, AS, HIS, and AA women who reported being bothered minimally and severely by various menstrual issues (*p < 0.05).

FIGURE 2 The percent of CA, AS, HIS, and AA women reporting mild, moderate, and severe dysmenorrhea (*p < 0.05).

No racial/ethnic differences in seeking treatment were noted, but AS women missed significantly fewer workdays than the other groups (Chi Square = 3.71; p = 0.05) (Table 1). Dysmenorrhea was a driver for both missing work and seeking medical care. Less than 4.4% of all women with mild dysmenorrhea missed work compared to 20.7% of those with moderate to severe dysmenorrhea. Clearly AS women reported missing less work: only 9.3% with moderate to severe dysmenorrhea missed ≥ 1 day of work for moderate or severe dysmenorrhea compared to 25.1% of all other women (Figure 3). Overall, women reporting moderate to severe dysmenorrhea were 5.7 times more likely to miss work than women with mild dysmenorrhea. Similar results were noted for seeking health care: only 9.2% of women with mild compared to 25.8% of those with moderate to severe (odds ratio = 3.44; p ≤ 0.001) dysmenorrhea sought health care. The relation between dysmenorrhea and seeking care did not differ by race/ethnicity except for severe dysmenorrhea: AS women sought health care at a significantly lower rate (21.4%) than the other groups (60.0%) (p < 0.01).

FIGURE 3 The percent of CA, AS, HIS, and AA women with moderate and severe dysmenorrhea who reported missing ≥ 1 day of work and seeking medical treatment for dysmenorrhea (*p < 0.05).

TABLE 4 Patterns of OC Use and Compliance by Race/Ethnicity

	CA (n = 141)	AS (n = 67)	HIS (n = 67)	AA (n = 184)
Ever used OC (%)	76.6	67.2	77.6	82.6
Used OC during deployment (%)	35.3	37.9	43.8	42.3
OC compliance (%)	54.0	40.0	60.0	65.0
Used OC continuously (%)	13.2	19.7	20.3	17.1
Desired education pre-deployment on OC (%)	81.6	88.9	91.4	83.7
Desired education in basic training on OC (%)	85.6	83.9	93.2	82.4
Would likely use continuously next deployment (%)	26.0	36.0	32.3	31.8

OC Agent Patterns

The majority of women, regardless of race/ethnicity, had used OC at some time in their lives. However, OC use during deployment was surprisingly low—less than 50% used any form of OC. No significant differences were noted across the four racial/ethnic groups in patterns of use, although compliance, defined as missing <1 pill per week, tended to be lower in AS relative to CA, HIS, and AA women (Table 4). Surprisingly few women took OC continuously. Some of the women (CA: 12.4%; AS: 18.0%; HI: 23.5%; and AA: 22.8%) reported never having heard about continuous OC use; about 20% of women were unsure of the risks associated with continuous use, but this did not differ by race/ethnicity. Interestingly, over 80% of all women, regardless of race/ethnicity, expressed a desire to learn more about continuous use of OC either prior to deployment or as part of their basic training. Despite desiring such education, less than 40% of women, regardless of race/ethnicity, reported they would or probably would use OC continuously during their next deployment.

DISCUSSION

Menstruation has been associated with significant morbidity in terms of physical distress, physical pain and/or disability, as well as missing school/work and seeking medical care. Selected reports have suggested racial/ethnic differences in the morbidity associated with menstruation (Gold et al., 2007; Harlow & Campbell, 1996; Wong & Khoo, 2009). The present study found that despite minimal differences in menstrual patterns or use of OC, AA women were significantly less bothered by menstruation and had lower rates of severe dysmenorrhea than the other groups. The rate of secondary amenorrhea did not differ by race/ethnicity, but was high. Patterns of OC use were comparable across racial/ethnic groups.

Secondary amenorrhea, a well recognized menstrual abnormality, has been defined in many ways (Harlow, 1986). We used two definitions with our military women who were not taking OC: 1) having three or fewer periods, or 2) having six or less periods during their year of deployment. A high percentage of amenorrhea was noted among deployed women—14.9% and 21.8% for the 3- and 6-month definitions, respectively, as compared to Patel et al. (2006), who reported an overall rate of 5.2% for women between 18–45 years living in Goa, India when amenorrhea was defined as no menstrual period in the previous 3 months. Their estimate was consistent with the 2.6% rate reported for college students by Bachmann and Kemmann (1982) and the 3.3% reported for Swedish women (Pettersson, Fries, & Nillius, 1973). The higher rates of this study may reflect the stressful nature of the deployment, as secondary amenorrhea has been associated with metabolic, physical, and/or psychological stress (Genazzani et al., 2006). However, the military women studied may have underestimated their total number of cycles over the year. Regardless, the percentage reporting secondary amenorrhea was high.

The prevalence of dysmenorrhea was compared by summing scores for cramps, headaches, nausea, bloating, and breast tenderness—symptoms consistent with multiple definitions of dysmenorrhea (Banikarim, Chacko, & Kelder, 2000; Gordley et al., 2000; Jamieson & Steege, 1996; O'Connell, Davis, & Westhoff, 2006; Patel et al., 2006; Wang et al., 2004; Wong & Khoo, 2009). Our prevalence of 54.2% for moderate and severe dysmenorrhea across all four racial/ethnic groups was within the 28% to 90% ranges reported in the literature (Banikarim, Chacko, & Kelder, 2000; Jamieson & Steege, 1996; Patel et al., 2006; Wang et al., 2004). However, it was much higher than the 31.2% reported for an Air Force population (Gordley et al., 2000). The rate of dysmenorrhea among AA women was significantly lower (43.1%) in this study than the 90% reported by Jamieson and Steege (1996) and 65% reported by Houston et al. (2006), and higher than the 30.8% reported by Gordley et al. (2000). Similar rate differences have been reported in the literature for AS and HIS women (Wang et al., 2004; Patel et al., 2006; Banikarim, Chacko, & Kelder, 2000). The rate of 64.2% for AS women was higher than the 28% among Chinese women (Wang et al., 2004) and the 33.4% among Indian women (Patel et al., 2006). In contrast, the finding of 71.2% among HIS women was comparable to the 75% among HIS women reported by Banikarim, Chacko, and Kelder (2000). Differences across studies may reflect how dysmenorrhea was defined. For example, Banikarim, Chacko, and Kelder (2000) defined dysmenorrhea as having painful menstruation during the previous 3 months, whereas Jamieson and Steege (1996) only asked about pain with their periods. Wang et al. (2004) defined dysmenorrhea as two or more days of pain within a menstrual cycle. Regardless of definition, HIS women had the highest rates of dysmenorrhea among the racial/ethnic groups.

The data of this study confirmed previously reported racial/ethnic differences in menstrual symptoms (Ferguson & Vermillion, 1957; Jamieson & Steege, 1996; Janiger, Riffenburgh, & Kersh, 1972; Woods, Most, & Dery, 1982; Gold et al., 2007). These differences remained after adjusting for confounders and regardless of which variables were considered confounders, significantly fewer AA women were moderately to severely bothered by menstrual bloating, headache, breast tenderness, cramping, and nausea compared to other racial/ethnic groups. This was consistent with the findings of Ferguson and Vermillion (1957) who found that only 24% and 34% of AA women as compared to 52% and 50% of CA women reported breast tenderness and cramping, respectively. However, they found 50.6% of AA women and only 23.5% of CA women reported headaches. Woods, Most, and Dery (1982) found only 8.6% of AA women experienced severe menstrual cramps as compared to 20.2% of CA women, but they also found AA women had a higher prevalence of headaches. Thus the data of this study were inconsistent with prior reports regarding headache, with the exception of the work of Gold et al. (2007) who reported HIS women had the highest rate of headaches. Likely explanations for the divergent results included sample size and wording of questions. Possible explanations for racial/ethnic differences in symptomatology may reflect differential perceptions of pain, ability to tolerate pain, daily life stressors, cultural expectations, and/or true differences in menstrual concerns. Further study will be required to address this issue.

In the current study sample AS women lost significantly fewer work days than the other groups: only 6.1% of AS women missed one day or more, and only 9.3% of AS women with moderate to severe dysmenorrhea missed 1 or more days. These rates were much lower than the 21.5% of Chinese girls who missed school due to dysmenorrhea (Wong & Khoo, 2009), but consistent with our overall data for other racial/ethnic groups when AS women were excluded: 25.1% of other women with moderate to severe dysmenorrhea missed work because of their menses. The high rate reported by Chinese girls may reflect their age and few years of menstrual cycles. In contrast, our prevalence rates were lower than those reported by Banikarim, Chacko, and Kelder (2000) who noted 38% of women with dysmenorrhea missed school, and O'Connell, Davis, and Westhoff (2006) who found 53.0% of adolescent girls with moderate to severe dysmenorrhea missed at least one school day each month. Patel et al. (2006) noted that 52.5% of all Indian women were unable to complete their regular activities because of dysmenorrhea, but rates changed as a function of severity. Rate differences may depend on whether the adolescent girls and women with mild dysmenorrhea were included.

With regard to seeking health care for menstrual symptoms, only 21.4% of AS women in this study with severe dysmenorrhea sought medical care as compared to 68.8%, 53.8%, and 54.5% of CA, HIS, and AA women, respectively. These rates were much higher than the 12% reported for Chinese adolescents (Wong & Khoo, 2009) and the 14% for HIS girls with moderate to severe dysmenorrhea (Banikarim, Chacko, & Kelder, 2000). However, overall only 18.1% of all military women sought health care for menstrual symptoms in contrast to the 37% of women reported by O'Connell, Davis, and Westhoff (2006). Other than for severe dysmenorrhea, our data indicated no significant racial/ethnic difference in seeking medical care.

No racial/ethnic differences in patterns of OC use were found, but relatively few women used OC during deployment and only half of those using OC took them continuously. This was surprising given the moderating effect of OC on menstrual symptoms and the belief that continuous use may improve quality of life (Wiegratz et al., 2004) and can be effective in terms of mediating menstrual symptoms (Edelman et al., 2005; Powell-Dunford et al., 2003). Over 80% of all women were interested in receiving education about continuous OC use. Whether introducing education about continuous OC use would confer benefit in terms of reducing lost workdays in the deployed setting will require further study.

Several limitations of the present study must be acknowledged. First, the results may not be generalizable to all women in the military as our study sample included only 22% of active duty military women, and selection from one center may not have provided a representative sample. Moreover, the smaller sample sizes for AS and HIS groups may have limited the power to detect statistically significant differences, particularly after correcting for confounders. Secondly, the data were subject to retrospective recall bias that may have limited the accuracy about menstrual patterns, dysmenorrhea and associated symptoms, lost duty days, and medical assessments. A prospective study, wherein menstrual-related symptoms and menstrual bleeding were tracked daily, and lost duty days and medical evaluations were determined via medical records, could address these limitations. However a prospective study would be difficult to conduct in the deployed environment due to significant operational and logistical demands. Also, a non-anonymous questionnaire may significantly under-represent menstrual concerns, lost workdays, and medical care sought. Despite the limitations of this study, important findings have emerged with regard to racial/ethnic differences in menstrual concerns.

In summary, racial/ethnic differences in menstrual cycle concerns were noted. In particular, AA women reported lower rates of dysmenorrhea and significantly fewer menstrual symptoms, and concerns about performing physical work during menstruation and the logistics of menses relative to CA, AS, and HIS women. Importantly, significantly fewer AA women reported being moderately to severely bothered by breast tenderness, bloating, headache, nausea, and cramping than other women. Interestingly, AS women were less likely to miss work and seek medical treatment for dysmenorrhea than the other groups. Reasons for racial/ethnic differences may reflect differential perceptions or true differences in menstrual concerns.

Notes

This article is not subject to U.S. copyright law.

The opinions and assertions expressed herein are those of the authors and should not be construed as reflecting those of the U.S. Army, the Department of Defense, or the Uniformed Services University of the Health Sciences.

*p < 0.05.

*Women not on OC reporting <22 days between cycles;

**Women not on OC reporting ≥35 days between cycles;

***Women not on OC reporting six or fewer periods during deployment;

****Women not on OC reporting 3 or fewer periods during deployment;

^#All women as break-through bleeding did not differ by OC use.

^aMeans with different letter superscripts differ significantly; p < 0.01.

^bMeans with different letter superscripts differ significantly; p < 0.01.

¹Ranges for Scores—Total Concerns: 9–90; Individual Concerns: 1–10; Dysmenorrhea: 5–50.

²Bothered by exercising or performing heavy physical labor or exerting at near maximal effort with heavy breathing.