Menstrual discomfort is common in women of reproductive age and may be related to mental, physical, and emotional health problems.1 The frequency or intensity of menstrual discomfort is typically based on women's self-report. The accuracy of self-report of menstrual discomfort in a questionnaire or an interview is unknown. Specifically, how well does a generalized self-report collected at a single time point reflect a woman's day-to-day experience of menstrual discomfort? Women may tend to overestimate their menstrual discomfort due to the common stereotype that menstruation is painful2,3 or because of a general human tendency to exaggerate pain experiences.4 Negative emotional states (such as depression and stress) at the time of pain may intensify pain recall, leading to overestimation.5,6 Alternatively, women may underreport their pain simply because they do not recall its intensity. Some authors hypothesize that new or acute pain may be better remembered than recurring pain, as menstrual pain tends to be.7 Another report suggested that people who rate their health higher tend to underreport chronic health conditions.8 Thus, women who consider themselves healthy may tend to underreport their recurring menstrual discomfort.
The causes of menstrual discomfort are not fully understood. Women's reports of the presence of menstrual pain have been correlated with circulating levels of vasopressin and prostaglandin metabolites.9 Prostaglandin metabolites are higher in the endometrial tissue10 or menstrual fluid11–13 of women with dysmenorrhea compared with women without pain. When women are treated with prostaglandin synthesis inhibitors, they report reduced premenstrual or menstrual symptoms, including pain.14,15 Decreases in intrauterine pressure after ibuprofen treatment have also been associated with decreases in reported pain severity among patients with dysmenorrhea.16
Authors of 2 studies have reported consistency in women's assessment of premenstrual and menstrual symptoms from 1 cycle to the next over 2 consecutive cycles,17,18 which suggests that cycles tend to be similar and that characterization of subjective menstrual symptoms is repeatable.
Our purpose was to compare women's summaries of their menstrual discomfort reported at enrollment with the frequency of any menstrual discomfort reported in a subsequent daily diary.
The study population comprised the daughters of women who had participated in a randomized trial of diethylstilbestrol (DES) during pregnancy in Chicago from 1950 to 1952. The daughters had been informed of their prenatal DES status. Details regarding recruitment, enrollment, and data collection are provided elsewhere.19 Briefly, the daughters of the trial participants were interviewed by telephone in 1990 when they were between the ages of 37 and 39. The telephone interview collected information on demographic, anthropometric, medical, reproductive, and menstrual characteristics.
Eligibility for a prospective study of menstrual cycles was determined during this interview. Women were ineligible if they were currently pregnant or breast-feeding, using oral contraceptives or other hormones, had undergone a hysterectomy, or had been through menopause. During the enrollment telephone interview, women were asked “approximately how often do you have cramps or backache with your menstrual periods? Would you say never, sometimes, often, or always?”
Those who were eligible and agreed to participate were sent a 6-month daily menstrual diary divided into weeks; women returned the data at the end of each week. In the daily diary, women reported any bleeding or spotting and were asked to “tell us about backache, cramps, or pain that you have with your period or ovulation.” Women were also asked if there were “any drugs taken for menstrual pain, including aspirin or Tylenol.” Three hundred fifty-eight women contributed data on at least 1 menstrual cycle in the prospective study. One woman did not answer the enrollment question, leaving 357 women in the analysis.
In the subsequent diary study, a woman was considered to have experienced discomfort in a given cycle if she reported any menstrual backache or cramps (with or without medication) at least once starting 5 days before the onset of menstrual bleeding and extending through the first 5 days of menses (“menses window”). This restricted window ensured that the opportunity to experience discomfort during menses was not dependent on the duration of menses. If any day was missing during this interval, menstrual discomfort was also treated as missing for that cycle, unless there was at least 1 day of pain reported on another day during the window. In that case, the cycle as a whole was treated as “with discomfort”; however, it was missing for the analyses of the number of days of pain within the window. If the woman reported taking medicine for pain, but did not specifically report menstrual discomfort, that day was considered to have been one in which the woman experienced pain. The number of menses windows with discomfort divided by the total number of observed menstrual cycles yielded the proportion of cycles with any menstrual discomfort. Because this proportion may not be well estimated with a small number of cycles, we limited this analysis to women with at least 4 prospectively observed menstrual cycles. This left 324 women for the proportion of cycles with discomfort analysis and 313 women in the number of days of pain analysis.
We compared the observed proportion of cycles with discomfort with the participants’ self-report of the frequency of menstrual discomfort. To describe agreement between the 2 sources, we divided women into 4 categories according to the proportion of their cycles with pain: 0, more than 0 but less than 0.5, at least 0.5 but less than 1.0, and 1.0. These 4 categories were compared with the enrollment classifications of never, sometimes, often, and always by means of a weighted κ coefficient. To determine the sensitivity of our categorization, we varied the cutpoint between the middle 2 categories from 0.3 to 0.7.
We also compared the average number of days with discomfort during each menses window with the questionnaire-reported frequency of menstrual discomfort (limited again to women with at least 4 cycles).
A previous analysis of this cohort found no differences in menstrual discomfort between women prenatally exposed to DES and those who were not.19 We conducted all analyses for both groups combined, but checked the main results for consistency between the 2 types of women.
The DES Reproductive Health study was approved by the National Institutes of Health institutional review board, and all participants gave their informed consent to participate in the study.
Fifty-four percent of diary participants had a college degree, 93% were white, 67% had a body mass index of less than 25 and 53% had never smoked. The frequency of discomfort across the menstrual cycle peaked on the first and second days of bleeding, with more than half of the women having discomfort on those 2 days (Fig. 1).
Of the 324 women who contributed data on at least 4 cycles, 7% recorded no days of discomfort during any of their menstrual windows. Among women who had reported at enrollment that they never experience menstrual discomfort, 35% reported no menstrual discomfort during the prospective study; this was the highest proportion of the 4 subgroups (Fig. 2). Only 6% of women who said they never experienced discomfort recorded discomfort in more than half of their observed cycles. Women who reported sometimes experiencing menstrual discomfort had a wide range of discomfort in their diaries; 22% had discomfort in less than half their cycles whereas 35% had discomfort in all of their cycles. Among women who reported often experiencing menstrual discomfort, 72% recorded discomfort in all of their cycles, whereas among women who reported always experiencing menstrual discomfort, 88% recorded discomfort in all of their cycles. These patterns were consistent between DES-exposed and -unexposed women.
Of the 1512 cycles with at least 1 day of recorded discomfort, two-thirds (1043) included at least 1 day of medication use for menstrual pain (59 cycles were missing medication information). When we restricted the analysis to menstrual pain that was accompanied by the use of pain medication, there was a stronger gradient with a woman's initial report at enrollment. Among women who reported never having discomfort, only 3% experienced at least 1 day of menstrual pain with medication use in at least half their cycles. This proportion increased to 36%, 67%, and 92% among women who said they sometimes, often, and always had discomfort.
The overall κ for agreement between the retrospective and prospective categorizations of menstrual discomfort was 0.35 (data not shown). When we varied the cutpoint between the prospective categories of sometimes and often, the κ ranged from 0.32 to 0.40. When we restricted the sample to medicated pain, κ was stronger and less variable (0.45–0.46).
For women who reported at enrollment that they never had discomfort, the average number of days of discomfort during the menses window was 0.3 (data not shown). This increased to 1.5 days among women who reported at enrollment that they sometimes had discomfort, 2.6 days for women who reported often, and 3.6 days for women who reported always.
We conjectured that a woman's interview report of menstrual discomfort may depend on the time elapsed since her most recent menses. We explored this by examining the time between the last menstrual period and the telephone interview, assessing whether the timing of the interview was related to her reported frequency of menstrual discomfort at enrollment. There was no difference in the proportions of women reporting never, sometimes, often, and always having menstrual discomfort (data not shown).
Women's responses to a single question about frequency of menstrual discomfort generally correlated with subsequent prospectively recorded frequency of discomfort. Almost all the women in this study recorded some menstrual discomfort, even if they had initially reported never having menstrual discomfort. Similarly, it was not uncommon for women who said they sometimes experienced discomfort to report at least 1 day of discomfort in all of their recorded cycles. This suggests a tendency to underreport the frequency of menstrual pain.
Women who reported more frequent discomfort at enrollment were also more likely to report more days of discomfort in each menses window. This could suggest some overlap in women's ideas of “frequency” of discomfort and “duration” of discomfort. Alternatively, duration of discomfort may be associated with severity, which in turn is associated with the woman's conception of frequency.
Our findings are not consistent with the results of Woods et al20 who found a tendency for women aged 18 to 35 to overstate their symptoms. In that study, 73 women filled out daily diaries for 2 months, and afterward they were asked to complete the Moos Menstrual Distress Questionnaire (asking about menstrual cramps and premenstrual backache during their most recent menstrual period).20 Although there was general agreement between prospective recording and reporting on the postdiary questionnaire, the symptoms rating was higher on the questionnaire compared with daily diary responses. It may be that keeping a menstrual diary before the questionnaire had sensitized these women to the later questions about their menstrual symptoms. Two other similarly structured studies also found that young women tended to overreport their premenstrual symptoms when asked to recall them retrospectively.21,22 Furthermore, women who believed premenstrual syndrome to be common tended to overreport more dramatically.21 This is consistent with the findings of McFarland et al23 who noted that among a group of women aged 17 to 31 the more a woman believed in the concept of “menstrual distress” the more she overreported the negative symptoms of the previous menses. One study asked women aged 18 to 41 to characterize retrospectively their premenstrual symptoms before a daily diary study of 1 menstrual cycle. The retrospective reports overestimated the symptoms reported during the daily diary.24
In another study, 40 college students were asked to predict their menstrual pain daily during menses.4 The level of pain predicted each morning was higher than the level reported at the end of the day. Women were also asked a week later to report their peak menstrual pain. Recall overestimated the prospective report (for a review of the assessment tools for menstrual cycle symptoms, see Haywood et al25).
In contrast, a study of 49 women found a slight underreporting of the intensity of dysmenorrhea 2 weeks after the onset of menses compared with measurements taken during the episode.26
These studies involve small samples and only 1 or 2 cycles per woman, whereas our study included 324 women and at least 4 cycles per woman. Additionally, most of these studies focused on younger women, whereas our population was older (37–39 years). Finally, several of these studies assessed menstrual and premenstrual symptoms broadly rather than pain or discomfort specifically. Such studies may reflect a tendency of women to overreport menstrual symptoms (including negative mood or irritability), whereas reports of pain or discomfort might be more consistent. Variability in the intensity of pain may be associated with overestimation of recalled pain.27 Unfortunately we were unable to assess this in our data as we did not have a measure of intensity of menstrual discomfort.
Our study is limited by the age range of our participants. Our study sample comprised the offspring of women enrolled in a clinical trial, all of whom were 37 to 39 years of age at interview. Thus, our results may not be generalizable to women of other ages. Furthermore, these women may be more aware of their reproductive characteristics because of their involvement with the trial (we did not, however, find any differences in reporting based on DES exposure itself). It is possible that the women in our study are the most vigilant participants from the DES trial (67% of the daughters were interviewed, 83% of those eligible for the menstrual study participated). If the participants in our study were selected for their vigilance, this study may suggest better consistency in reporting than would be found in a random sample of the population. One advantage of this population, however, is that selection for the study was not based on the woman's prior experience with menstrual discomfort. This allows for greater variability in menstrual discomfort frequency. Another limitation of the study is that neither the enrollment questionnaire nor the daily diary directly queried severity of menstrual discomfort. Finally, we did not evaluate the validity of discomfort reporting because there was no gold standard for comparison.
Menstrual discomfort is a relatively common occurrence among reproductive age women, and its causes are not well understood. There are currently no biologic methods to measure or quantify menstrual discomfort, and we must rely on women's self-reports to describe it. If women can accurately describe their menstrual symptoms in a single questionnaire, then clinical trials investigating effects on menstrual discomfort would have less need for baseline prospective data collection.28 We found that a single question about frequency of menstrual discomfort was predictive of several measures of discomfort in prospective records, and diary data provided a more detailed depiction of the pattern of discomfort experience.
We would like to thank Robert McConnaughey for his help with data management. We also thank Ruby Nguyen and Pablo Nepomnaschy for their comments on an earlier draft of this manuscript.
1. Strine TW, Chapman DP, Ahluwalia IB. Menstrual-related problems and psychological distress among women in the United States. J Womens Health (Larchmt)
. 2005;14:316 –323.
2. Koeske RW, Koeske GF. An attributional approach to moods and the menstrual cycle. J Pers Soc Psychol
3. Ruble DN. Premenstrual symptoms: a reinterpretation. Science
. 1977; 197:291–292.
4. Rachman S, Eyrl K. Predicting and remembering recurrent pain. Behav Res Ther
5. Niven CA, Murphy-Black T. Memory for labor pain: a review of the literature. Birth
. 2000;27:244 –253.
6. Gedney JJ, Logan H. Memory for stress-associated acute pain. J Pain
7. Erskine A, Morley S, Pearce S. Memory for pain: a review. Pain
8. Jabine TB. Reporting chronic conditions in the National Health Interview Survey. A review of findings from evaluation studies and methodological test. Vital Health Stat 2
9. Stromberg P, Akerlund M, Forsling ML, et al. Vasopressin and prostaglandins in premenstrual pain and primary dysmenorrhea. Acta Obstet Gynecol Scand
10. Lundstrom V, Green K. Endogenous levels of prostaglandin F2alpha and its main metabolites in plasma and endometrium of normal and dysmenorrheic women. Am J Obstet Gynecol
11. Chan WY, Hill JC. Determination of menstrual prostaglandin levels in non-dysmenorrheic and dysmenorrheic subjects. Prostaglandins
. 1978; 15:365–375.
12. Rees MC, Anderson AB, Demers LM, et al. Prostaglandins in menstrual fluid in menorrhagia and dysmenorrhoea. Br J Obstet Gynaecol
13. Seifert B, Liedtke MP, Kempf HJ, et al. Prostaglandins and primary dysmenorrhoea. Acta Biol Med Ger
. 1978;37:949 –950.
14. Jakubowicz DL, Godard E, Dewhurst J. The treatment of premenstrual tension with mefenamic acid: analysis of prostaglandin concentrations. Br J Obstet Gynaecol
15. Shangold MM, Aksel S, Schomberg DW, et al. Plasma prostaglandin F 2-alpha levels in dysmenorrehic women. Fertil Steril
. 1976;27: 1171–1175.
16. Milsom I, Andersch B. Effect of ibuprofen, naproxen sodium and paracetamol on intrauterine pressure and menstrual pain in dysmenorrhoea. Br J Obstet Gynaecol
. 1984;91:1129 –1135.
17. Mortola JF, Girton L, Beck L, et al. Diagnosis of premenstrual syndrome by a simple, prospective, and reliable instrument: the calendar of premenstrual experiences. Obstet Gynecol
18. Shaver JF, Woods NF. Concordance of perimenstrual symptoms across two cycles. Res Nurs Health
19. Hornsby PP, Wilcox AJ, Weinberg CR, et al. Effects on the menstrual cycle of in utero exposure to diethylstilbestrol. Am J Obstet Gynecol
. 1994;170:709 –715.
20. Woods NF, Most A, Dery GK. Estimating perimenstrual distress: a comparison of two methods. Res Nurs Health
21. Marvan ML, Cortes-Iniestra S. Women’s beliefs about the prevalence of premenstrual syndrome and biases in recall of premenstrual changes. Health Psychol
. 2001;20:276 –280.
22. Boyle GJ, Grant AF. Prospective versus retrospective assessment of menstrual-cycle symptoms and moods—role of attitudes and beliefs. J Psychopathol Behav Assess
23. McFarland C, Ross M, Decourville N. Womens theories of menstruation and biases in recall of menstrual symptoms. J Pers Soc Psychol
24. van den Akker O, Sharifian N, Packer A, et al. Contribution of generalized negative affect to elevated menstrual cycle symptom reporting. Health Care Women Int
25. Haywood A, Slade P, King H. Assessing the assessment measures for menstrual cycle symptoms: a guide for researchers and clinicians. J Psychosom Res
26. Brodie EE, Niven CA. Remembering an everyday pain: the role of knowledge and experience in the recall of the quality of dysmenorrhoea. Pain
. 2000;84:89 –94.
27. Stone AA, Schwartz JE, Broderick JE, et al. Variability of momentary pain predicts recall of weekly pain: a consequence of the peak (or salience) memory heuristic. Pers Soc Psychol Bull
. 2005;31:1340 –1346.
28. Miller MN, McGowen KR, Miller BE, et al. Lessons learned about research on premenstrual syndrome. J Womens Health Gend Based Med
. 1999;8:989 –993.