Methods

The study subjects were female students at three colleges and two nursing schools between 1998 and 2001. A total of 362 women agreed to participate in the present study and responded to a self-administered questionnaire that asked about menstrual history, demographic characteristics, smoking and drinking habits, diet, exercise, and past medical and reproductive histories. The response rate was 90.0%. The present study was approved by the institutional review board.

The severity of menstrual pain was measured using the verbal multidimensional scoring system reported by Andersch and Milson (1982). This scoring system grades pain as none, mild, moderate , or severe and takes into account the effect of pain on daily activity, systemic symptoms, and analgesic requirements. Each woman was asked to report the date of the beginning of her last menses, the length of her usual menstrual cycle, and the number of days of bleeding. For woman who reported irregular menstrual cycles, we asked the range of the length of cycles and allotted the median as her cycle length.

Exercise was assessed by asking the average hours per week spent performing various kinds of activities during the past year. The details including its validity are described elsewhere (Suzuki et al, 1998).

Diet including soy, fat, and dietary fiber intakes was assessed by a semiquantitative food-frequency questionnaire. The women were asked to indicate the average frequency that they consumed 169 food items during the year prior to the study and the usual serving size of each item. We included nine food items for soy products (miso soup, tofu, deep-fried tofu, fried bean curd, dried bean curd, fermented soy beans, houba-miso, soymilk, and boiled soy beans). These nine items and some other dishes including soy products as ingredients were taken account for to obtain the estimates for total amount (g) of soy products and isoflavone intake. Isoflavone intake (mg/day) from soy products was estimated using isoflavone concentration in these soy foods (Wakai et al, 1999). The intakes of foods and nutrients were estimated from the frequency of ingestion and portion size using the Japanese Standard Tables of Food Composition, 4th and 5th editions, published by the Science and Technology Agency of Japan (2001). Fatty acid composition was evaluated using data published by Sasaki et al (1999). Detailed information on the questionnaire including its validity and reproducibility has been described elsewhere (Shimizu, 1996; Shimizu et al, 1999; Nagata et al, 2001). For example, the Spearman correlation coefficients comparing estimates of soy product intake from this questionnaire with the estimates from 12 daily diet records kept over a year period was 0.68. The corresponding figures for total fat and dietary fiber were 0.52 and 0.60.

Because of incomplete or unreliable responses to the dietary questionnaire (criteria shown in the reference by Shimizu, 1996), we did not assess the diets of 44 women. The response to the menstrual pain scale was missing for six women. One woman did not report her age. Therefore, the ultimate response rate was 77.4%. We restricted study subjects to women aged 24 y or less, because the frequency of secondary dysmenorrhea is likely to be higher in the elder women (Wentz, 1988). Therefore, 29 women were excluded. We further excluded women who had been taking steroid hormones during the previous 6 months (n=9) or who had a history of thyroid diseases (n=1) or other endocrine diseases (n=2). No one reported ovariectomy or use of oral contraceptives. The remaining 276 women aged 19–24 y consisted of the present study. Age distribution of the study subjects were 81 (29.3%), 106 (38.4%), 70 (25.4%) , 11 (3.9%), and 8 (2.9%) for 19, 20, 21, 22, 23+ y of age, respectively.

Spearman's correlation coefficients were used to examine the associations of severity of menstrual pain with study variables. Dietary values were log-transformed and adjusted for total energy using the method proposed by Willett (1990). Adjustment for potential confounders of the associations between dietary variables and the severity of menstrual pain was accomplished by regressing the menstrual pain scale and dietary values separately upon confounders. Spearman's correlation coefficients were then calculated. Several nondietary factors including weight, height, body mass index, smoking, exercise, marital status, age at menarche, menstrual cycle, days of bleeding, and number of births or pregnancies and intakes of macro- and micronutrients were examined as potential confounders. Age was always included in the model as a covariate to calculate the partial correlation coefficients. All statistical analyses were performed using SAS programs (Version 8, SAS Institute, Cary, NC, USA).

Results

The distribution of menstrual pain scores among the study subjects was 46 (16. 7%), 111 (40.2%), 95 (34.4%), and 24 (8.7%) for grades 0–3 (none, mild, moderate, and severe), respectively. Characteristics of subjects according to menstrual pain scale are shown in Table 1. Group comparison for any variable except age at menarche did not reveal a significant association with menstrual pain scale.

Table 1 - Characteristics of study subjects according to menstrual pain score.

Full table

Table 2 shows the correlation coefficients between selected nondietary variables and the menstrual pain scale. Age at menarche was significantly inversely correlated with the menstrual pain scale. Smoking status was positively associated with the menstrual pain scale, but this association was of borderline significance (P=0.06).

Table 2 - Correlations of selected nondietary variables with menstrual pain scale.

Full table

Dietary fiber was significantly inversely correlated with the menstrual pain scale after controlling for age, smoking status and age at menarche (r=-0.12, P=0.04) (Table 3). There were no significant correlations between the menstrual pain scale and intakes of soy product or isoflavone as well as any type of fat. The positive association between saturated fat intake and the menstrual pain scale was of borderline significance (P=0.08). The additional adjustment for marital status and numbers of days of bleeding did not alter the results substantially (for example, the correlation coefficient between dietary fiber and the menstrual pain scale was -0.13, P=0.03). Reanalysis restricting the subjects to those who reported a regular menstrual cycle with length of 25–35 days (n=156) did not attenuate the association between dietary fiber intake and the menstrual pain scale (r=-0.14, P=0.10).

Table 3 - Correlation coefficients between dietary intakes and menstrual pain scale.

Full table

Discussion

In spite of the relatively low intake levels of dietary fiber in our study subjects, we found a moderate but significant inverse association between dietary fiber intake and menstrual pain. It is well known that primary dysmenorrhea occurs only in ovulatory cycles (Friederich, 1983), indicating that adequate uterine exposure to estrogen and then to progesterone is necessary. Studies have suggested that fiber intake decreases blood estrogen levels in women (Kaneda et al, 1997; Rose et al, 1997). Although fat intake has been associated with increased estrogen levels (Wu et al, 1999), we failed to find a significant positive association between fat intake and menstrual pain. Neither soy product nor isoflavone intake was associated with menstrual pain. We expected that dietary soy would be inversely associated with menstrual pain through its effects on estrogens or on the cyclooxygenase pathway. However, such effects did not appear to be clinically relevant regarding dysmenorrhea. It is also possible that a limited range of soy intake as well as fat intake among the study subjects may have obscured a real association. Additional findings on smoking and age at menarche in relation to menstrual pain were consistent with previous results from other studies (Klein & Litt, 1981; Sundell et al, 1990; Parazzini et al, 1994; Harlow & Park, 1996; Hornsby et al, 1998).

So far, to our knowledge, five studies have assessed the relationship between diet and menstrual pain (Deutch, 1995; Harel et al, 1996; Di Cintio et al, 1997; Balbi et al, 2000; Barnard et al, 2000). One of them (Harel et al, 1996) was based on dietary intervention using supplementation of n-3 fatty acids. In two other studies (Di Cintio et al, 1997; Balbi et al, 2000), dietary fiber as well as fat intake could not be estimated because the questionnaires used for measuring diet, which were apparently not validated, included a limited number of food items. Barnard et al (2000) reported that a low-fat vegetarian diet with a change of total fiber from 26.7 to 31.3 g was associated with an increase in sex hormone-binding globulin levels and with reductions in dysmenorrhea duration and intensity. Their findings are not contradictive with our results. In the remaining study reported by Deutch (1995), diet was measured by a 4-day diet record, and fat and dietary fiber intakes were not significantly associated with menstrual pain after controlling for covariates.

One of the limitations of our study is that we could not perform physical imaging and surgical examinations, such as uterosonography and laparoscopy, to rule out secondary cause of dysmenorrhea. The frequency of secondary dysmenorrhea is much lower than that of primary dysmenorrhea in this age group (Balbi et al, 2000). However, we cannot deny the possibility that our findings from a study with a small number of subjects were due to the inclusion of secondary dysmenorrhea.

The lack of endocrinologic measures of ovarian activity was another limitation of the present study. Women with anovulatory cycles do not experience menstrual pain. We could not determine whether each woman was ovulatory or not. Thus, there might be a concern that dietary fiber intake may be associated with anovulation rather than menstrual pain. However, when we reanalyzed data restricted to subjects with normal cycle lengths of 25–35 days, whom we thought to be ovulatory (Harlow & Ephross, 1995), the association between dietary fiber and the menstrual pain scale was not altered.

We employed a widely used scaling system to assess menstrual pain. However, pain is difficult to measure because it cannot be confirmed by any instrumental or clinical evaluation. Therefore, measurement error may have affected the results. However, it seems unlikely that women who had a lower intake of dietary fiber reported their pain more inaccurately or perceived more pain than those with a higher intake of dietary fiber. The food-frequency questionnaire, like all methods of dietary assessment, is subject to measurement error. Our questionnaire was designed to measure an individual's relative intakes of foods and nutrients rather than absolute values. The data presented for soy products may have been overestimated because soy product intake estimated from the questionnaire was 40% higher than that estimated from the 12 daily diet records. The estimate for dietary fiber was 8% higher than that from the diet records. However, again, it is likely that this measurement error was unrelated to menstrual pain and led to an underestimation of the true associations.

Owing to the cross-sectional study design, we can only infer associations. Neuroendocrine functions of the body, mental attitude, and food choice may be mutually related. It is possible that the decreased intake of dietary fiber might be a consequence of menstrual pain. However, if this is true, intakes of other nutrients should also have been affected. None of the other measured nutrients or food groups was associated with the menstrual pain scale. Although the cross-sectional differences in dietary fiber intake across the level of menstrual pain was small in magnitude, more attention should be paid to the role of diet, including soy, fat, and dietary fiber, in the etiology of dysmenorrhea cases that are amenable to public health intervention.

References

1. Andersch B & Milson I (1982): An epidemiologic study of young women with dysmenorrhea. Am. J. Obstet. Gynecol. 144, 655–660. | PubMed | ChemPort |
2. Barnard ND, Scialli ARS, Hrlock D & Bertron P (2000): Diet and sex-hormone binding globulin, dysmenorrhea, and premenstrual symptoms. Obstet. Gynecol. 95, 245–250. | Article | PubMed | ChemPort |
3. Balbi C, Musone R, Menditto, Di Prisco L, Cassese E, D'Ajello nM, Ambrosio D & Cardone AA (2000): Influence of menstrual factors and dietary habits on menstrual pain in adolescence age. Eur. J. Obstet. Gynecol. 91, 143–148. | Article | ChemPort |
4. Deligeoroglou E (2000): Dysmenorrhea. Ann. NY Acad. Sci. 900, 237–244. | PubMed | ChemPort |
5. Deutch B (1995): Menstrual pain in Danish women correlated with low n-3 polyunsaturated fatty acid intake. Eur. J. Clin. Nutr. 49, 508–516. | PubMed | ChemPort |
6. Di Cintio E, Parazzini F, Tozzi L, Luchini L, Mezzopane R, Marchini M & Fedele L (1997): Dietary habits, reproductive and menstrual factors and risk of dysmenorrhoea. Eur. J. Epidemiol . 13, 925–930. | Article | PubMed | ChemPort |
7. Friederich MA (1983): Dysmenorrhea. Women Health 8, 91–106. | Article | PubMed | ChemPort |
8. Harel Z, Biro FM, Kottenhahn RK & Rosenthal SL (1996): Supplementation with omega-3 polyunsaturated fatty acids in the management of dysmenorrhea in adolescents. Am. J. Obstet. Gynecol. 174, 1335–1338. | PubMed | ChemPort |
9. Harlow SD & Ephross SA (1995): Epidemiology of menstruation and its relevance to women's health. Epidemiol. Rev . 17, 265–286. | PubMed | ChemPort |
10. Harlow S & Park M (1996): A longitudinal study of risk factors for the occurrence, duration and severity of menstrual cramps in a cohort of college women. Br. J. Obstet. Gynaecol. 103, 1134–1142. | PubMed | ChemPort |
11. Hornsby PP, Wilcox AJ & Weinberg CR (1998): Cigarette smoking and disturbance of menstrual function. Epidemiology 9, 193–198. | PubMed | ChemPort |
12. Kaneda N, Nagata C, Kabuto M & Shimizu H (1997): Fat and fiber intakes in relation to serum estrogen concentration in premenopausal Japanese women. Nutr. Cancer 27, 279–283. | PubMed | ChemPort |
13. Klein JR & Litt IF (1981): Epidemiology of adolescent dysmenorrhea. Pediatrics 68, 661–664. | PubMed | ChemPort |
14. Kurzer MS (2002): Hormonal effects of soy in premenopausal women and men. J. Nutr. 132, 570S–573S. | PubMed | ISI |
15. Liang Y-C, Huang Y-T, Tsai S-H, Lin-Shiau S-Y, Chen C-F & Lin J-K (1999): Suppression of inducible cyclooxygenase and inducible nitric oxide synthase by apigenin and related flavonoids in mouse macrophages. Carcinogenesis 20, 1945–1952. | Article | PubMed | ISI | ChemPort |
16. Miyagi M, Morishita M & Iwamoto Y (1993): Effects of sex hormones on production of prostaglandin E2 by human peripheral monocytes. J. Periodontol. 64, 1075–1078. | PubMed | ChemPort |
17. Nagata C, Takatsuka N, Kawakami N & Shimizu H (2001): Soy product intake and hot flashes in Japanese women: results from a community-based prospective study. Am. J. Epidemiol. 153, 790–793. | Article | PubMed | ChemPort |
18. Pavan B, Biondi C, Ferretti ME, Lunghi L & Paganetto G (2001): 17 -estradiol medulates prostaglandin E2 release from human amnion-derived WISH cells. Biol. Reprod. 64, 1677–1681. | Article | PubMed | ISI | ChemPort |
19. Parazzini F, Tozzi L, Mezzopane R, Luchini L, Marchini M & Fedele L (1994): Cigarette smoking, alcohol consumption, and risk of primary dysmenorrhea. Epidemiology 5, 469–472. | PubMed | ChemPort |
20. Picherit C, Dalle M, Néliat G, Lebecque P, Davicco MJ, Barlet JP & Coxam V (2000): Genistein and daidzein modelate in vitro rat uterine contractile activity. J. Steroid Biochem. Mol. Biol. 75, 201–208. | Article | PubMed | ChemPort |
21. Pickles VR, Hall WJ, Best FA & Smith GN (1965): Prostaglandins in endometrium and menstrual fluid from normal and dysmenorrhoeic subjects. J. Obstet. Gynaecol. Br. Comm. 72, 185–192. | ChemPort |
22. Rose DP, Lubin M & Connolly JM (1997): Effects of diet supplementation with wheat bran on serum estrogen levels in the follicular and luteal phases of the menstrual cycle. Nutrition 13, 535–539. | Article | PubMed | ChemPort |
23. Sasaki S, Kobayashi M & Tsugane S (1999): Development of substituted fatty acid composition table for the use in nutritional epidemiologic studies for Japanese populations: its methodological backgrounds and the evaluation. J. Epidemiol. 9, 190–207. | PubMed | ChemPort |
24. Shimizu H (1996): The Basic Report on Takayama Study. Gifu, Japan: Department of Public Health, Gifu University School of Medicine.
25. Shimizu H, Ohwaki A, Kurisu Y, Takatsuka N, Kawakami N, Ido M, Nagata C & Inaba S (1999): Validity and reproducibility of a quantitative food frequency questionnaire for a cohort study in Japan. Jpn. J. Clin. Oncol. 29, 38–44. | Article | PubMed | ChemPort |
26. Steusloff A, Paul E, Semenchuk LA, Di Salvo J & Pfitzer G (1995): Modulation of Ca²⁺ sensitivity in smooth muscle by genistein and protein tyrosine phosphorylation. Arch. Biochem. Biophys. 320, 236–242. | Article | PubMed | ChemPort |
27. Strom BL, Schinnar R, Ziegler EE, Barnhart KT, Sammel MD, Macones GA, Stallings VA, Drulis JM, Nelson SE & Hanson SA (2001): Exposure to soy-based fomula in infancy and endocrinological and reproductive outcomes in young adulthood. JAMA 286, 807–814. | Article | PubMed | ChemPort |
28. Sundell G, Milson I & Andersch B (1990): Factors influencing the prevalence and severity of dysmenorrhea in young women. Br. J. Obstet. Gynecol. 97, 588–594. | ChemPort |
29. Suzuki I, Kawakami N & Shimizu H (1998): Reliability and validity of a questionnaire for assessment of energy expenditure and physical activity in epidemiological studies. J. Epidemiol. 8, 152–159. | PubMed | ChemPort |
30. Willett W (1990): Implication of total energy intake for epidemiological analyses. In Nutritional Epidemiology ed. W Willett, pp 245–271. Oxford: Oxford University Press.
31. Wu AH, Pike MC & Stram DO (1999): Meta-analysis: dietary fat intake, serum estrogen levels, and the risk of breast cancer. J. Natl. Cancer Inst. 91, 529–534. | Article | PubMed | ChemPort |
32. Wentz AC (1988): Dysmenorrhea, premenstrual syndrome, and related disorders. In Novak's Textbook of Gynecology eds. Jones HW, Wentz AC, Burnett LB, pp 240–251. London: Williams & Willkins.
33. Wakai K, Egami I, Kato K, Kawamura T, Tamakoshi A, Lin Y, Nakayama T, Wada M & Ohno Y (1999): Dietary intake and sources of isoflavones among Japanese. Nutr. Cancer 33, 139–145. | Article | PubMed | ISI | ChemPort |
34. Yamaki K, Kim D-H, Ryu N, Kim YP, Shin KH & Ohuchi K (2001): Effects of naturally occurring isoflavones on prostaglandin E2 production. Planta Med. 68, 97–100. | Article |