1 Istituto di Ricerche Farmacologiche Mario Negri, 20157 Milano, 2 Clinica Ostetrico Ginecologica, Università di Milano, 20122 Milano and 3 Studi di via Fontana, 20122 Milano, Italy
4 To whom correspondence should be addressed at: Istituto di Ricerche Farmacologiche Mario Negri, via Eritrea, 6220157 Milano, Italy. Email: parazzini{at}marionegri.it
![]() |
Abstract |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Key words: casecontrol study/diet/endometriosis/risk factors
![]() |
Introduction |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The role of diet in the development of hormone-related diseases has become a topic of interest in recent years (Ingram et al., 1987; Fentiman et al., 1988
). For example, diet may have some influence on ovarian and endometrial carcinogenesis and on the development of benign gynaecological conditions, such as fibroids and ovarian cysts (La Vecchia et al., 1987
; Mori et al., 1988
; Chiaffarino et al., 1999
; Kushi et al., 1999
; Britton et al., 2000a
,b
; Bosetti et al., 2001
). Endometriosis is hormone-related (Olive and Schwartz, 1993
), so diet may play a role in its aetiopathogenesis. A casecontrol study in the USA suggested that the risk of endometrioid cysts was elevated for high intake of total, vegetable, non-saturated and polyunsaturated fats (Britton et al., 2000a
).
In order to obtain information on the relationship between diet and risk of pelvic endometriosis, we analysed data collected in two casecontrol studies (Parazzini et al., 1989, 1995
).
![]() |
Materials and methods |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The studies included 504 women aged <65 years (median age 33 years, range 2065) with a laparoscopically confirmed diagnosis of endometriosis, admitted to a network of obstetrics and gynaecology departments in Milan and university obstetrics and gynaecology clinics in Brescia and Pavia.
Controls were women aged <65 years admitted for acute non-gynaecological, non-hormonal, non-neoplastic conditions to the Ospedale Maggiore (comprising the four major teaching and general hospitals in Milan) and several university clinics, serving a catchment area comparable to that of the hospitals where cases had been identified, and the university hospitals of Brescia and Pavia. They were recruited as controls in one of the casecontrol studies on endometriosis (Parazzini et al., 1995) and in a casecontrol study of female genital neoplasms (Parazzini et al., 1989
, 1992
). A total of 2422 controls was identified and 504 (age range 2061 years; median 34 years) were matched with cases in a 1:1 ratio, and randomly selected within strata of 5 year age groups, centre and calendar year of interview. Of these, 31% were admitted for traumatic conditions (mostly fractures and sprains), 23% had non-traumatic orthopaedic disorders (mostly low back pain and disc disorders), 12% acute abdominal diseases requiring surgery, and 34% other miscellaneous illnesses, such as disorders of the ear, nose, throat, or teeth.
Trained interviewers identified and questioned cases and controls. All interviews were conducted in hospital. Less than 3% of cases and controls refused to be interviewed.
Information was obtained using a structured questionnaire, on general socio-demographic factors, personal characteristics and habits, gynaecological and obstetric history, and lifetime oral contraceptive use. Women were also asked about their frequency of consumption per week (i.e. in 14 meals) of portions of selected dietary items including the major sources of retinoids and carotenoids in the Italian diet, and alcohol and coffee drinking in the year before interview. Subjective scores (low, intermediate and high) were used to collect information on fat intake (butter, margarine and oil) and consumption of whole-grain foods. Reproducibility of the questionnaire was satisfactory (D'Avanzo et al., 1997).
The thresholds for the analysis of dietary factors were based on the best possible approximations of tertiles of control group. Specifically they were: milk (0, 0.56, 7 portions/week), meat (03, 46,
7 portions/week), liver (0,
0.5 portions/week), carrots (0, 1,
2 portions/week), green vegetables (06, 712,
13 portions/week), fresh fruit (
6, 713,
14 portions/week), eggs (0, 1,
2 per week), ham (
1, 2,
3 portions/week), fish (0, 1,
2 portions/week), cheese (
2, 35,
6 portions/week). In some cases the numbers of subjects were not equally distributed in the tertiles because of the large number reporting the same frequency of consumption. The items green vegetables and fruits included all types, specifically all the main sources in the Italian diet such as spinach/other greens, cruciferae, green and red salads, zucchini, artichokes; fruits included citrus, apple, peach, melon, strawberries/cherries, banana and pear (Franceschi et al., 1993
).
An estimate of the total daily average alcohol intake was derived assuming a comparable ethanol content in each type of beverage (125 ml wine=333 ml beer=40 ml spirits=15 g pure alcohol). Wine accounted for >80% of the alcohol consumed.
To account simultaneously for the effects of several potential confounding factors, we performed unconditional multiple logistic regression, with maximum likelihood fitting, to obtain the odds ratios (OR) of endometriosis, their corresponding 95% confidence intervals (CI), and the test for trend when appropriate (Baker and Nelder, 1978). The variables included in the model are listed in the footnotes to the tables.
Since a total of 11 2-tests for trend were done in the analysis of dietary factors, P<0.004 can be considered statistically significant after taking into account the effect of multiple tests, according to the Bonferroni test (Perneger, 1998
).
![]() |
Results |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
|
Table III presents the OR for an increase of one serving per day for beef and other red meat, green vegetables, fruit, and ham in strata of selected variables. All food groups significantly associated with endometriosis were simultaneously included in the same multiple logistic model, to allow for mutual confounding. The associations were generally consistent in strata of age, body mass and parity; the OR for beef and ham were higher in strata of low education and high education, but no significant heterogeneity was observed. No difference emerged in the OR when we analysed the effect of diet in women living in different areas of Italy.
|
![]() |
Discussion |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Some limitations must be considered in interpreting the results. First of all, the dietary section in this study was restricted to a few selected indicator foods. Information was limited to the number of portions per week of a restricted list of dietary items, with no estimate of portion size. Thus, no estimate of total caloric intake could be obtained (Willett and Stampfer, 1986). However, a major role of information bias is unlikely, since the possible relationship between diet and endometriosis was probably not known to interviewers and to the majority of women interviewed. The diet questionnaire was satisfactorily reproducible (D'Avanzo et al., 1997
). We collected information on ham, but not on ground pork or pork chops; these latter, however, are rarely eaten in Italy (Turrini et al., 2001
).
A major effect of selection bias is unlikely because the control group included only women with acute conditions and we excluded women with digestive tract diseases or any condition potentially related to long-term dietary changes. Controls were not examined by laparoscopy, so we cannot exclude that some may have had undiagnosed endometriosis. This can be considered a limitation of this study, but the potential misclassification should only underestimate any difference between cases and controls.
We analysed several dietary items, so the association between endometriosis risk and green vegetable, fruit and meat intake might be due to chance. However, after taking into account multiple tests for all dietary items in the catagories, except ham which was a single catagory, P-values were still significant.
Selection bias should be considered in interpretating of the findings. Green vegetables, fruit and fish may be general indicators of a more health-oriented attitude toward diet and other lifestyle habits. Closer attention to health may also favour the diagnosis of endometriosis, thus producing an underestimate of the real association. The diagnosis of endometriosis was more frequent among more educated women of higher social class (Mangtani and Booth, 1933; Cramer et al., 1986
; Parazzini et al., 1995
; Signorello et al., 1997
) and could to some extent reflect the greater attention such women pay to relatively minor health problems. The association between socio-economic status and endometriosis risk may also involve the inverse association between parity and socio-economic status, since more educated women are more likely to be nulliparous in Italy.
Socio-economic status, body weight, and potential reproductive and hormonal risk factors for endometriosis did not explain the results. The estimated OR were not markedly affected by the inclusion of terms for education and parity in the multivariate models (data not shown). Although the OR estimates for beef and ham were different in strata of education, there was no significant heterogeneity.
Epidemiological data on the relationship between endometriosis and diet are scanty. A study conducted in the USA on ovarian endometrioid cysts reported elevated risks of endometriosis for higher intakes of polyunsaturated and vegetable fats, but no reduction in risk for high intake of vegetables and fruits (Britton et al., 2000a).
There are, however, some indications that a diet poor in vegetables and fruits and rich in fat increases the risk of endometrial cancer (Armstrong, 1979) and fibroids (Chiaffarino et al., 1999
), two diseases known to be associated with estrogens, and of ovarian benign and malignant epithelial diseases (Risch et al., 1994
). For example, for endometrial and ovarian cancer and fibroids, there was a direct association with the frequency of consumption of meat and ham in this Italian population, whereas high intake of vegetables and fruits conferred some protection (Levi et al., 1993
; Chiaffarino et al., 1999
; Bosetti et al., 2001
). In biological terms, fats may influence prostaglandin concentrations, which may affect ovarian function (Smith, 1986
). Hormonal factors are a potential link between diet and endometriosis, since the risk may be increased by exposure to unopposed estrogens, and a diet rich in fat increases circulating unopposed estrogens (Armstrong et al., 1981
; Goldin et al., 1982
; Gorbach and Goldin, 1987
). More difficult to explain in biological terms is the protective effect of a diet rich in green vegetables and fruits. However, similar findings emerge for the risk of breast and endometrial cancer, two estrogen-related diseases. A diet rich in green vegetables and fruits includes high levels of vitamin C, carotenoids, folic acid and lycopene, micronutrients which may help to protect against cell proliferation (Bosetti et al., 2002
). We did not find any association between alcohol intake and risk of endometriosis. Some studies have reported that women with endometriosis tend to drink more alcohol than those without the disease (Grodstein et al., 1994
; Missmer and Cramer, 2003
), also after taking into account the potential effect of alcohol on fertility. Most of these studies, however, have been conducted in North European or American countries, where alcohol intake is likely to be higher than in Italy. This may partly explain the lack of association found in the present study.
In accordance with previous studies, we found an inverse association between body mass and risk of endometriosis (Darrow et al., 1993; Signorello et al., 1997
; Missmer and Cramer, 2003
).
In conclusion, despite its limitations, this study suggests that there is some link between diet and risk of endometriosis. These findings suggest the need for a proper prospective interventional investigation designed to study these factors.
![]() |
Acknowledgements |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
![]() |
References |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Armstrong BK, Brown JB, Clarke HT, Crooke DK, Hähnel R, Masarei JR and Ratajczak T (1981) Diet and reproductive hormones: a study of vegetarian and nonvegetarian postmenopausal women. J Natl Cancer Inst 67, 761767.[Medline]
Baker RJ and Nelder JA (1978) The GLIM System. In Release. Numerical Algorithms Group, Oxford.
Bosetti C, Negri E, Franceschi S, Pelucchi C, Talamini R, Montella M, Conti E and La Vecchia C (2001) Diet and ovarian cancer risk: a case-control study in Italy. Int J Cancer 93, 911915.[CrossRef][Medline]
Bosetti C, Altieri A and La Vecchia C (2002) Diet and environmental carcinogenesis in breast/gynaecological cancers. Cur Opin Obstet Gynecol 14, 1318.[CrossRef]
Britton JA, Westhoff C, Howe G and Gammon MD (2000a) Diet and benign ovarian tumors (United States). Cancer Causes Control 11, 389401.[CrossRef][Medline]
Britton JA, Westhoff C, Howe GR and Gammon MD (2000b) Lactose and benign ovarian tumours in a case-control study. Br J Cancer 83, 15521555.[CrossRef][Medline]
Chiaffarino F, Parazzini F, La Vecchia C, Chatenoud L, Di Cintio E and Marsico S (1999) Diet and uterine myomas. Obstet Gynecol 94, 395398.
Cramer DW, Wilson E, Stillman RJ, Berger MJ, Belisle S, Schiff I, Albrecht B, Gibson M, Stadel BV and Schoenbaum SC (1986) The relation of endometriosis to menstrual characteristics, smoking and exercise. J Am Med Assoc 255, 19041908.[Abstract]
Darrow SL, Vena JE, Batt RE, Zielezny MA, Michalek AM and Selman S (1993) Menstrual cycle characteristics and the risk of endometriosis. Epidemiology 4, 135142.[Medline]
D'Avanzo B, La Vecchia C, Katsouyanni K, Negri E and Trichopolous D (1997) An assessment, and reproducibility of food frequency data provided by hospital controls. Eur J Cancer Prev 6, 288293.[Medline]
Fentiman IS, Caleffi M, Wang DY, Hampson SJ, Hoare SA, Clark GM, Moore JW, Bruning P and Bonfrer JM (1988) The binding of blood-borne estrogens in normal vegetarian and omnivorous women and the risk of breast cancer. Nutr Cancer 11, 101106.[Medline]
Franceschi S, Negri E, Salvini S, Decarli A, Ferraroni M, Filiberti R, Giacosa A, Talamini R, Nanni O and Panarello G (1993) Reproducibility of an Italian food frequency questionnaire for cancer studies: results for specific food items. Eur J Cancer 29A, 22982305.[CrossRef][Medline]
Goldin BR, Adlercreutz H, Gorbach SL, Warram JH, Dwyer JT, Swenson L and Woods MN (1982) Estrogen excretion patterns and plasma levels in vegetarian and omnivorous women. N Engl J Med 307, 15421547.[Abstract]
Gorbach SL and Goldin BR (1987) Diet and the excretion and enterohepatic cycling of estrogens. Prev Med 16, 525531.[CrossRef][Medline]
Grodstein F, Goldman MB and Cramer DW (1994) Infertility in women and moderate alcohol use. Am. J Publ Health 84, 14291434.[Abstract]
Houston DE (1984) Evidence for the risk of pelvic endometriosis by age, race and socioeconomic status. Epidemiol Rev 6, 167191.[Medline]
Ingram DM, Bennett FC, Willcox D and de Klerk N (1987) Effect of low-fat diet on female sex hormone levels. J Natl Cancer Inst 79, 12251229.[Medline]
Kushi LH, Mink PJ, Folsom AR, Anderson KE, Zheng W, Lazovich D and Sellers TA (1999) Prospective study of diet and ovarian cancer. Am J Epidemiol 149, 2131.[Abstract]
La Vecchia C, Decarli A, Franceschi S, Gentile A, Negri E and Parazzini F (1987) Dietary factors and the risk of breast cancer. Nutr Cancer 10, 205214.[Medline]
Levi F, Franceschi S, Negri E and La Vecchia C (1993) Dietary factors and the risk of endometrial cancer. Cancer 71, 35753581.[Medline]
Mangtani P and Booth M (1933) Epidemiology of endometriosis. J Epidemiol Commun Health 47, 8488.
Missmer SA and Cramer DW (2003) The epidemiology of endometriosis. Obstet Gynecol Clin N Am 30, 119.[Medline]
Mori M, Harabuchi I, Miyake H, Casagrande JT, Henderson BE and Ross RK (1988) Reproductive, genetic, and dietary risk factors for ovarian cancer. Am J Epidemiol 128, 771777.[Abstract]
Olive DL and Schwartz LB (1993) Endometriosis. New Engl J Med 328, 17591769.
Parazzini F, La Vecchia C, Franceschi S, Negri E and Cecchetti G (1989) Risk factors for endometrioid, mucinous and serous benign ovarian cysts. Int J Epidemiol 18, 108112.[Abstract]
Parazzini F, Negri E, La Vecchia C, Restelli C and Franceschi S (1992) Family history of reproductive cancers and ovarian cancr risk: an Italian case-control study. Am J Epidemiol 135, 3540.[Abstract]
Parazzini F, Ferraroni M, Fedele L, Bocciolone L, Rubessa S and Riccardi A (1995) Pelvic endometriosis: reproductive and menstrual risk factors at different stages in Lombardy, northern Italy. J Epidemiol Commun Health 49, 6164.[Abstract]
Perneger TV (1998) What's wrong with Bonferroni adjustments. Br Med J 316, 12361238.
Risch HA, Jain M, Marrett LD and Howe GR (1994) Dietary fat intake and risk of epithelial ovarian cancer. J Natl Cancer Inst 86, 14091415.[Abstract]
Signorello LB, Harlow BL, Cramer DW, Spiegelman D and Hill JA (1997) Epidemiologic determinants of endometriosis: a hospital-based case-control study. Ann Epidemiol 7, 267274.[CrossRef][Medline]
Smith MF (1986) Recent advances in corpus luteum physiology. J Dairy Sci 69, 911926.[Medline]
Turrini A, Saba A, Perrone D, Cialfa E and D'Amicis A (2001) Food consumption patterns in Italy: the INN-CA study 1994-1996. Eur J Clin Nutr 55, 571588.[CrossRef][Medline]
Willett W and Stampfer MJ (1986) Total energy intake: implications for epidemiologic analyses. Am J Epidemiol 124, 1727.[Abstract]
Submitted on July 25, 2003; accepted on November 20, 2003.
|