Human Reproduction, Vol.24, No.6 pp. 1501 1506, 2009 Advanced Access publication on March 11, 2009 doi:10.

1093/humrep/dep048

ORIGINAL ARTICLE Reproductive epidemiology

Relative weight at ages 10 and 16 years and risk of endometriosis: a case control analysis
C.M. Nagle 1,7, T.A. Bell 1,2, D.M. Purdie 3, S.A. Treloar 1,4, C.M. Olsen 1,5, S. Grover6, and A.C. Green1
1 Queensland Institute of Medical Research, PO Royal Brisbane Hospital, Herston, Brisbane 4006, Australia 2Faculty of Sciences, Engineering and Health, Central Queensland University, Bundaberg, Australia 3Genentech Inc., 1 DNA Way, South San Francisco, CA, USA 4Centre for Military and Veterans Health, The University of Queensland, Brisbane, Australia 5School of Population Health, University of Queensland, Brisbane, Australia 6Mercy Hospital for Women, Department of Obstetrics and Gynaecology, University of Melbourne, Melbourne, Australia 7

Corresponding author. E-mail: christina.nagle@qimr.edu.au

Downloaded from humrep.oxfordjournals.org by guest on October 4, 2011

background: Although previous epidemiological studies have shown that women with endometriosis are more likely to be thinner and underweight, it is currently not clear whether this is a true characteristic of women who develop endometriosis or a consequence of their disease and its symptoms. The aim of this study was to investigate the relationship between endometriosis and relative weight in childhood and adolescence, prior to diagnosis. methods: This case control study included 268 Australian women with surgically conrmed moderate to severe endometriosis (cases) and 244 women without endometriosis (controls). Relative weight at ages 10 and 16 years, as recalled and classied (underweight, average weight and overweight) separately by the women themselves and their mothers, was analyzed. results: Women who reported being overweight at 10 years had an increased risk of endometriosis (OR 2.8; 95% CI 1.17.5).
Mothers reports and concordant responses among mother daughter pairs were consistent with this association. There was no clear evidence of an association between relative weight at 16 years and risk of endometriosis.

conclusions: These data suggest that being overweight during late childhood is associated with the development of endometriosis; however, the results warrant conrmation in larger study populations.
Key words: endometriosis / relative weight / childhood / adolescence

Introduction
A number of characteristics of womens menstrual patterns and reproductive histories, such as shorter menstrual cycle lengths and longer ow (Cramer et al., 1986), earlier age at menarche (Moen and Schei, 1997; Signorello et al., 1997) and parity (Missmer et al., 2004a), have been associated with endometriosis. Other personal and lifestyle characteristics, in addition to certain environmental factors such as dioxin-like compounds (Heilier et al., 2008), have also been implicated, but most of these remain equivocal as risk factors. Many epidemiological studies evaluating the effects of putative risk factors that change over time (i.e. menstrual cycle patterns, weight) have collected data as current exposures in women affected by endometriosis (Cramer and Missmer, 2002). This is problematic since current personal characteristics and habits are likely to be highly associated with disease symptoms and even a consequence of

endometriosis rather than a preceding factor involved in its development. A more accurate approach to identifying factors that are causally related to endometriosis is to measure these factors in early life, prior to diagnosis, because they correctly reect the temporal relationship between exposures and the subsequent development of endometriosis (Cramer and Missmer, 2002). Previous studies have reported that women with endometriosis are more likely to be thinner and underweight (Darrow et al., 1993; McCann et al., 1993; Signorello et al., 1997; Hemmings et al., 2004; Missmer et al., 2004b: Ferrero et al., 2005; Hediger et al., 2005; Matalliotakis et al., 2008), but because these reports were based on current weight, it is difcult to determine whether this association is a true characteristic of women who develop endometriosis or a consequence of their disease symptoms. Although two recent studies (Missmer et al., 2004b; Hediger et al., 2005) have reported that women who were thinner and leaner in late

& The Author 2009. Published by Oxford University Press on behalf of the European Society of Human Reproduction and Embryology. All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.org

1502
adolescence have an increased risk of endometriosis, the potential role of relative weight in childhood and in early adolescence in the development of endometriosis remains an open question. The aim of this novel study was to determine whether relative weight in early life, as reported by cases, controls and their mothers, is associated with subsequent diagnosis of endometriosis. The structured data collection of our study enabled analysis of associations in relation to specic time periods of exposure, including in childhood and in adolescence, prior to diagnosis. This methodological approach should be highly relevant to evaluating the etiological relationship between relative weight and the development of endometriosis (Missmer and Cramer, 2003). This approach is also particularly important in light of recent suggestions of the possible role of in utero and childhood environment in the causation of endometriosis (Missmer et al., 2004c; Buck Louis et al., 2007).

Nagle et al.

were also asked to classify their daughters relative weight at the same ages and in the same pre-dened categories. Relative weights at 10 and 16 years were analyzed three ways: using participants reports of their own relative weight (self-report); each mothers report of her daughters relative weight (mothers report); and reports from both mothers and daughters where the reports between the two sources were in agreement (concordant pairs).

Statistical analysis
Multivariable logistic regression was used to quantify associations and calculate odds ratios (OR) and 95% condence intervals (CI). All analyses were adjusted for known confounding factors including state of residence, age at menarche and relative weight at 10 or 16 years (i.e. when relative weight at age 10 years was the variable of interest, it was adjusted for relative weight at age 16 years, and vice versa). Menstrual factors, smoking, alcohol consumption and other personal characteristics such as education were also considered as potential confounders but were not included in the nal models as they did not materially alter the effect estimates. The level of concordance between the responses of the participants and their mothers was examined using a weighted kappa and its 95% CI. Discordant responses were weighted by the square of the deviation from exact agreement (Landis and Koch, 1977; Graham and Jackson, 1993). All analyses were performed using the Statistical Packages for Social Sciences for Windows, Version 13.0 (SPSS Inc., Chicago, IL, USA). Ethics approval for this study was received from the Queensland Institute of Medical Research, Bancroft Research Ethics Committee, the Australian Twin Registry, and the University of Queensland, Medical Research Ethics Committee.

Materials and Methods

Study population and sample
Cases included 268 Australian women aged between 18 and 55 years, with surgically conrmed endometriosis, recruited between 1996 and 2002 through the Australian component of the International Endogene Study (Treloar et al., 2002). Cases were drawn from volunteer women, with no affected rst-degree relative and no known family history of endometriosis, who had been diagnosed with moderate to severe endometriosis (revised American Fertility Society stage III/IV) (Holt and Weiss, 2000; Zondervan et al., 2002). Of 310 affected women who were eligible, two were residing overseas and seven were unable to be contacted. Among the remaining 301 eligible cases, 268 (89%) participated. Potential control participants were pairs of female twins enrolled with the Australian Twin Registry. One twin per pair was randomly selected and frequency-matched one-to-one with cases based on age (5-year groups) and geographic location (urban/rural). In all, 511 women were contacted. Of these, 40 were ineligible because they were either residing overseas (n 11) or had a diagnosis of endometriosis (n 29) and a further 39 were unable to be contacted. Of the remaining 432 eligible women, 244 (56%) agreed to participate. Permission was obtained from participating cases and controls to contact their mothers in order to invite them also to participate in the study. Among the 268 cases, 220 (82%) gave permission for their mothers to be contacted, and of these 196 (90%) mothers of cases participated. Among the control group of 244 women, 144 (59%) gave permission for their mother to be contacted and 132 mothers (93%) participated.

Downloaded from humrep.oxfordjournals.org by guest on October 4, 2011

Results
There were 268 women with moderate to severe endometriosis and 244 controls included in the study. There was no difference in the mean ages of cases and controls, but state of residence did differ between the case and control groups (P , 0.001), with more cases residing in Queensland, where the study was based, and more controls in Victoria, where the twin registry is based, and in nearby Tasmania (Table I). A signicantly higher proportion of cases were born in Asia, Africa and Southern and Eastern Europe compared with controls (P 0.04), whereas controls had more years of schooling than cases (P 0.002). The average age at menarche for cases was 12.6 years [standard deviation (SD) +1.4 years] compared with 13.0 years (SD + 1.4 years) for controls (P 0.003), and 23% of cases experienced menarche before age 12 years compared with 13% of controls. Cases did not differ from controls with regard to marital status, religion, smoking status or alcohol consumption (Table I). Relative weight at age 10 years was reported by 503 participants and 324 mothers, whereas relative weight at 16 years was reported by 504 participants and 323 mothers. Relative weight at ages 10 and 16 years were available for 318 daughter mother pairs. There was moderate concordance between daughters and mothers reports of relative weight at age 10 years (weighted kappa 0.4, 95% CI 0.3 0.5), with cases and their mothers attaining a better level of agreement (weighted kappa 0.5) than controls and their mothers (weighted kappa 0.3) (Table II). Similarly, there was moderate concordance between self-reports and mothers reports of relative weight at 16 years (weighted kappa 0.4, 95% CI 0.3 0.5), and again there was a better level of agreement between cases and their mothers

Data collection and exposures

Cases and controls completed a comprehensive health and lifestyle questionnaire, which included questions about their personal details, menstrual cycle patterns and habits such as smoking and talc use, as well as physical characteristics. Cases and controls were asked to report their relative weight at ages 10 and 16 years. These ages were thought to best represent relative weight pre- and post-menarche, respectively (Morabia and Costanza, 1998; Thomas et al., 2001). Relative weight was described in three pre-specied categories: underweight, average weight and overweight. Participants were not asked to record weight in kilograms because of the potential biases of such data, particularly among women (Engstrom et al., 2003). The recall of childhood and adolescent relative weight has been shown in several studies to have reasonable validity (Must et al., 1993; Must et al., 2002). Mothers of cases and controls

Relative weight in early life and endometriosis

1503

Table I Selected demographic, personal and lifestyle characteristics of cases and controls
Casesa (n 5 268) Age (years) (mean + SD) State of residence Queensland [n (%)] New South Wales Victoria/Tasmania South Australia/Northern Territory Western Australia Ethnicity Australia/New Zealand North-West Europe Otherb Marital status Married/Defacto Never Married Divorced/Separated Education ,15 years of schooling 15 years of schooling Religion No religion Catholic Anglician Other Age at menarche ,12 years 1213 years 14 years Smoking status Lifelong non-smoker Current smoker Ex-smoker Alcohol consumption Never/occasional Infrequent Regular
a b

Controlsa (n 5 244) 36.3 + 7.6 35 (14) 50 (21) 113 (46) 25 (10) 21 (9) 222 (91) 18 (7) 4 (2) 182 (75) 44 (18) 17 (7) 26 (11) 214 (89) 76 (31) 61 (25) 44 (18) 63 (26) 31 (13) 126 (52) 83 (35) 147 (60) 46 (19) 50 (21) 47 (19) 150 (62) 47 (19)

Table II Level of agreement between self-reported and mothers report of relative weight at ages 10 and 16 years
Mothers reports Self-report

P-value

........................................................................................
36.4 + 7.1 117 (44) 61 (23) 54 (20) 13 (5) 23 (8) 242 (90) 12 (5) 14 (5) 197 (74) 38 (14) 32 (12) 56 (21) 210 (79) 67 (25) 77 (29) 60 (22) 64 (24) 61 (23) 143 (53) 64 (24) 139 (52) 52 (19) 77 (29) 39 (15) 164 (61) 65 (24) 0.61

........................................................................................
Relative weight at 10 years Underweight Average weight Overweight Underweight Average weight Overweight Relative weight at 16 years Underweight Average weight 20 36 0 5 218 5 0 23 11 24 44 0 6 221 4 1 12 6

,0.001

Overweight

0.04

Table III Relative weight (reported by participants) at ages 10 and 16 years and risk of endometriosis
Downloaded from humrep.oxfordjournals.org by guest on October 4, 2011

0.11

Casesa n (%) Relative weight at 10 years Self-reportc Underweight Average Overweight Mothers report Underweight Average Overweight Concordant pairs Underweight Average Overweight
c c

Controlsa n (%)

Adjustedb OR (95% CI)

.......................................................................................
0.002

54 (20) 187 (72) 20 (8) 21 (11) 165 (85) 7 (4) 16 (11) 131 (86) 5 (3)

58 (24) 176 (73) 8 (3) 11 (9) 117 (89) 3 (2) 8 (8) 90 (91) 1 (1)

0.9 (0.51.7) 1.0 2.8 (1.17.5) 1.2 (0.52.9) 1.0 2.1 (0.411.0) 0.6 (0.12.4) 1.0 6.2 (0.573.0)

0.29

0.003

0.08

Relative weight at 16 years Self-reportd Underweight Average Overweight Mothers reportd Underweight Average Overweight Concordant pairsd 21 (11) 159 (83) 12 (6) 17 (11) 129 (83) 9 (6) 4 (3) 122 (93) 5 (4) 3 (3) 89 (95) 2 (2) 5.9 (1.620.9) 1.0 1.0 (0.33.6) 8.1 (1.738.4) 1.0 2.2 (0.413.6) 48 (18) 185 (71) 29 (11) 41 (17) 167 (69) 34 (14) 1.4 (0.72.7) 1.0 0.5 (0.31.0)

0.21

Numbers may not sum to total because some data missing. Other includes Asia, Africa, Southern and Eastern Europe.

(weighted kappa 0.6) than controls and their mothers (weighted kappa 0.2) (Table II). The associations between relative weight at 10 and 16 years and endometriosis are shown in Table III. In multivariate analysis adjusting for potential confounding variables, self-report of being overweight at age 10 years was associated with subsequent development of endometriosis; the odds ratio for being overweight at 10 years compared with being average weight was 2.8 (95% CI 1.1 7.5). Analysis of

Underweight Average Overweight

a b

Numbers may not sum to total because some data missing. Adjusted for age at menarche; state of residence. c Additionally adjusted for relative weight at 16 years. d Additionally adjusted for relative weight at 10 years.

1504
mothers reports and concordant responses among motherdaughter pairs were also suggestive of a positive association between being overweight at 10 years and endometriosis, but these results did not reach statistical signicance. Being underweight at 10 years (as reported by self, mothers and concordant mother daughter pairs) was not associated with a subsequent diagnosis of endometriosis. There was no association between self-report of being underweight at age 16 years and endometriosis (OR 1.4, 95% CI 0.7 2.7). Similar results were noted when we restricted the analysis by the age of onset of rst symptoms to exclude women who reported rst symptoms before the age 16 years. However, there was a suggestion of a positive association between mothers and concordant mother daughter pair reports of being underweight at 16 years and endometriosis. We found no signicant association between a reported history of being overweight at age 16 years (as reported by self, mothers and concordant mother daughter pairs) and endometriosis.

Nagle et al.

Discussion
We have presented novel data from a case control analysis of the associations between relative weight in childhood and in adolescence (before diagnosis) and subsequent development of endometriosis. We found that women who reported being overweight in childhood at age 10 years were more likely to be diagnosed with endometriosis. We found no clear evidence of an association between relative weight at 16 years (under- or over-weight) and endometriosis. There are few published data pertaining to the association between body size or weight in adolescence and development of endometriosis (Missmer et al., 2004b; Hediger et al., 2005). One large prospective study found that women with a low body mass index (BMI) (,19) at 18 years had a higher rate of endometriosis than other women (Missmer et al., 2004b). Likewise, a recent study of 84 women (32 cases and 52 controls) found that women diagnosed with endometriosis reported a signicantly leaner body shape and size during adolescence (15 19 years) and young adulthood (20 24 years) (Hediger et al., 2005). We found no signicant association between relative weight at 16 years and endometriosis. Our principal nding of an increased risk of endometriosis among women who are overweight at 10 years is consistent with current etiological hypotheses. A single hypothesis to explain the pathogenesis of endometriosis has not yet been elucidated, despite extensive research. Theories of retrograde menstruation and implantation of endometrial cells, probably through errors in the immune response, remain dominant, although evidence also shows that estradiol plays an important role in stimulating the development of endometriosis (Cramer and Missmer, 2002; Batt and Mitwally, 2003; Baldi et al., 2008). We can speculate on reasons for the observed positive relationship between being overweight at 10 years and endometriosis. Plump girls on average reach menarche earlier than their thin peers (Kaplowitz et al., 2001; Blell et al., 2008), so childhood body fatness could be assumed to increase exposure to menstrual ow and subsequently increase the risk of endometriosis. Furthermore, increased levels of bioavailable estradiol, which may result from increased production of estrogen by aramotase in adipose tissue could also explain the link between overweight in early life and endometriosis. Accumulating evidence supports a role of overweight/obesity in the etiology of many chronic diseases including inammatory and

immune-mediated disorders (Calle and Kaaks, 2004; Wellen and Hotamisiligil, 2005; Mannino et al., 2006). The distinct possibility exists that overweight/obesity may be linked to endometriosis through altered adipocytokine secretion and circulating levels, in particular adiponectin and leptin. These adipocytokines are known to have inammatory and angiogenic actions (Tilg and Moschen, 2006), factors widely believed to be central to the development of endometriosis (Agic et al., 2006; May and Becker, 2008). Several other lines of evidence indicate a potential role for adiponectin and leptin in the pathogenesis of endometriosis (Takemura et al., 2005; Milewski et al., 2008; Styer et al., 2008). One of the strengths of this study was the use of time period data collection points [i.e. relative weight data collected before menarche (age 10 years) and after menarche (age 16 years)] which methodologically aimed to capture the timing of any causal relationship between relative weight and endometriosis. We found that the mean age at menarche was signicantly later in controls than cases and to ensure that our results did not simply reect an association between age of menarche and relative weight, all multivariate analyses were adjusted for age at menarche. Our models for childhood relative weight were also adjusted for relative weight at 16 years (and vice versa) because these two variables are correlated. Given that a large proportion of women with endometriosis have symptoms that begin during early adolescence (Greene et al., 2009), we repeated the analysis for relative weight at age 16 years restricted to women with no symptoms at this age, and effect estimates were unchanged. The ndings do, however, require cautious interpretation. As with all studies of endometriosis, the choice of control group is likely to inuence the value of the study ndings. Our control group comprised single-sex twins who had volunteered to participate in Australian Twin Registry. Although we know that the childhood weight of twins is similar to that of singletons by age 9 years (Leon, 2001), it is possible that some volunteer bias remains. Another potential limitation was the relatively low participation rate among controls (56%). To assess the possible effect of this, we compared the distribution of key variables in our participating control group with data from the non-participating controls. We found that the distributions of key exposures (age, state of residence, education, age at menarche, smoking status and alcohol consumption) among participating and non-participating control women were not notably different. Unfortunately no exposure information (childhood/adolescent weight) was available for the nonparticipating controls, so the role of selection bias may explain the observed association between being overweight at 10 years and endometriosis. Even though our controls were women who reported no past history of endometriosis, it is possible that women with undiagnosed disease may have been included in our control group. However, as the community prevalence of more severe stages of endometriosis is less than 2% (Zondervan et al., 2002), an estimated maximum of ve women with undiagnosed (more severe stage) disease may have been inadvertently included in our control group. These would be highly unlikely to inuence our effect estimates. Although the response rate for cases was high (89%), 33 eligible cases did not participate. Of these women, three actively refused to participate, whereas the remaining 30 women did not return questionnaires despite initial agreement in some cases. Comparison of sociodemographic factors (age, state, residential area) showed no signicant differences between participating and non-participating

Downloaded from humrep.oxfordjournals.org by guest on October 4, 2011

Relative weight in early life and endometriosis

1505
Cramer DW, Missmer SA. The epidemiology of endometriosis. Ann N Y Acad Sci 2002;955:11 22. Cramer DW, Wilson E, Stillman RJ, Berger MJ, Belisle S, Schiff I, Albrecht B, Gibson M, Stadel BV, Schoenbaum SC. The relation of endometriosis to menstrual characteristics, smoking, and exercise. JAMA 1986;255:1904 1908. Darrow SL, Vena JE, Batt RE, Zielezny MA, Michalek AM, Selman S. Menstrual cycle characteristics and the risk of endometriosis. Epidemiology 1993;4:135 142. Engstrom JL, Paterson SA, Doherty A, Trabulsi M, Speer KL. Accuracy of self-reported height and weight in women: an integrative review of the literature. J Midwifery Womens Health 2003;48:338 345. Ferrero S, Anserini P, Remorgida V, Ragni N. Body mass index in endometriosis. Eur J Obstet Gynecol Reprod Biol 2005;121:94 98. Graham P, Jackson R. The analysis of ordinal agreement data: beyond weighted kappa. J Clin Epidemiol 1993;46:1055 1062. Greene R, Stratton P, Cleary SD, Ballweg ML, Sinaii N. Diagnostic experience among 4,334 women reporting surgically diagnosed endometriosis. Fertil Steril 2009;91:32 39. Hediger ML, Hartnett HJ, Louis GMB. Association of endometriosis with body size and gure. Fertil Steril 2005;84:1366 1374. Heilier JF, Donnez J, Lison D. Organochlorines and endometriosis: a mini-review. Chemosphere 2008;71:203 210. Hemmings R, Rivard M, Olive DL, Poliquin-Fleury J, Gagne D, Hugo P, Gosselin D. Evaluation of risk factors associated with endometriosis. Fertil Steril 2004;81:1513 1521. Holt VL, Weiss NS. Recommendations for the design of epidemiologic studies of endometriosis. Epidemiology 2000;11:654 659. Kaplowitz PB, Slora EJ, Wasserman RC, Pedlow SE, Herman-Giddens ME. Earlier onset of puberty in girls: relation to increased body mass index and race. Pediatrics 2001;108:347 353. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1977;33:159 174. Leon DA. The foetal origins of adult disease: interpreting the evidence from twin studies. Twin Res 2001;4:321 326. Mannino DM, Mott J, Ferdinands JM, Camargo CA, Friedman M, Greves HM, Redd SC. Boys with high body masses have an increased risk of developing asthma: ndings from the National Longitudinal Survey of Youth (NLSY). Int J Obes (Lond) 2006;30:6 13. Matalliotakis IM, Cakmak H, Fragouli YG, Goumenou AG, Mahutte NG, Arici A. Epidemiological characteristics in women with and without endometriosis in the Yale series. Arch Gynecol Obstet 2008; 277:389 393. May K, Becker C. Endometriosis and angiogenesis. Minerva Ginecol 2008; 60:25 54. McCann SE, Freudenheim JL, Darrow SL, Batt RE, Zielezny MA. Endometriosis and body fat distribution. Obstet Gynecol 1993; 82:545 549. Milewski L, Barcz E, Dziunycz P, Radomski D, Kaminski P, Roszkowski PI, Korczak-Kowalska G, Malejczyk J. Association of leptin with inammatory cytokines and lymphocyte subpopulations in peritoneal uid of patients with endometriosis. J Reprod Immunol 2008; 79:111 117. Missmer SA, Cramer DW. The epidemiology of endometriosis. Obstet Gynecol Clin North Am 2003;30:1 19. Missmer SA, Hankinson SE, Spiegelman D, Barbieri RL, Malspeis S, Willett WC, Hunter DJ. Reproductive history and endometriosis among premenopausal women. Obstet Gynecol 2004a;104:965 974. Missmer SA, Hankinson SE, Spiegelman D, Barbieri RL, Marshall LM, Hunter DJ. Incidence of laparoscopically conrmed endometriosis by demographic, anthropometric, and lifestyle factors. Am J Epidemiol 2004b;160:784 796.

cases. Another potential source of bias in this study results from the fact that cases were volunteers. Although we know that individuals who self-select into studies often differ in key ways from the target or source population, we have attempted to account for the potential biases of volunteers by choosing a comparison volunteer control group. Misclassication of exposure, as reported by mothers and mother daughter pairs, is also a potential source of bias since fewer controls than cases gave permission to contact their mothers. Although such misclassication is likely to be non-differential, the focus of our interpretation rests with the results from the primary source, namely self-reports of relative weight, rather than mothers and mother daughter pairs. Study data on relative weight at 10 and 16 years were self-reported, and although the correlation between cases, controls and mothers reports of relative weight at these ages offers some support to the validity of these measures, we are aware that data of this type are subject to recall error whereby women with higher BMI are more likely to underestimate body size at young ages, whereas underweight women are more likely to overestimate body size (Troy et al., 1995; Tehard et al., 2002; Nyholm et al., 2007). This might have led to non-differential misclassication that could have attenuated the true association between relative weight and endometriosis. Finally, the extent to which we have adequately addressed confounding factors is unknown, but every effort was made to measure and adjust for all factors currently thought to be related to the development of endometriosis. In summary, we have provided new evidence that high relative weight in childhood is associated with the subsequent development of endometriosis. Conrmation of these ndings and understanding the biological relations underpinning these observations is paramount.

Downloaded from humrep.oxfordjournals.org by guest on October 4, 2011

Funding
C.M.N. and T.A.B. were funded by Fellowships from the National Health and Medical Research Council of Australia and C.M.O. was funded by the University of Queensland Postdoctoral fellowship for Women. S.A.T. was supported by the Cooperative Research Centre for Discovery of Genes for Common Human Diseases and by NHMRC (339430).

References
Agic A, Xu H, Finas D, Banz C, Diedrich K, Hornung D. Is endometriosis associated with systemic subclinical inammation? Gynecol Obstet Invest 2006;62:139 117. Baldi A, Campioni M, Signorile PG. Endometriosis: pathogenesis, diagnosis, therapy and association with cancer (review). Oncol Rep 2008; 19:843 846. Batt RE, Mitwally MF. Endometriosis from thelarche to midteens: pathogenesis and prognosis, prevention and pedagogy. J Pediatr Adolesc Gynecol 2003;16:337 347. Blell M, Pollard TM, Pearce MS. Predictors of age at menarche in the newcastle thousand families study. J Biosoc Sci 2008;40:563 575. Buck Louis GM, Hediger ML, Pena JB. Intrauterine exposures and risk of endometriosis. Hum Reprod 2007;22:3232 3236. Calle EE, Kaaks R. Overweight, obesity and cancer: epidemiological evidence and proposed mechanisms. Nat Rev Cancer 2004;4:579 591.

1506
Missmer SA, Hankinson SE, Spiegelman D, Barbieri RL, Michels KB, Hunter DJ. In utero exposures and the incidence of endometriosis. Fertil Steril 2004c;82:1501 1508. Moen MH, Schei B. Epidemiology of endometriosis in a Norwegian county. Acta Obstet Gynecol Scand 1997;76:559 562. Morabia A, Costanza MC. International variability in ages at menarche, rst livebirth, and menopause. World Health Organization Collaborative Study of Neoplasia and Steroid Contraceptives. Am J Epidemiol 1998; 148:1195 1205. Must A, Willett WC, Dietz WH. Remote recall of childhood height, weight, and body build by elderly subjects. Am J Epidemiol 1993;138:5664. Must A, Phillips SM, Naumova EN, Blum M, Harris S, Dawson-Hughes B, Rand WM. Recall of early menstrual history and menarcheal body size: after 30 years, how well do women remember? Am J Epidemiol 2002; 155:672 679. Nyholm M, Gullberg B, Merlo J, Lundqvist-Persson C, Rastam L, Lindblad U. The validity of obesity based on self-reported weight and height: Implications for population studies. Obesity (Silver Spring) 2007; 15:197 208. Signorello LB, Harlow BL, Cramer DW, Spiegelman D, Hill JA. Epidemiologic determinants of endometriosis: a hospital-based case control study. Ann Epidemiol 1997;7:267 741. Styer AK, Sullivan BT, Puder M, Arsenault D, Petrozza JC, Serikawa T, Chang S, Hasan T, Gonzalez RR, Rueda BR. Ablation of leptin signaling disrupts the establishment, development, and maintenance of endometriosis-like lesions in a murine model. Endocrinology 2008; 149:506 514.

Nagle et al.

Takemura Y, Osuga Y, Harada M, Hirata T, Koga K, Yoshino O, Hirota Y, Morimoto C, Yano T, Taketani Y. Concentration of adiponectin in peritoneal uid is decreased in women with endometriosis. Am J Reprod Immunol 2005;54:217 221. Tehard B, van Liere MJ, Com Nougue C, Clavel-Chapelon F. Anthropometric measurements and body silhouette of women: validity and perception. J Am Diet Assoc 2002;102:1779 1784. Thomas F, Renaud F, Benece E, de Meeus T, Guegan JF. International variability of ages at menarche and menopause: patterns and main determinants. Hum Biol 2001;73:271 290. Tilg H, Moschen AR. Adipocytokines: mediators linking adipose tissue, inammation and immunity. Nat Rev Immunol 2006;6:772 783. Treloar S, Hadeld R, Montgomery G, Lambert A, Wicks J, Barlow DH, OConnor DT, Kennedy S, International Endogene Study. The international endogene study: a collection of families for genetic research in endometriosis. Fertil Steril 2002;78:679 685. Troy LM, Hunter DJ, Manson JE, Colditz GA, Stampfer MJ, Willett WC. The validity of recalled weight among younger women. Int J Obes Relat Metab Disord 1995;19:570 572. Wellen KE, Hotamisligil GS. Inammation, stress, and diabetes. J Clin Invest 2005;115:1111 1119. Zondervan KT, Cardon LR, Kennedy SH. What makes a good case control study? Design issues for complex traits such as endometriosis. Hum Reprod 2002;17:1415 1423.
Submitted on August 18, 2008; resubmitted on February 2, 2009; accepted on February 5, 2009

Downloaded from humrep.oxfordjournals.org by guest on October 4, 2011