Uterine leiomyomata are common and concerning. Leiomyomata have been associated with adverse pregnancy outcomes including difficulty conceiving, spontaneous abortion, preterm birth, placental abruption, and cesarean delivery.1–8 By age 35, among nonpregnant women, more than 60% of African-American women and almost 40% of white women have leiomyomata identifiable by imaging.9 Thus, overall prevalence in pregnancy could average 10% to 20%, with higher prevalence in women of advancing maternal age. Little is known from prospective studies in pregnant populations that reflect the general population of the United States about either prevalence of leiomyomata or their reproductive risks.10 Available information has been gathered from subgroups, such as women seeking care for infertility, from European populations, and from ultrasound databases of academic medical centers.11–15
Advances in ultrasonography have made some prior operational definitions obsolete. The Muram criteria have been used widely in research to define the presence of a leiomyoma by ultrasonography. The original criteria included visualization of a spherical mass with a diameter of 3 cm or more, distortion of the adjacent myometrium by the mass, and a distinctive echogenicity differentiating the mass from myometrium.16 Improvements in ultrasound resolution suggest we now can identify confidently and measure smaller leiomyomata. Prior literature reports that the prevalence of leiomyomata in pregnancy is less than 1% to 5%.8,14,17–20 A low prevalence in these studies may result from 1) operational definitions requiring a large diameter to define presence of leiomyomata, 2) inconsistent documentation of leiomyomata in clinical databases subsequently used for research, 3) difficulty detecting leiomyomata as pregnancy progresses and uterine anatomy, fetus, and placenta interfere with complete assessment of the myometrium, or 4) study populations that are highly selected for younger women.
Using a prospective study design with community-based recruitment and research-quality ultrasonography for all participants, we sought to: 1) estimate what proportion of women have leiomyomata 0.5 cm or more in maximum diameter in the first trimester of pregnancy, 2) describe the size, type, and location of leiomyomata identified, and 3) report any variation by race/ethnicity.
MATERIALS AND METHODS
Right from the Start is a prospective, cohort study of pregnancy that includes women from four metropolitan areas in three states (North Carolina, Tennessee, and Texas). Institutional review board approval was obtained from each of four participating academic institutions. Women included in this analysis were recruited from 2001 to 2007 using multiple approaches, including print materials in community practices, direct home mailings, pregnancy test kit coupons at pharmacies, newsletters, and advertisements.21 The study is described in outreach materials as a study of early-pregnancy health and never has been advertised as a study about leiomyomata in pregnancy. Women were eligible to participate if they were 18 years or older, enrolled by 12 6/7 weeks of gestation based on last menstrual period, did not use assisted reproductive technology, intended to carry the pregnancy to term, spoke English or Spanish, and did not plan to move for the next 18 months. Women could reenroll for subsequent pregnancies, but only first enrollments with ultrasound data on leiomyoma status were included in this analysis.
The study was designed to enroll women very early in pregnancy. Women who were planning a pregnancy were preenrolled for up to 6 months and were enrolled once they reported a pregnancy (n=858); the balance enrolled in the first trimester. An abbreviated interview was done at intake, and a more detailed, computer-assisted telephone interview was conducted in the first trimester, including extensive reproductive and medical history. Women self-reported race and ethnicity. This interview also gathered information about diagnosis and treatment of leiomyomata before the current pregnancy.
Endovaginal ultrasonography without Doppler (supplemented if needed by transabdominal images), was scheduled for all participants, aiming for the beginning of the sixth week of gestation. Beginning in January 2005, participants who reported a pregnancy loss before the scheduled ultrasound examination also were invited to have ultrasonography; this was scheduled within 2 weeks of their report (n=56). Sonographers at each study site were required to have 3 or more years of clinical obstetric–gynecologic experience. They received additional research training on uniformity in identifying and measuring leiomyomata and were instructed not to discuss leiomyoma history with participants.
Presence of a leiomyoma was defined by the Muram criteria with modification to include masses of maximum diameter of 0.5 cm or more.16 When a leiomyoma was identified, the diameter was measured in three perpendicular planes. During the ultrasound examination, sonographers returned twice more to each leiomyoma to record the same measurements (intervening time was used to record gestational structures). The triplicate measures were intended to reduce the chance that focal contractions would be misclassified as leiomyomata. Leiomyoma diameters were averaged across all three measures, and a mean diameter was calculated for each leiomyoma. We refer to mean diameter of the largest leiomyoma when we use the term leiomyoma size.
Leiomyomata were “mapped” onto a uterine diagram and were categorized by location (fundus, corpus, lower segment), position (anterior, posterior, right, and left), and type. Type was defined in mutually exclusive categories: submucous—any leiomyoma in contact with or distorting the uterine cavity without identifiable myometrium between the leiomyoma and the endometrium; subserous—distorting the external contour of the uterus; intramural—within the myometrium, distorting neither contour nor cavity; and pedunculated—attached to the uterus with an identifiable stalk. Women with multiple leiomyomata had each leiomyoma documented separately. Leiomyoma and fetal images were saved initially as still print images and later as digital images on CD-ROM and sent to the study office for review by study investigators.
For analyses describing prevalence of leiomyomata, we adjusted for correlation within the three study sites (North Carolina, Tennessee, and Texas) using generalized estimating equations (22, 23). For analyses of fibroid size, we used the logarithm of the diameter for normalization and the geometric means are reported. Mixed models24 were used to adjust for correlations within study site when describing fibroid size among women with leiomyomas present. Age-related changes in leiomyoma prevalence, multiplicity, and size were not examined for Hispanic women because of limited numbers. The χ2 test was used for comparing categorical data on leiomyomata (type, location, and position) by ethnicity. We used two-sided significance testing and considered P<.05 statistically significant in presentation of results. Data were analyzed with SAS 9.1 (SAS Institute, Cary, NC) and Stata 10.0 (Stata Corp., College Station, TX).
We recruited 4,582 women, of whom 4,271 had ultrasound examinations and were enrolled for the first time (Table 1). Women who did not complete ultrasound examinations (n=311) were similar to those who did in terms of maternal age, race/ethnicity, parity, marital status, and education level. Ultrasound examinations were completed an average of 61 days±13 days from self-reported last menstrual period. Leiomyoma prevalence did not differ by gestational age at ultrasound examination (P=.3).
The prevalence of one or more leiomyomata was 10.7% (95% confidence interval [CI] 8.5–13.6). Among 458 women with one or more leiomyomata, we identified a total of 687 leiomyomata. Maximum leiomyoma diameter ranged from 0.5 cm to 12.9 cm. The mean size of the women’s largest leiomyoma was 2.3 cm (95% CI 1.8–2.8 cm). Subserosal (42%) and intramural (35%) leiomyomata were most common; 17% were submucous, and 5% were pedunculated. Nearly half of all leiomyomata occurred in the uterine corpus, and 35% were in the fundus. Leiomyomata were distributed evenly in the anterior and posterior as well as on the right and left sides of the uterus (P=.2). Leiomyoma type and location did not differ between African-American women and white women (P≥.1 for each comparison), except that African-American women had slightly more leiomyomata in the posterior wall of the uterus than did white women (P=.03). Seventy-two percent of women with a leiomyoma did not report a diagnosis of leiomyomata before this pregnancy.
Although overall leiomyoma prevalence was 10.7%, prevalence differed by race/ethnicity. African-American women (n=915) had a prevalence of 18% (95% CI 13–25), white women (n=2,826) 8% (95% CI 7–11), Hispanic women (n=335) 10% (95% CI 5–19), and women in the “other” group (predominantly Asian [n=186]) 13% (95% CI 10–16). Leiomyomata were present in women as young as 19 years old. For African-American and white women, we had sufficient numbers of participants to examine differences with age. Prevalence among those under 25 years old was 6% (95% CI 3–10) in African-American women and 4% (95% CI 3–7) in white women. Prevalence increased with age for both African-American and white women (Fig. 1). The rise in prevalence by age was steeper among African-American women (P=.02).
Among those with leiomyomata, having two or more leiomyomata also differed by race/ethnicity: 39% of African-American women with leiomyomata had multiple tumors (95% CI 37–42), 20% of white women (95% CI 20–21), and 22% of Hispanic women (95% CI 11–42). The presence of multiple tumors increased with age for African-American women, from 22% (95% CI 14–35) for women younger than 30 to 58% (CI 52–65) for women 35 years and older. For white women, the proportion with multiple tumors increased slightly across age, from 19% (95% CI 16–22) in women younger than 30 to 25% (95% CI 22–29) for women 35 years and older (Fig. 2). This difference in the proportion of women with multiple tumors as age advances was significantly different between African-American women and white women (P=.005).
The average size of the women’s largest leiomyoma differed by race/ethnicity as well. Average size was 2.5 cm (95% CI 2.1–3.1) for African-American women. This was not different from the average leiomyoma size for Hispanic women (2.4 cm, 95% CI 1.8–3.1) but was significantly larger than the average leiomyoma size for white women (2.0 cm, 95% CI 1.6–2.4, P<.001). Leiomyoma size tended to be larger in older compared with younger African-American women, but there was little variation in size with age for white women. However, the difference between African-American women and white women was not statistically significant (P=.08).
Prevalence estimates are determined intrinsically by the operational definition of the size required to classify a mass as a leiomyoma. If we were to define prevalence based on the original Muram criteria, restricting leiomyoma diameter to 3.0 cm or more, prevalence would be 7% in African-American women (95% CI 5–10) and 3% in white women (95% CI 2–3) (P<.001). Another common cutpoint, requiring a diameter of 1.0 cm or more, results in a prevalence of 16% in African-American women (95% CI 12–22) and 7% in white women (95% CI 6–9) (P<.001).
We have estimated the prevalence of leiomyomata in pregnancy by screening women uniformly during the first trimester. Our findings indicate a higher prevalence than previously reported. These higher estimates are more compatible with those required to reach the prevalence documented in imaging studies of older, nonpregnant women. We have confirmed that leiomyomata are common and occur with increasing frequency with advancing age.
Our data on race/ethnicity, for a younger population than that previously screened for leiomyomata, are consistent with prior reports. We found that African-American women are more likely to have leiomyomata and to have more numerous and larger leiomyomata. Hispanic women appeared to have leiomyoma characteristics similar to those of white women; however, screening of larger groups of Hispanic women will be required.
The majority of our participants were unaware of their leiomyoma status before the study ultrasound examination. Thus, we do not believe the sample was biased by a predominance of women with leiomyomata seeking participation. However, owing to the nature of this cohort, with predominantly planned pregnancies, we were not surprised that the participants were older, more educated, less likely to smoke, and more likely to be married than the general population.25 Because our participants were pregnant, our cohort is limited to fertile women. Leiomyomata may be associated with impaired fertility.7 If so, our findings would underestimate the true prevalence in this age group. Leiomyomata also might be related to pregnancy loss. If so, this too will result in an underestimate of tumor prevalence because those with a loss before their scheduled ultrasound examination were invited to have ultrasound examinations only in the final few years of data collection, and those invited were less likely to keep postloss ultrasound appointments than their pregnant counterparts. The current data set includes only 56 of these women.
In 1970, the median age of women who gave birth in the United States was 25.4 years; in 2000 it was 27.1.26 This shift in the demographics of maternity makes a more nuanced understanding of leiomyomata in pregnancy of increasing importance because larger proportions of women and their pregnancies are affected. We lack knowledge about the precise relationship of leiomyomata to adverse pregnancy outcomes and the risks and benefits of leiomyoma intervention in reproductive-age women.10 Additionally, African-American women experience a disproportionate share of pregnancy complications. Potentially as a result of the higher prevalence of and larger leiomyomata, they also have increased likelihood of surgical interventions for leiomyomata. To provide the best informed care, the research community must continue to pursue an understanding of the degree to which racial and ethnic differences in prevalence of leiomyomata contribute to disparities in events such as miscarriage and prematurity. If leiomyomata do increase adverse pregnancy outcomes, determining which interventions to use in the treatment of leiomyomata may reduce poor outcomes and help mitigate disparities.
1.Benson CB, Chow JS, Chang-Lee W, Hill JA 3rd, Doubilet PM. Outcome of pregnancies in women with uterine leiomyomas identified by sonography in the first trimester. J Clin Ultrasound 2001;29:261–4.
2.Coronado GD, Marshall LM, Schwartz SM. Complications in pregnancy, labor, and delivery with uterine leiomyomas: a population-based study. Obstet Gynecol 2000;95:764–9.
3.Katz VL, Dotters DJ, Droegemeuller W. Complications of uterine leiomyomas in pregnancy. Obstet Gynecol 1989;73:593–6.
4.Klatsky PC, Tran ND, Caughey AB, Fujimoto VY. Fibroids and reproductive outcomes: a systematic literature review from conception to delivery. Am J Obstet Gynecol 2008;198:357–66.
5.Lev-Toaff AS, Coleman BG, Arger PH, Mintz MC, Arenson RL, Toaff ME. Leiomyomas in pregnancy: sonographic study. Radiology 1987;164:375–80.
6.Ouyang DW, Economy KE, Norwitz ER. Obstetric complications of fibroids. Obstet Gynecol Clin North Am 2006;33:153–69.
7.Practice Committee of the American Society for Reproductive Medicine. Myomas and reproductive function. Fertil Steril 2006;86:S194–9.
8.Vergani P, Ghidini A, Strobelt N, Roncaglia N, Locatelli A, Lapinski RH, et al. Do uterine leiomyomas influence pregnancy outcome? Am J Perinatol 1994;11:356–8.
9.Day Baird D, Dunson DB, Hill MC, Cousins D, Schectman JM. High cumulative incidence of uterine leiomyoma in black and white women: ultrasound evidence. Am J Obstet Gynecol 2003;188:100–7.
10.Viswanathan M, Hartmann K, McKoy N, Stuart G, Rankins N, Thieda P, et al. Management of uterine fibroids: an update of the evidence. Evid Rep Technol Assess (Full Rep) 2007;154:1–122
11.Bulletti C, DE Ziegler D, Levi Setti P, Cicinelli E, Polli V, Stefanetti M. Myomas, pregnancy outcome, and in vitro fertilization. Ann N Y Acad Sci 2004;1034:84–92.
12.Davis JL, Ray-Mazumder S, Hobel CJ, Baley K, Sassoon D. Uterine leiomyomas in pregnancy: a prospective study. Obstet Gynecol 1990;75:41–4.
13.Eldar-Geva T, Meagher S, Healy DL, MacLachlan V, Breheny S, Wood C. Effect of intramural, subserosal, and submucosal uterine fibroids on the outcome of assisted reproductive technology treatment. Fertil Steril 1998;70:687–91.
14.Exacoustòs C, Rosati P. Ultrasound diagnosis of uterine myomas and complications in pregnancy. Obstet Gynecol 1993;82:97–101.
15.Stovall DW, Parrish SB, Van Voorhis BJ, Hahn SJ, Sparks AE, Syrop CH. Uterine leiomyomas reduce the efficacy of assisted reproduction cycles: results of a matched follow-up study. Hum Reprod 1998;13:192–7.
16.Muram D, Gillieson M, Walters JH. Myomas of the uterus in pregnancy: ultrasonographic follow-up. Am J Obstet Gynecol 1980;138:16–9.
17.Parker WH. Etiology, symptomatology, and diagnosis of uterine myomas. Fertil Steril 2007;87:725–36.
18.Qidwai GI, Caughey AB, Jacoby AF. Obstetric outcomes in women with sonographically identified uterine leiomyomata. Obstet Gynecol 2006;107:376–82.
19.Rice JP, Kay HH, Mahony BS. The clinical significance of uterine leiomyomas in pregnancy. Am J Obstet Gynecol 1989;160:1212–6.
20.Strobelt N, Ghidini A, Cavallone M, Pensabene I, Ceruti P, Vergani P. Natural history of uterine leiomyomas in pregnancy. J Ultrasound Med 1994;13:399–401.
21.Promislow JH, Makarushka CM, Gorman JR, Howards PP, Savitz DA, Hartmann KE. Recruitment for a community-based study of early pregnancy: the Right From The Start study. Paediatr Perinat Epidemiol 2004;18:143–52.
22.Liang, K-Y, Zeger SL. Longitudinal data analysis using generalized linear models. Biometrika 1986;73:13–22.
23.Zeger SL, Liang KY. Longitudinal data analysis for discrete and continuous outcomes. Biometrics 1986;42:121–30.
24.Laird NM, Ware JH. Random-effects models for longitudinal data. Biometrics 1982;38:963–74.
26.Mathews TJ, Hamilton BE. Mean age of mother, 1970-2000. Natl Vital Stat Rep 2002;51:1–13.