It is currently understood that pelvic organ prolapse develops as a result of uniform failure of pelvic floor support. The key risk factors for pelvic floor insufficiency are pregnancy, child birth, advancing age, obesity, and genetic predisposition.1,2 Some bulging of vaginal walls in parous women is a physiological phenomenon that does not cause any symptoms or require any treatment. In contrast, the descending uterus, with or without vaginal wall prolapse, may cause a number of troublesome symptoms (eg, bulging, bladder, bowel, and sexual dysfunction), and in Finland approximately 12% of women face a life-time risk for prolapse-related vaginal hysterectomy, with or without vaginal wall repair.3
Receptors for estrogen (ER) and progesterone (PR) are present in the pelvic floor (eg, musculature, pubocervical fascia, and uterosacral ligaments). These receptors mediate the effects of estrogen and progesterone on collagens, elastin, smooth muscle cells, and fibroblasts that are crucial for pelvic floor support.4-9 The possible molecular mechanisms are complex, but they are likely linked to extracellular regulators, such as transforming growth factor-β (TGF-β) and tropoelastin which estradiol modulates.9 Estradiol may also affect the structure of extracellular matrix.8 The impact of sex steroids on the pelvic floor can be seen, for example, during pregnancy, when high circulating levels of estrogens and progesterone, alone or in combination with relaxin, soften the pelvic floor to facilitate the forthcoming delivery.10 Sensitivity for steroid hormones in the pelvic floor gains additional support from the rapid development of pelvic organ prolapse in women using some new selective estrogen receptor modulators in initial/phase III clinical studies.11,12
The use of postmenopausal systemic hormone therapy also stimulates steroid hormone receptors in the pelvic floor.4-8 Based on the Cochrane analysis, the impact of postmenopausal hormone therapy on the pelvic floor support is, however, still unclear.13 In the majority of these studies, conjugated equine estrogens (CEE), rather than estradiol, have been used. Estradiol, however, differs in many biological effects from CEE,14 and thus, the impact of estradiol containing hormone therapy on pelvic organ prolapse needs to be assessed separately. Progestogens, of which noretisterone acetate (NETA) and medroxyprogesterone acetate (MPA) are the most popular, at least in Finland,15 also differ in many biological effects,16 but their possible differences in the regulation of the pelvic floor are not known. Likewise, no data exist for the levonorgestrel releasing intrauterine device (Levo-IUD) or tibolone in possible connection with pelvic organ prolapse. Thus, in the present study, we evaluated the impact of various systemic postmenopausal combined hormone therapy regimens (EPT) containing estradiol and progestogen and that of tibolone on the risk of uterine prolapse (UP) in Finnish women.
PARTICIPANTS AND METHODS
We collected from the Nationwide Hospital Discharge Register 12,072 women who had undergone vaginal hysterectomy for UP from 1998 to 2012. Although the uterus was the dominant prolapsing organ, hysterectomy could have been accompanied by vaginal repair, but these additional surgeries were not recorded into our data registry. The cases had not undergone any previous vaginal operation, and the index operations had been performed in 19 hospital districts located evenly throughout Finland
For each woman with UP three control women without a history of hysterectomy or any prolapse operation were collected from the Finnish National Population Register. The control women (N = 36,216) were matched in regards to age (±1 mo), number of deliveries, and hospital district. A number of selected women (N = 2,512) failing to fulfill inclusion criteria were excluded. Thus, the control group consisted of 33,704 women (Table 1).
The use of postmenopausal hormone therapy
Because all the cases had uterus in situ, their postmenopausal hormone therapy had been exclusively EPT, the use of which was assessed from the National Drug Reimbursement Register for the cases and the controls. In Finland, various EPT regimens, available only with a physician's prescription, contain only estradiol as an estrogenic component. A 3-month use of EPT can be purchased at each pharmacy visit, and therefore, repeat visits and EPT purchases entered into the drug reimbursement register confirmed continuous use of the regimen. The register was opened in 1994, and thus a woman buying EPT regimens at the opening year 1994 (N = 5,826, 43.8%) was considered to have started the use of EPT at the age of 52 years, the mean starting EPT age in this population,17 and the duration of preregister EPT use was approximated accordingly. From 1995 onward to the end of 2011, the initiation date of EPT use was accurately known (N = 7,490, 56.2%).
Among the whole study population 13,316 (29.1%) women had used EPT. Of various progestogens, sole NETA-EPT was used by 2,923 women (21.9%) and sole MPA-EPT by 2,299 women (17.3%). The rest of the EPT users who had switched from one regimen to another and were exposed to NETA, MPA, dydrogesterone, megestrol acetate, lynestrenol, drospirenone, or trimegestone were categorized into mixed-EPT (N = 7,622, 57.2%). The EPT regimen could have been sequential or continuous and given orally or transdermally, but all EPT users were considered as one group. Altogether 317 women (2.4%) had used a Levo-IUD (Mirena) in addition to systemic use of estradiol, and 155 women (1.2%) had used tibolone. The use of local vaginal estrogen, alone or in addition to EPT was not recorded.
The various background factors between the cases and controls were tested with test of Equal or Given Proportions. The odds ratio of UP (OR; 95% CI) was assessed with a conditional logistic regression analysis. Subgroup analyses were carried out in women with NETA, MPA, Levo-IUD, mixed progestogen, and tibolone. Separate subanalyses were done for women who had started the use of EPT at and under or over 55 years of age. Moreover, to analyze the possible impact of EPT exposure duration subgroup analyses were done for exposures lasting under 3 years, 3 to 5 years, and over 5 years.
The appropriate permissions for the study were obtained from the National Institute of Health and Welfare (THL/1370/5.05.00/2010), the Finnish National Population Register (901/410/14), and Social Insurance Institution of Finland (KELA 40/522/2010).
At the time of UP operation 36.2% of the cases had been under 60 years, 31.5% between 60 and 69 years, and 32.3% were 70 years or older. The matching succeeded in age, parity and hospital district. The groups were also comparable with regard to age at the initiation of EPT regimens (Table 1).
The cases had used EPT more often than the controls (34.2% vs 27.3%; P < 0.005). The use of MPA-EPT was also more common among cases, whereas the distribution of NETA-EPT and estradiol with Levo-IUD showed no difference between the groups (Table 2). Instead, tibolone had been used less frequently (P < 0.005) by the cases than the controls. The EPT exposure durations and follow-up times did not differ between the cases and controls. (Table 2)
The use of any EPT was associated with a 41% elevation in the risk for UP (Table 3). The risk increase did not differ between the sole NETA-EPT and the sole MPA-EPT users (Table 3). Furthermore, the use of estradiol with the Levo-IUD was associated with a risk increase (52%), but in contrast, the use of tibolone was accompanied with a risk decrease of 70% (Table 3).
The elevations in UP risk were independent of the EPT initiation age (Table 4). The rise in the UP risk was seen already with an EPT use shorter than 3 years, but it increased further with longer exposures (Table 4).
The present data indicate that estradiol, if given in combination with systemic or intrauterine progestogen, is accompanied with an elevated risk of UP in postmenopausal women. Because EPT users became exposed to both estradiol and various progestogens, it is not possible to distinguish whether estradiol or progestogen or their combination accounted for risk increase in the development of UP. The Cochrane meta-analysis of four studies could not draw any definite conclusions on the role of estrogen (mostly CEE) on pelvic organ prolapse.13 This could be in part due to diagnostic difficulties because the diagnosis of pelvic organ prolapse has been based on subjective questionnaires in many previous studies.13,18 Questionnaires have often been used for recording the use of hormone therapy, although such a method is prone to recall biases. We focused solely on UP because the uterus is the leading prolapsing pelvic organ. Moreover, UP should have been severe enough to require hysterectomy, and thus, the UP diagnosis was accurate. We also benefited from using hormone register data which allowed us to accurately trace the use of EPT in our study women. Our data imply that the use of EPT may decrease the pelvic floor support to the extent that UP ensues. On the whole, our finding is in line with data from the WHI trial that hysterectomy with retained ovaries is accompanied with a higher prolapse risk than hysterectomy with bilateral oophorectomy.19 Thus, the common use of EPT, particularly some decades ago, may have contributed to the elevation in the incidence of pelvic floor failures during the last few decades, at least in our country.3
Although we conducted our large epidemiological study in a case-controlled manner, it can only show associations between preceding use of EPT and UP but no definite cause and effect relationship. In view of the presence of estrogen and progesterone receptors in the pelvic floor,4-8 it is, however, possible that EPT-derived estradiol and/or various progestogens could have had a role in the development of prolapse in our cases. Although the exact mechanisms are not known, they could resemble those in pregnancy, when high circulating levels of both estrogens and progesterone alone or together with relaxin, soften the pelvic floor. Pelvic organ prolapses are associated with degradative changes in collagen, metalloproteins, elastin, smooth muscle cells, and fibroblasts in the pelvic floor2,8,20,21 which, if developed for any reason, can perhaps be exaggerated by EPT. The fact that estradiol in combination with the Levo-IUD-EPT also carried UP risk increase suggests that estradiol rather than progestogens is the key factor for UP because the circulating levels of levonorgestrel are very low.22 This could be supported by the findings that estradiol affects extracellular matrix mediators, such as TGF-β and tropoelastin,9 and postmenopausal use of estradiol may also cause the structure of the extracellular matrix to be more vulnerable.8 Progestogens may, however, oppose or potentiate the effects of estradiol in many tissues,23 and such a phenomenon may also exist in the pelvic floor. Our data, however, imply that the two leading progestogens, MPA and NETA, in combination with estradiol, show no differences for pelvic floor support.
Tibolone is metabolized to estrogenic, progestogenic, and androgenic derivatives in a tissue-specific manner; for example, in breast, endometrium, bone, and blood vessel walls.24 This may also hold true for the pelvic floor because the androgenic metabolites of tibolone occur in higher levels, for instance, in the myometrium and vagina than in circulating blood.25 Androgenic metabolites may stimulate the growth of pelvic floor muscles,26 and thus, this could at least in part explain the decreased UP risk detected in our study, even if the number of women with tibolone was small. This finding, if confirmed in future larger studies, may favor the use of tibolone in women being at high risk of UP.
Our study has limitations. First, we did not have a placebo group, and therefore an EPT prescription bias should be considered. This is, however, unlikely because in Finland severe hot flushes and other subjective menopausal symptoms, but not pelvic organ prolapse, are primary indications for the initiation of hormone therapy, and no data link these symptoms with subsequent UP. Second, the indications for UP operations are to some extent relative, and one can argue that women seeking a better quality of life from the use of EPT were also more likely to have surgery due to milder forms of UP than similar women without EPT history. This argument is unlikely because UP surgery is done only with nationally approved uniform indications. Moreover, in Finland no socioeconomic differences are observed between EPT users and nonusers,27 and UP surgery is basically free of charge. Third, it is possible that women with UP may have started using EPT before the UP operation, wishing to relieve prolapse symptoms, but such a possible prolapse-induced bias in EPT prescription would have been against the uniform national guidelines for optimal use of EPT and most likely concerned local vaginal regimens that were excluded from our study. Moreover, the fully comparable EPT exposure times between cases and controls also speak against this bias. Fourth, we acknowledge as a weakness the small number of cases with estradiol + Levo-IUD and with tibolone, and thus, all conclusions for these therapies must be drawn with caution. Fifth, our patient series, although a large national one, was nevertheless too small to analyze the relationship between various modes of administration and doses of systemic EPT. Furthermore, we were unable to study the possible association between the degree of prolapse and doses of EPT. And finally, obesity is a risk factor for UP,28 but we had no data on body weight. However, it is unlikely that EPT users would have been more obese than the controls.
The strengths of our study include first a large number of cases and controls. Second, the UP diagnosis was accurate, because all UP patients had been operated on using nationally scrutinized indications. Third, the history of EPT use was also accurate because it is unlikely that a woman would have repeatedly bought, but not used EPT. And finally, we were able to compare MPA and NETA from a new point of view and to produce novel data for Levo-IUD and tibolone.
Women with UP had used systemic estradiol-EPT regimens and estradiol with Levo-IUD more often than women without UP. This may have weakened the pelvic floor, resulting in uterine prolapse. Therefore, our data must be incorporated into information for the present and future users of EPT.
1. Tinelli A, Malvasi A, Rahimi S, et al. Age-related pelvic floor modifications and prolapse risk factors in postmenopausal women. Menopause
2. Chen B, Yeh J. Alterations in connective tissue metabolism in stress incontinence and prolapse. J Urol
3. Kurkijarvi K, Aaltonen R, Gissler M, Makinen J. Pelvic organ prolapse
surgery in Finland from 1987 to 2009: a national register based study. Eur J Obstet Gynecol Reprod Biol
4. Skala CE, Petry IB, Albrich S, Puhl A, Naumann G, Koelbl H. The effect of genital and lower urinary tract symptoms on steroid receptor expression in women with genital prolapse. Int Urogynecol J
5. Skala CE, Petry IB, Albrich SB, Puhl A, Naumann G, Koelbl H. The effect of hormonal status on the expression of estrogen and progesterone receptor in vaginal wall and periurethral tissue in urogynecological patients. Eur J Obstet Gynecol Reprod Biol
6. Fuermetz A, Schoenfeld M, Ennemoser S, Muetzel E, Jeschke U, Jundt K. Change of steroid receptor expression in the posterior vaginal wall after local estrogen therapy. Eur J Obstet Gynecol Reprod Biol
7. Gebhart JB, Rickard DJ, Barrett TJ, et al. Expression of estrogen receptor isoforms alpha and beta messenger RNA in vaginal tissue of premenopausal and postmenopausal women. Am J Obstet Gynecol
8. Jackson S, James M, Abrams P. The effect of oestradiol on vaginal collagen metabolism in postmenopausal women with genuine stress incontinence. BJOG
9. Chakhtoura N, Zhang Y, Candiotti K, Medina CA, Takacs P. Estrogen inhibits vaginal tropoelastin and TGF-beta1 production. Int Urogynecol J
10. Rahn DD, Ruff MD, Brown SA, Tibbals HF, Word RA. Biomechanical properties of the vaginal wall: effect of pregnancy, elastic fiber deficiency, and pelvic organ prolapse
. Am J Obstet Gynecol
11. Goldstein SR, Nanavati N. Adverse events that are associated with the selective estrogen receptor modulator levormeloxifene in an aborted phase III osteoporosis treatment study. Am J Obstet Gynecol
12. Albertazzi P, Sharma S. Urogenital effects of selective estrogen receptor modulators: a systematic review. Climacteric
13. Ismail SI, Bain C, Hagen S. Oestrogens for treatment or prevention of pelvic organ prolapse
in postmenopausal women. Cochrane Database Syst Rev
14. Bhavnani BR, Stanczyk FZ. Pharmacology of conjugated equine estrogens: efficacy, safety and mechanism of action. J Steroid Biochem Mol Biol
15. Lyytinen H, Pukkala E, Ylikorkala O. Breast cancer risk in postmenopausal women using estradiol
-progestogen therapy. Obstet Gynecol
16. Binkowska M, Woron J. Progestogens in menopausal hormone therapy
. Prz Menopauzalny
17. Savolainen-Peltonen H, Tuomikoski P, Korhonen P, et al. Cardiac Death Risk in Relation to the Age at Initiation or the Progestin
Component of Hormone Therapies. J Clin Endocrinol Metab
18. Cardenas-Trowers OO, Borgstrom M, Addis I. Associations between type and route of hormone use on urinary incontinence and pelvic organ prolapse
in premenopausal and postmenopausal women. Female Pelvic Med Reconstr Surg
19. Shveiky D, Kudish BI, Iglesia CB, et al. Effects of bilateral salpingo-oophorectomy at the time of hysterectomy on pelvic organ prolapse
: results from the Women's Health Initiative trial. Menopause
20. Kerkhof MH, Ruiz-Zapata AM, Bril H, et al. Changes in tissue composition of the vaginal wall of premenopausal women with prolapse. Am J Obstet Gynecol
21. Ruiz-Zapata AM, Kerkhof MH, Ghazanfari S, et al. Vaginal fibroblastic cells from women with pelvic organ prolapse
produce matrices with increased stiffness and collagen content. Sci Rep
22. Nilsson CG, Lahteenmaki P, Luukkainen T. Levonorgestrel plasma concentrations and hormone profiles after insertion and after one year of treatment with a levonorgestrel-IUD. Contraception
23. Sitruk-Ware R. Pharmacology of different progestogens: the special case of drospirenone. Climacteric
2005; 8 (Suppl. 3):4–12.
24. Formoso G, Perrone E, Maltoni S, et al. Short-term and long-term effects of tibolone in postmenopausal women. Cochrane Database Syst Rev
25. Verheul HA, Blok LJ, Burger CW, Hanifi-Moghaddam P, Kloosterboer HJ. Levels of tibolone and estradiol
and their nonsulfated and sulfated metabolites in serum, myometrium, and vagina of postmenopausal women following treatment for 21 days with tibolone, estradiol
, or estradiol
plus medroxyprogestrone acetate. Reprod Sci
26. Ho MH, Bhatia NN, Bhasin S. Anabolic effects of androgens on muscles of female pelvic floor and lower urinary tract. Curr Opin Obstet Gynecol
27. Topo P, Luoto R, Hemminki E, Uutela A. Declining socioeconomic differences in the use of menopausal and postmenopausal hormone therapy
in Finland. Maturitas
28. Lee UJ, Kerkhof MH, van Leijsen SA, Heesakkers JP. Obesity and pelvic organ prolapse
. Curr Opin Urol