Figures
Abstract
Objective
Most women live to experience menopause and will spend 4–8 years transitioning from fertile age to full menstrual stop. Biologically, reproductive ageing is a continuous process, but by convention, it is defined categorically as pre-, peri- and postmenopause; categories that are sometimes supported by measurements of sex hormones in blood samples. We aimed to develop and validate a new tool, a reproductive ageing score (RAS), that could give a simple and yet precise description of the status of reproductive ageing, without hormone measurements, to be used by health professionals and researchers.
Methods
Questionnaire data on age, menstrual regularity and menstrual frequency was provided by the large multicentre population-based RHINE cohort. A continuous reproductive ageing score was developed from these variables, using techniques of fuzzy mathematics, to generate a decimal number ranging from 0.00 (nonmenopausal) to 1.00 (postmenopausal). The RAS was then validated with sex hormone measurements (follicle stimulating hormone and 17β-estradiol) and interview-data provided by the large population-based ECRHS cohort, using receiver-operating characteristics (ROC).
Results
The RAS, developed from questionnaire data of the RHINE cohort, defined with high precision and accuracy the menopausal status as confirmed by interview and hormone data in the ECRHS cohort. The area under the ROC curve was 0.91 (95% Confidence interval (CI): 0.90–0.93) to distinguish nonmenopausal women from peri- and postmenopausal women, and 0.85 (95% CI: 0.83–0.88) to distinguish postmenopausal women from nonmenopausal and perimenopausal women.
Conclusions
The RAS provides a useful and valid tool for describing the status of reproductive ageing accurately, on a continuous scale from 0.00 to 1.00, based on simple questions and without requiring blood sampling. The score allows for a more precise differentiation than the conventional categorisation in pre-, peri- and postmenopause. This is useful for epidemiological research and clinical trials.
Citation: Triebner K, Johannessen A, Svanes C, Leynaert B, Benediktsdóttir B, Demoly P, et al. (2020) Describing the status of reproductive ageing simply and precisely: A reproductive ageing score based on three questions and validated with hormone levels. PLoS ONE 15(6): e0235478. https://doi.org/10.1371/journal.pone.0235478
Editor: Krasimira Tsaneva-Atanasova, University of Exeter, UNITED KINGDOM
Received: January 9, 2020; Accepted: June 16, 2020; Published: June 30, 2020
Copyright: © 2020 Triebner et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: The data in the tables of the uploaded manuscript contain all information to completely replicate the reproductive ageing score. Data used for the validation of the score, such as hormone measurements and other covariates are sensitive patient information and due to Norwegian ethical and legal restrictions only available upon request to qualified researchers. The institutional body responsible for restricting the data is the European Community Respiratory Health Survey consortium. Requests for data access can be directed to Haukeland University Hospital, 5021 Bergen, Norway, Att. Med. Director Alf H. Andreassen; e-mail: postmottak@helse-bergen; phone: +47 55 97 50 00. Org. nr: 983 974 724.
Funding: Kai Triebner has received a postdoctoral fellowship from the University of Bergen. The present analyses are part of a project funded by the Norwegian Research Council (Project No. 228174) as well as part of the Ageing Lungs in European Cohorts (ALEC) Study (www.alecstudy.org), which has received funding from the European Union's Horizon 2020 research and innovation program (Grant No. 633212). The European Commission supported the European Community Respiratory Health Survey, as part of the “Quality of Life” program. Bodies funding the local studies are listed in the supporting information. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
1. Introduction
Menopause marks the cessation of menstruations and the end of the reproductive part of life [1]. This transition, which occurs in women around 50 years of age, takes on average five years and is a major part of reproductive ageing [2, 3]. It also implies profound hormonal changes: a woman’s estrogen levels start to decline, and her gonadotropin levels begin to rise. The result is that her body gradually changes into a non-reproductive state [4, 5]. Epidemiological research traditionally uses categories to describe the reproductive status of a woman, such as perimenopausal or postmenopausal. These categories are based on arbitrary thresholds; they may be rather heterogeneous and lead to an aggravated interpretation of research findings. As an example, the current consensus, the Stages of Reproductive Ageing Workshop (STRAW) defines late perimenopause as a stage lasting one to three years, during which women experience amenorrhea for 60 days or more [6]. This evidently conglomerates a range of women of different reproductive ages, who would then be jointly analysed in an epidemiological study. Categorization may facilitate communication between clinicians and the public, but it does not reflect the underlying biology and it introduces clear limitations, especially for epidemiological research on women undergoing the menopausal transition [7]. The frequent use of categories can however be understood, as there is to date no single biomarker sufficiently describing menopause [6, 8].
We aim to mathematically describe the status of reproductive ageing and create an easy to replicate reproductive ageing score based on the age of a woman and the number and regularity of menstruations she experiences.
2. Materials and methods
We used data from two population-based cohorts, the Respiratory Health in Northern Europe (RHINE) study and the European Community Respiratory Health Survey (ECRHS). Ethical approval was obtained for each centre of the RHINE study and the ECRHS from the appropriate institutional or regional ethics committee and each participant provided informed written consent prior to inclusion in the studies.
2.1 RHINE
Bergen: Regional Committee for Medical and Health Research Ethics in Western Norway (2010/759); Reykjavik: National Bioethics Committee of Iceland (VSN 11–121); Uppsala, Umeå, Göteborg: Ethics Committee of Uppsala University (2010/068); Aarhus: not required for questionnaire-only studies; Tartu: Research Ethics Committee of the University of Tartu (209/T-17);
2.2 ECRHS
Aarhus: Scientific ethical committee for Region Midtlylland; Albacete: Comité Ético de Investigación Clínica del Hospital Universitario de Albacete; Galdakao: Comité Ético de Investigación Clínica del Hospital de Galdakao-Usansolo; Huelva: Comisión de Ética de Investigación Sanitarias del Hospital Juan Ramón Jiménez de Huelva; Bergen: Regional Committee for Medical and Health Research Ethics in Western Norway (2010/759); Bordeaux, Grenoble, Montpellier, Paris: Comite De Protection Des Personnes (2011-A00013-38); Erfurt, Hamburg: Ethikkommission der Bayerischen Landesärztekammer (Reg Nr. 10015); Uppsala, Umeå, Göteborg: Ethics Committee of Uppsala University (2010/068); Reykjavik: National Bioethics Committee of Iceland (VSN 11–121); Tartu: Research Ethics Committee of the University of Tartu (209/T-17);
2.3 Development and validation population
The reproductive ageing score (RAS) was developed using data of subjects participating in the questionnaire-based RHINE study and it was validated in the population of the ECRHS where objective sex hormone measurements were available.
The RHINE study is a longitudinal, international, multi-centre study (www.rhine.nu), which includes seven Northern European centres (Bergen in Norway; Reykjavik in Iceland; Umeå, Uppsala and Göteborg in Sweden; Aarhus in Denmark and Tartu in Estonia). For the current paper, we used data from the most recent wave, carried out between 2010 and 2012 with a response rate of 63% [9]. We used data from 3107 women with a mean age of 52 years (range: 38–66 years) to develop the RAS.
The ECRHS also is a longitudinal, international, multicentre study (www.ecrhs.org) [10, 11]. The validation population includes women from 16 centres in nine countries who participated during 2010 to 2012 (Aarhus in Denmark; Albacete, Galdakao and Huelva in Spain; Bergen in Norway; Bordeaux, Grenoble, Montpellier and Paris in France; Erfurt and Hamburg in Germany; Göteborg, Umeå and Uppsala in Sweden; Reykjavik in Iceland and Tartu in Estonia). The available serum samples were analysed for concentrations of follicle stimulating hormone (FSH) and 17β-estradiol. The database contained 1056 women with a mean age of 55 years (range: 40–67 years). For the seven Northern European centres in ECRHS, participants were also a subsample of the RHINE study. Women who participated in both surveys were included into the validation population but not the development population. We further excluded women from both populations who currently used exogenous sex hormones like contraceptives or hormone replacement therapy (including intermittent progestin therapy), women being pregnant or breastfeeding at the time of the surveys and women reporting irregular menstruation unrelated to menopause (Fig 1) [6, 12, 13].
RHINE: Respiratory Health in Northern Europe study, ECRHS: European Community Respiratory Health Survey.
2.4 Score development
To develop the score we used techniques of fuzzy set theory, a mathematical concept to depict the biology of physiological processes [14, 15]. It was created in 1964 and successfully implemented in biology, artificial intelligence, and linguistics [14, 16, 17]. Unlike conventional mathematics, which does not allow vague expressions and demands that an object either is a member of a set or not, fuzzy sets are defined by a function (μ) assigning a value between 0.00 and 1.00 to an observation, representing the degree of belonging to a fuzzy set. The value 1.00 means that an object completely belongs to the set and the value 0.00 means the object does not at all belong to the set.
We first defined a function μA based on the number of periods per year and the menstrual regularity. The function requires the proportion P(period) of regularly menstruating women for every single number of reported periods as answers to the question: “How may periods did you have in the last twelve months?”. A cross-tabulation (Table 1) of the menstrual regularity (answers to the question: “Do you have regular periods?”, possible answers: “Yes”, “No, they have been irregular for a few months”, “No, my periods have stopped”) and the number of periods, illustrates the entirety of all values for P(period) calculated with Eq 1.
Equation 1. Proportion of women who have regular menstruation for each number of reported menstruations in the last year (with period = number of periods per year, x = number of women answering “Yes” to the question: “Do you have regular periods?”, y = number of women answering “No, they have been irregular for a few months” and z = number of women answering “No, my periods have stopped”, e.g. x(11) = number of women reporting regular menstruation among those who report 11 menstruations in the last 12 months).
For women reporting zero menstruations per year this proportion is expected to be zero, as those women are most likely menopausal, while for women reporting twelve menstruations per year it is expected to approach one, as those women are most likely nonmenopausal. For women who report more than twelve menstruations per year it is expected to decline again, as shortening as well as lengthening cycles are an indicator of the beginning menopausal transition [18]. Plotting the complementary proportion 1-P(period) versus the corresponding number of periods per year, depicts a discrete function, from which the continuous function μA can be approximated using the least squares function approximation.
A second function μB was defined based on age. The construction of μB requires the proportion P(age) of women whose menstruations have already stopped for every single reported age. A cross-tabulation of the menstrual status with the reported ages illustrates the entirety of all values for P(age) calculated with Eq 2 (Table 2).
Equation 2. Proportion of women whose menstruations have already stopped, for each reported year of age (with age = age in years, x = number of women answering “Yes” to the question: “Do you have regular periods?”, y = number of women answering “No, they have been irregular for a few months”, z = number of women answering “No, my periods have stopped”, e.g. x(40) = number of women reporting regular menstruations among those who are 40 years old).
This proportion increases with age and for younger women this proportion is expected to be low. Plotting the proportion P(age) versus the corresponding age depicts a discrete function, from which the continuous function μB can be approximated using the least squares function approximation.
Additionally, two optional modifiers were introduced to μB: smoking and unilateral oophorectomy. According to recent literature, current smoking is associated with two years [19] and unilateral oophorectomy with one year [20] younger age at menopause. Therefore, two years were added to the age of current smokers and one year was added to the age of women reporting unilateral oophorectomy, defining the new variable mage (modified age) (Eq 3).
Equation 3. Age modification by smoking and oophorectomy.
Subsequently, the value for the function μB is defined from P(mage).
The calculated probability of being not regularly menstruating or without menses, according to the number of periods within the last twelve months (μA), does not depend on the woman’s age, thus allowing to calculate the probability for each year of being either amenorrheic or not regularly menstruating. Thus, the RAS can be calculated as the union of μA and μB by adding the two sets (μA + μB) and subtracting the overlap, i.e. the repeated elements of the intersection (μAμB):
Equation 4. The reproductive ageing score as an aggregation function of μA and μB.
The union of μA and μB can be imagined within a three-dimensional Cartesian coordinate system, where μA is projected on the planes parallel to the one spanned by the x- and z-axis for each increment of μB and is projected on the planes parallel to the one spanned by the y- and z-axis for each increment of μB. Subsequently the function, representing the union of μA and μB is formed by the maximum value of either function (μA or μB) for any given coordinates of the plane spanned by the x- and y-axis. Thus, Eq 4 can be applied for any woman, knowing the following variables: age, number of periods in the last year, age, oophorectomy (“Never”, “One ovary”, “Both ovaries”) and the smoking status (“Yes” / “No”).
2.5 Validation of the developed score
We used receiver operating characteristics (ROC) to validate the calculated score against established hormone cut-offs in our validation population (ECRHS) [21] (nonmenopausal: FSH ≤20IU/L and 17β-estradiol ≥147pmol/L, postmenopausal: FSH ≥80IU/L and 17β-estradiol ≤73pmol/L [22]. ROC curves show the true positive rate (sensitivity) against the false positive rate (specificity) and compare to random guessing. An area under the curve (AUC) of one would indicate perfect performance. We validated two concepts. First, we tested how well the score separated nonmenopausal women from the remaining (perimenopausal and postmenopausal) women and second, how well the score separated postmenopausal women from the remaining (nonmenopausal and perimenopausal) women. As perimenopausal women score by definition intermediate values, they were evaluated by comparing intermediate cut-offs of the score to hormone levels and menstrual pattern. For the sake of completeness, this was also done for nonmenopausal and postmenopausal women. Additionally, in order to visualize the performance of the RAS, we include a boxplot of the RAS against three commonly used categories defined by hormonal measurements in the supporting information (S1 Fig).
The least square approximations to develop the RAS were performed using the Maxima CAS (Computer Algebra System) software [23]. All other calculations, including the validation of the RAS, were performed using the R statistical package [24].
3. Results
The cross-tabulated data to derive the first function μA based on P(period) is presented in Table 1. This function can be approximated by a biquadratic exponential function with a mean-squared error (MSE) of 0.011 (Eq 5). It shows, consistently with existing literature, that the transition to menopause is characterized by an increased frequency of both very long and very short cycles (Fig 2) [18].
Data points: Inverse proportion of women with regular menstruation for every response to the number of periods during the last year, observed in the RHINE dataset 1-P(period); Line: biquadratic exponential function μA with best fit to observed values 1-P(period).
Equation 5. Biquadratic exponential function μA depending of the number of periods.
The tabulated data to derive the second function μB (Eq 6) based on P(age) is presented in Table 2.
The calculated proportion of women who indicate being menopausal by answering “No” to the question "Do you have regular periods?" of all women of the same age (μB) can be approximated by a quadratic logistic function (Fig 3) with a MSE for this approximation of 0.002.
Equation 6. Quadratic logistic function approximating the function μB (with age in years).
Data points: Proportion of women without menstruations according to age observed in the RHINE dataset P(age); Line: quadratic logistic function μB with best fit to observed values P(age).
Finally, the function of the RAS (Eq 7) according to the aggregation expressed in Eq 4:
Equation 7. Final formula to calculate the reproductive ageing score (RAS) (with period being the number of periods per year and age as the age in years, modified according to smoking status and oophorectomy).
Graphically the RAS can be represented as a three-dimensional figure (Fig 4) using the variables period along the x-axis and age along the y-axis.
Both, older age and fewer menstruations contribute to a higher RAS and indicate progression into a postmenopausal state. Smoking and oophorectomy act as modifiers of the function μB and thus the RAS. We calculated the RAS for women within the validation population (ECRHS) with Eq 7, which was derived from the development population (RHINE).
The area under the ROC curve was 0.91 (95% CI: 0.90–0.93) to distinguish nonmenopausal women from the remaining perimenopausal and postmenopausal women (Fig 5, black curve) and 0.85 (95% CI: 0.83–0.88) to distinguish postmenopausal women from the remaining nonmenopausal and perimenopausal women (Fig 5, grey curve), as defined by concentrations of FSH and 17β-estradiol.
Black curve: Nonmenopausal women versus perimenopausal and postmenopausal women; Grey curve: Postmenopausal women versus perimenopausal and nonmenopausal women.
Table 3 shows reproductive characteristics for quartiles of the RAS, to illustrate the reproductive characteristics of women with intermediate reproductive ageing scores (0.26–0.75), who are presumed to be in the menopausal transition.
4. Discussion
We developed a continuous reproductive ageing score based on age and number of menstruations. Calculating this score for each woman in the validation population (ECRHS) showed that women with the lowest scores featured nonmenopausal characteristics [21, 25]. Women who scored intermediate values showed characteristics of advancing degrees of the menopausal transition and women who scored highest strongly resembled typical postmenopausal women [21, 25]. The RAS can be interpreted as an indicator of the progress of reproductive ageing. Women with a score of 0.00 can be considered premenopausal and women with a score of 1.00 can be considered postmenopausal, while the intermediate values can be considered advancing degrees of reproductive ageing in terms of decreasing fertility, respectively depletion of the ovarian reserve.
The validation with ROC curves and the practical example (supporting information) presented the RAS as a useful tool for epidemiologists. The performance of the RAS for women who were either nonmenopausal or postmenopausal was very good, with AUC values of 0.91 and 0.85, respectively. The ROC curve validation shows that the RAS discriminates nonmenopausal women slightly better than postmenopausal women. The score was able to quantify degrees of the menopausal transition, which has so far not been possible, resulting in women at different stages of reproductive ageing being defined as one heterogeneous group. This tool has great potential to offer new insights into health and disease, e.g. whether women are vulnerable to a certain health condition during a narrow time window within the perimenopausal phase.
In addition, the mean square deviation of the calculated functions reaches values very close to zero, implying a strong nonlinear correlation between the independent variables (number of periods in the last year and age) and the proportions, calculated within the development population. Concerning μA, possibly less precise data in the upper range of menstruations per year (>12), due to the lower number of data points, is largely being adjusted for by the relation with age (μB) after forming the final function (RAS). The RAS is based on the strong association between age and the changing number of menstrual periods and its major strength is to quantify reproductive ageing continuously and that it is based on answers to a few simple questions. Important factors influencing reproductive ageing are unilateral oophorectomy and current smoking behaviour, both related to a younger age at menopause, which we accounted for as modifiers. In the case that it is desirable to evaluate smoking behaviour and/or unilateral oophorectomy separately or in a different manner, these modifiers may be removed from the calculation of the RAS.
Another strength is that the RAS can be easily used with all common statistical software and spread sheets by applying the final formula (Eq 7) to a dataset. To replicate the development of the RAS (least square approximations) in other settings we recommend using the open-source software Maxima CAS (https://sourceforge.net/projects/maxima/files/), a specialized computer algebra system that yields high precision numerical results and is capable of including and evaluating complex operations, such as Taylor series and Laplace transforms as well as linear algebra tools like matrix operations, which have been used for the current calculations.
The proportions of women with regular menstruation for the various reported number of periods as well as ages correspond well to the general consensus and what has been described in other studies [6, 8, 22]. Both the development population and the validation population are representative for the women in the relevant age groups in Europe [9], thus the external validity is high for Caucasian populations. For other ethnicities, the functions might have to be slightly modified as age at menopause might differ.
A limitation of the RAS is that potential factors affecting reproductive ageing such as chronic disease, gynaecological disorders and the use of exogenous hormones, are not considered. These limitations are however also acknowledged in the STRAW +10 model [26], which, today, is considered to be the gold standard for assessing reproductive ageing.
It must also be noted that a continuous RAS, indicating how far along a woman is on her way from fertile age to menopause (0.00–1.00) should not be confused with menopause scores assessing women’s health after menopause or climacteric symptoms [27, 28].
5. Conclusion
The RAS provides a new, innovative and useful tool to describe the current status of reproductive ageing accurately, on a continuous scale from 0.00 to 1.00, based on simple questions and without a need for blood measurements. The score allows for a more precise differentiation between women during this period than the current, conventional categorisation into pre-, peri- and postmenopause. It thus is useful for epidemiological research and in the design of clinical trials, e.g. studies on hormone replacement therapy.
Acknowledgments
We thank all participants, field workers and coordinators of the RHINE study and the ECRHS for their efforts as well as Ersilia Bifulco and Sandra Suske from the Core Facility for Metabolomics at the University of Bergen where the hormone measurements were performed. Further we are very grateful for the support of the Department of Applied Mathematics of the University of Málaga (Spain) and we want to thank Elinor Bartle for revising grammar and language.
References
- 1. Lumsden MA, Sassarini J. The evolution of the human menopause. Climacteric. 2019;22(2):111–6. Epub 2019/02/05. pmid:30712396.
- 2. Allshouse A, Pavlovic J, Santoro N. Menstrual Cycle Hormone Changes Associated with Reproductive Aging and How They May Relate to Symptoms. Obstet Gynecol Clin North Am. 2018;45(4):613–28. Epub 2018/11/08. pmid:30401546; PubMed Central PMCID: PMC6226272.
- 3. Gracia CR, Freeman EW. Onset of the Menopause Transition: The Earliest Signs and Symptoms. Obstet Gynecol Clin North Am. 2018;45(4):585–97. Epub 2018/11/08. pmid:30401544.
- 4. Randolph JF Jr., Zheng H, Sowers MR, Crandall C, Crawford S, Gold EB, et al. Change in follicle-stimulating hormone and estradiol across the menopausal transition: effect of age at the final menstrual period. The Journal of clinical endocrinology and metabolism. 2011;96(3):746–54. Epub 2010/12/17. pmid:21159842; PubMed Central PMCID: PMC3047231.
- 5. Sowers MR, Zheng H, McConnell D, Nan B, Harlow S, Randolph JF Jr. Follicle stimulating hormone and its rate of change in defining menopause transition stages. The Journal of clinical endocrinology and metabolism. 2008;93(10):3958–64. Epub 2008/07/24. pmid:18647816; PubMed Central PMCID: PMC2579655.
- 6. Harlow SD, Gass M, Hall JE, Lobo R, Maki P, Rebar RW, et al. Executive summary of the Stages of Reproductive Aging Workshop + 10: addressing the unfinished agenda of staging reproductive aging. Menopause. 2012;19(4):387–95. pmid:22343510; PubMed Central PMCID: PMC3340903.
- 7. Guthrie J, Dennerstein L, Burger H. How reliably does 12-month amenorrhea define final menstrual period? Data from a longitudinal study. Climacteric. 2002;5(1):92. pmid:11974565.
- 8. Burger HG, Robertson DM, Baksheev L, Collins A, Csemiczky G, Landgren BM. The relationship between the endocrine characteristics and the regularity of menstrual cycles in the approach to menopause. Menopause. 2005;12(3):267–74. pmid:15879915.
- 9. Johannessen A, Verlato G, Benediktsdottir B, Forsberg B, Franklin K, Gislason T, et al. Longterm follow-up in European respiratory health studies—patterns and implications. BMC pulmonary medicine. 2014;14:63. pmid:24739530; PubMed Central PMCID: PMC4021078.
- 10. Burney PG, Luczynska C, Chinn S, Jarvis D. The European Community Respiratory Health Survey. Eur Respir J. 1994;7(5):954–60. pmid:8050554.
- 11. Jarvis D KJ, Burney P, Chinn S, Luczynska C, Anto JM, Cerveri I, et al. The European Community Respiratory Health Survey II. The European respiratory journal. 2002;20(5):1071–9. pmid:12449157.
- 12. Taponen S, Martikainen H, Jarvelin MR, Laitinen J, Pouta A, Hartikainen AL, et al. Hormonal profile of women with self-reported symptoms of oligomenorrhea and/or hirsutism: Northern Finland birth cohort 1966 study. The Journal of clinical endocrinology and metabolism. 2003;88(1):141–7. pmid:12519843.
- 13. West S, Lashen H, Bloigu A, Franks S, Puukka K, Ruokonen A, et al. Irregular menstruation and hyperandrogenaemia in adolescence are associated with polycystic ovary syndrome and infertility in later life: Northern Finland Birth Cohort 1986 study. Hum Reprod. 2014;29(10):2339–51. pmid:25085801; PubMed Central PMCID: PMC4164150.
- 14. Zadeh LA. Fuzzy sets. Information and Control. 1965;8(3):338–53. https://doi.org/10.1016/S0019-9958(65)90241-X.
- 15.
Zadeh LA. Biological application of the theory of fuzzy sets and systems. Proc of the Symp on the Biocybernetics of the Higher Nervous System; Boston: Little, Brown and Co.; 1969. p. 199–206.
- 16. Belohlavek R, Cordero P, Enciso M, Mora Á, Vychodil V. Automated prover for attribute dependencies in data with grades. International Journal of Approximate Reasoning. 2016;70:51–67. https://doi.org/10.1016/j.ijar.2015.12.007.
- 17. Rodríguez-Jiménez JM, Cordero P, Enciso M, Mora A. Data mining algorithms to compute mixed concepts with negative attributes: an application to breast cancer data analysis. Mathematical Methods in the Applied Sciences. 2016:n/a-n/a. pmid:28133404
- 18. Harlow SD, Paramsothy P. Menstruation and the menopausal transition. Obstetrics and gynecology clinics of North America. 2011;38(3):595–607. pmid:21961722.
- 19. Dratva J, Gomez Real F, Schindler C, Ackermann-Liebrich U, Gerbase MW, Probst-Hensch NM, et al. Is age at menopause increasing across Europe? Results on age at menopause and determinants from two population-based studies. Menopause. 2009;16(2):385–94. pmid:19034049.
- 20. Bjelland EK, Wilkosz P, Tanbo TG, Eskild A. Is unilateral oophorectomy associated with age at menopause? A population study (the HUNT2 Survey). Hum Reprod. 2014;29(4):835–41. pmid:24549218.
- 21. Speroff L. The perimenopausal transition. Annals of the New York Academy of Sciences. 1999;900:375–92. pmid:10818426.
- 22.
Speroff L, Glass RH, Kase NG. Clinical gynecologic endocrinology and infertility. In: Mitchell C, editor. Clinical gynecologic endocrinology and infertility. 6th ed. Baltimore: Lippincott Williams & Wilkins; 2011.
- 23. Ney de Souza P, Fateman RJ, Moses J, Yapp C. The Maxima Book. 2004.
- 24.
R Development Core Team. R: A language and environment for statistical computing.: R Foundation for Statistical Computing, Vienna, Austria; 2018.
- 25. Hale GE, Zhao X, Hughes CL, Burger HG, Robertson DM, Fraser IS. Endocrine features of menstrual cycles in middle and late reproductive age and the menopausal transition classified according to the Staging of Reproductive Aging Workshop (STRAW) staging system. The Journal of clinical endocrinology and metabolism. 2007;92(8):3060–7. pmid:17550960.
- 26. Harlow SD, Gass M, Hall JE, Lobo R, Maki P, Rebar RW, et al. Executive summary of the Stages of Reproductive Aging Workshop +10: addressing the unfinished agenda of staging reproductive aging. Climacteric. 2012;15(2):105–14. pmid:22338612; PubMed Central PMCID: PMC3580996.
- 27. Sydora BC, Fast H, Campbell S, Yuksel N, Lewis JE, Ross S. Use of the Menopause-Specific Quality of Life (MENQOL) questionnaire in research and clinical practice: a comprehensive scoping review. Menopause (New York, NY). 2016. Epub 2016/06/15. pmid:27300115.
- 28. Zollner YF, Acquadro C, Schaefer M. Literature review of instruments to assess health-related quality of life during and after menopause. Qual Life Res. 2005;14(2):309–27. pmid:15892422.