PREDICTOR FACTORS OF THE MENOPAUSE INSTALLATION BEFORE 45 YEARS AT PATIENTS WITH AUTOIMMUNE
, I. GH. TOTOIANU2
, MIHAELA STANCIU3
1PhD candidate University „Lucian Blaga“ of Sibiu, 2,3University „Lucian Blaga“of Sibiu
menopauza timpurie, insuficienţă ovariană prematură, boală tiroidiană autoimună
Rezumat: Obiective: Scopul studiului a fost identificarea unor factori predictivi pentru instalarea menopauzei timpurii (early menopause, EM), la pacientele cu boală tiroidiană autoimună (autoimmune thyroid disease, AITD). Material şi metodă: Studiul a inclus 162 paciente cu AITD diagnosticată înainte de 45 de ani. Am efectuat analiza univariată a mai multor factori independenţi şi analiza multivariată prin regresie logistică a factorilor cu semnificaţie statistică (p<0,01), cu modelarea unui scor de predicţie. Rezultate: Scorul de predicţie a EM = –2.2022 x Vârsta <35 ani la diagnosticul AITD (1/0) + 1.9328 x Infertilitatea (1/0) – 0.3792 x Anticorpii antiovarieni (U I /ml), cu criteriul ≤ –1.8359, sensibilitatea 90%, specificitatea 88%, AUC – 0.91. Concluzii: Doi factori predictivi sunt legaţi de momentul diagnosticării AITD: vârsta < 35 ani care scade riscul şi infertilitatea care creşte riscul şi ar putea avea o corelaţie cu îmbătrânirea ovariană şi respectiv insuficienţă ovariană ocultă.
Keywords: early menopause, premature ovarian failure, autoimmune thyroid disease
Abstract: Objectives: The aim of this study was to identify predictive factors for early menopause (EM) to autoimmune thyroid disease patients (AITD). Material and methods: The study included 162 patients with AITD diagnosed before 45 years. We performed univariate analysis of several independent factors by multivariate logistic regression analysis of the factors with statistical significance (p <0.01), with a score of predictive modeling. Results: Score EM prediction = -2.2022 x Age <35 years at the diagnosis of AITD (1 / 0) + 1.9328 x Infertility (1 / 0) -0.3792 x Antiovarian antibodies (IU / ml), the criterion ≤ – 1.8359, sensitivity 90%, specificity 88 %, AUC - 0.91. Conclusions: Two predictors are linked AITD for diagnosis: age <35 years which decreases the risk and infertility which increase the risk, and may have a correlation with ovarian aging and occult ovarian failure.
1Corresponding Author: Mihaela Ciocan, Clinical Emergency County Hospital Mureş, 1,Sportivilor street, Tîrgu Mureş, Romania; e-mail:
[email protected]; tel +40-0265 262288
Article received on 18.04.2011 and accepted for publication on 08.08.2011 ACTA MEDICA TRANSILVANICA September 2011; 2(3)418-422
Menopause is the moment of the definitive cessation of the periods, due to the cessation and of the ovarian activity. It is declared, conventionally, at one year of amenorrhea, in the absence of a known cause , the hormonal marker being the growth of the foliculinostimulating hormon (FSH) over 40UI/l, the expression of the primary ovarian insufficiency . The age of the natural menopause communicated by the cohort studies is uniformly distributed between 40 and 60 years, with median or average in the interval 49-52 years [1, 3, 4] and is the result of a combination of factors: genetics, reproductive and a way of life [4-7].
The term of precocious menopause is used for the instalation of the menopause before 40 years. Premature ovarian insuficiency (POF) has a wide meaning, defining the precocious depletion of the ovarian folicules, including also the women with primary amenorrhea and is the term agreed in literature [3, 8]. Epidemiological studies showed that ~ 1% of the females in the general population experience the menopause before 40 years, in exchange in many females , ~5% before 45 years .
As the age limit of 40 years is fixed conventionally, more rationally for the description of the phenomenon seemed the age of 45 years, at 2 standard deviation toward the expected age of the natural menopause, named early menopause (EM) [9, 10].
The physiological ovarian ageing, defines the decline of the ovarian reserve predictibly, in certain limits, with debut at
the age of 35 , and a critical level around 37.5 ani and its exhaustion around 51 years, in parallel with the decline of the fecundability and natural fertility. Aproximativelly 10% of the females present a deviation from the standard, the so called premature ovarian ageing or ocult ovarian insuffiency.
Excepting the recognized causes of POF,among the risk factors is well-known the personnal hystorical of autoimmune diseases [10, 11]. The frequent association is rapported with the autoimmun thyroidian disease (AITD), especially chronic limphocitar thyroiditis Hashimoto, between 12-33% ,.
The thyroidian pathologyis a frequent pathology, the NHANES III study rapporting the prevalence of the hypothyroidism in the SUA of 4.6%, of the hyperthyroidism of 1.3% and of the antithyroperoxidasis antibodies(ATPO) of 11.3±0.4%, with a predominence in women. It is recognized the influence of the abnormal thyroidian funtion on the hypothalamo-pituitary-gonadal ax and implicitely on the menstruation function and fertility, but its problematic becomes actual due to the tendency of postponement of the first birth after years , the growing incidence of the infertile couples and acceding the proceeds of assisted reproduction .
THE AIM OF THE PAPER
The aim of the paper was to identify the predictive factors for the installation of the early menopause (EM), in patients with autoimmune thyroidian disease (AITD).
MATERIAL AND METHOD
The study was developed during 01.08.2008- 31.03.2011 and included a number of 162 patients, that addressed to the ambulatory of the Gynecology Clinic II in Tîrgu Mureş.
The criterion of inclusion in the study was the presence of the AITD with debute before 45 years: Hashimoto thyroiditis, ATPO >50 UI/ml or Graves-Basedow disease, antibodies antireceptor for the hormone of thyroidian stimulation (TRab)>1,5 UI/l). The criterion of exclusion were : AITD in the context of the autoimmune polyglandular syndroms and secundary menopause due to a genetic or iatrogenic cause The menopause was defined as amenorrhea of 1 year or as in the women with the age under 45 years as amenorrhea of at least 4 months, with FSH levels of >40UI/ml, at least two determinations in a one month interval .
In all cases the anamnesis has been performed, the clinical general exam, the determination of the IMC, the gynecological exam, transvaginal ultrasound, the determination of the anti-ovarian antibodies (AOA). In the case of the AITD or of the installated menopause former to the enlist of the information based on the medical documents and on the patients declaration, in case of diagnosing at the enlist there have been effectuated the serial determination of the FSH, of the estradiol, prolactin, thyrostimulant hormone (TSH), free thyroxine , ATPO or Trab, thyroidian ultrasound, endocrinological exam. In all cases the informed consent of the patients was obtained .
In every case there have been pursued several factors that could be correlated with the age of the menopause installation in this group of patients and have been analysed as independent variables: general (smoking, the level of instruction, IMC), linked to the debut of AITD (the diagnosis age, the thyroidian funtion, the type of AITD, the level of the TSH, the level of ATPO), reproductives (menarrh, the characteres of the period , parity, infertility), the level of AOA at the moment of the enroll in the study, the associated pathology (autoimmune, hyperprolactinemy, hyperandrogenism, uterine fibroma, endometrum hyperplasia, the syndrome of the polycystic ovary, benign ovary tumors, fibrocystic mastopaty, depressive syndrome ).
For the statistical data processing were used the estimation programes SPSS for the t-Student test, the programe
of estimation Microsoft Excel for the arrithmetic media, the standard deviation (DS), the minimal and maximal values, the median, the programe MedCalc for the ROC curves (Receiver Operating Characteristic) and multivariate analysis,through logistic regression. The statistical interpretatiopn of the data considered p<0,01 potent as a statistical significance. The variables for which the p was significant statistically were used at the logistic remodeling and a prediction score was obtained for the installation of the EM.
The study included 162/217 (75%) patients with AITD diagnosed before 45 years, selected from 217 paciente with AITD, that addressed during the study period in an ambulatory service of gynecology. Among those 24/162 (15.0%) were at menopause, 19/162 (12%) with EM and 11/162 (7.0%) with POF. In the two subgroups in the case of 10/19 (10/162, 6.2%) and respectivelly 7/11 cases (7/162, 4.3%) the diagnosis of AITD preceded or was concomitently with the menopause pronouncement. The first subgroup meets the minimal number of cases that permit the logistic remodeling.
The threshold of statistical signifiance for a number of 26 independent variables , among which only 5 had p<0,01 (the age of diagnosis AITD <35 years, infertility, the level of AOA UI/ml, menopause buffs and perspiration).
Those were used for the multivariate analysis through logistic regression through anterograde method step by step.
After the modelation a configuration was obtained with 3 predictive factors and a prediction score of the installation of premature menopause:
SCORE = -2.2022 x diagnosis age AITD < 35 years (1/0 ) +1.9328 x Infertility (1/0) - 0.3792 x anti-ovarian antibodies (UI/ml).
The analitical performance against the configuration with the age the AITD numeric of the score is confirmed by the ROC.curve The optimal operating point showed the score criterion ≤ -1.8359, with specificity of 88% and sensibility of 90%, the area under the curve AUC1=0.91 > AUC2=0.82.
We have calculated the risks and we have represented them graphically, rezulting 6 curves that evidentiate very clearly the risk differences.
Table no. 1. The descriptive characteristics of the studied lote
Table no. 2. The univariate analysis of the independent variables
Table no. 3. The logistic model for the estimation of the prediction score of the menopause <45 ani
Figure no. 1. The ROC curve of the logistic score
Figure no. 2. The curves of the risks after the logistic score
To the eredity a contribution of at least 50% is atributted at the determination of the menopause age, considered a quantitative characteristic, influenced by several genes [5, 15]. The genomic studies have found 4 loci associated with the determination of the age of the natural menopause (on the chromosoms 19, 20, 6 and 5) the first two of them having an substantial effect, of reduction with 3 months respectivelly of growing with11 month per allele. The two variants mentioned are also risk factors sigficantly for EM, but not for POF. .
Cohort studies have proved that the că race/ethnicity is significanlty linked to the age of menopause [1, 16].
The influence of the factors of environment on the menopause was the object of numerous studies.The most significant results are linked with the smoking , that determines the installation of the menopause with 1-2 years earlier [1, 7].
The results regarding the action of other factors are less concordant: the low level of education, nuliparity and the small number of children, often mentruation cycle, precocious menarche, not using the oral contraceptives, low weight , physical activity depressive syndrome [1, 16].
determination of the AOA being a specificity and sensibility test when the are low [2, 3, 12, 17].
Having as a point of departure the researches previously mentioned we have been searching the influence of independent factors on the menopause age. At the univariate analysis no statistical significative association with smoking, menarche, nuliparity, the aspect of the menstrual cycle, the level of education, IMC, the depressive syndrome, but not with factors linked to the thyroidian pathology: the TSH level, the level of ATPO, Basedow disease or Hashimoto thyroiditis, thyroidian funtion at debute. In exchange there was found a significant association with the age at the AITD diagnosis< 35 years, infertility, the level of AOA (UI/ml), buffs and perspirations.
The logistic regression makes the prediction if a certain event will produce or not, in the situation in which it is influenced by a series of identified factors; shapes linkes between numerical variables and categorial variables .
The obtained result was SCOR = -2.2022 x diagnosis age AITD < 35 years (1/0 ) +1.9328 x Infertility (1/0) - 0.3792 x anti-ovarian antibodies (UI/ml), with the score criterion ≤ - 1.8359, with specificity of 88% and sensibility of 90%,the area under the curve AUC1=0.91. The score containes only 3 of the 5 variables introduced at modellation.
The sign of the logistic coefficient of regresion shows that the sense of correlation between the predictive factor and the interest event (the installation of the menopause before 45 years).
The debute of AITD before 35 years is correlated negativelly with precocious menopause, with the coefficient - 2.2022, fact that suggests the installation of the AITD before the debute of the process of ovarian ageing influences it less. The rate of accelerated follicular growth begins at 35 years and reaches the critical point at 37.5 years.
The infertility is correlated positivelly with early menopause, with the coefficient 1.9328 and OR=6.9, the factor may suggest that there is already an alteration of the ovarian reserve and of the natural fecundability.It is possible the subestimation of this factor through the postponement of the first birth after 35 years. Regarding the risk representation it is observed that the infertility affects dramatically the risk of menopause if it is present.
The level of AOA (UI/ml), is negatively correlated with early menopause, with the coefficient - 0.3792. The rezult should be interpreted considering that the AOA values in our study are in proportion of 99.4 % in the reference interval, that the test was done at the enroll in the study, in many cases at distance from the debute of AITD or menopause and that it didn′t proved its sensibility in the identification of the autoimmune ovarian failure.
The sources of uncertainty in our research could be:
selecting the cases of AITD through a gynecology service, the criterion of age at diagnosis for the censoring of the cases, the imprecision in the menopause data anamnestically supplied, the determination of the AOA at the enroll in the study.
According to our study results it is confirmed the higher frequency of the association of early menopause and /or premature with AITD than its prevalence in the general
The debute of AITD in an infertile patient ,with the age ≥ 35 years, with the value AOA in the reference interval may warn the clinician that he finds himself in ahead a case with occulte ovarian failure, that may affect the prognosis of a procedure of assisted reproduction.
1. Gold, E.B., et al., Relation of demographic and lifestyle factors to symptoms in a multi-racial/ethnic population of women 40-55 years of age. Am J Epidemiol, 2000. 152(5):
2. Bachelot, A., et al., Phenotyping and genetic studies of 357 consecutive patients presenting with premature ovarian failure. Eur J Endocrinol, 2009. 161(1): p. 179-87.
3. Goswami, D. and G.S. Conway, Premature ovarian failure.
Horm Res, 2007. 68(4): p. 196-202.
4. Murabito, J.M., et al., Heritability of age at natural menopause in the Framingham Heart Study. J Clin Endocrinol Metab, 2005. 90(6): p. 3427-30.
5. Murabito, J.M., et al., Genome-wide linkage analysis to age at natural menopause in a community-based sample: the Framingham Heart Study. Fertil Steril, 2005. 84(6): p.
6. Murray, A., et al., Common genetic variants are significant risk factors for early menopause: results from the Breakthrough Generations Study. Hum Mol Genet. 20(1):
7. Kinney, A., et al., Smoking, alcohol and caffeine in relation to ovarian age during the reproductive years. Hum Reprod, 2007. 22(4): p. 1175-85.
8. Rebar, R.W., Premature ovarian failure. Obstet Gynecol, 2009. 113(6): p. 1355-63.
9. Luborsky, J., Ovarian autoimmune disease and ovarian autoantibodies. J Womens Health Gend Based Med, 2002.
11(7): p. 585-99.
10. Alviggi, C., et al., Biological versus chronological ovarian age: implications for assisted reproductive technology.
Reprod Biol Endocrinol, 2009. 7: p. 101.
11. Gleicher, N., Autoantibodies in infertility: current opinion.
Hum Reprod Update, 1998. 4(2): p. 169-76.
12. Weetman P. Anthony, Humana Press, 2008, New Jersey, p 331-356
13. Rees M - Management of the Menopause: The Handbook, in Purdie DW (eds), 2006, Royal Societz of Medicine Press, London, 141-149
14. Hollowell, J.G., et al., Serum TSH, T(4), and thyroid antibodies in the United States population (1988 to 1994):
National Health and Nutrition Examination Survey (NHANES III). J Clin Endocrinol Metab, 2002. 87(2): p.
15. Kok, H.S., et al., Genetic studies to identify genes underlying menopausal age. Hum Reprod Update, 2005.
11(5): p. 483-93.
16. Henderson, K.D., et al., Predictors of the timing of natural menopause in the Multiethnic Cohort Study. Am J Epidemiol, 2008. 167(11): p. 1287-94.
17. Boelaert, K., et al., Prevalence and relative risk of other autoimmune diseases in subjects with autoimmune thyroid disease. Am J Med. 123(2): p. 183 e1-9.