Przegląd Antropologiczny � Anthropological Review • Vol. 64, 57-72 (2001) 57

Age at menopause and reproductive determinantsin Polish women

Maria Kaczmarek, Anita SzwedInstitute of Anthropology, Adam Mickiewicz University, Fredry 10, 61-701 Poznań,Poland; E-mail: makac@main.amu.edu.pl

ABSTRACT Selected reproductive determinants of age at menopause wereanalyzed using data from a representative sample of 4354 women aged 35-65years. The data were collected in cross-sectional surveys carried out inWielkopolska and Western part of Poland in the years 1998 to 2001. Meno-pausal status and median and mean ages at menopause were estimated with theuse of retrospective and status-quo methods. It was found that the median ageat natural menopause, estimated by probit analysis, was 50.1 years. Women,who had been treated with HRT (Hormonal Replacement Therapy) in theperiod of perimenopause, underwent menopause significantly later � theirmedian age at menopause was 52.9 years. Results of ANOVA revealed that ofall characteristics of female reproduction taken into consideration, age atmenarche and length of menstruation cycle had significant influence on ageat menopause. Menopause occurred markedly later in women who began tomenstruate after 15 years of age and whose menstrual cycles lasted longer than32 days. Reproductive determinants were considered as a complex factor.Stepwise multiple regression analysis with backward elimination revealed thatthe reproductive determinants including the length of menstruation cycle, ageat menarche, age at first and last pregnancies and length of lactation in combi-nation affected age at menopause.KEY WORDS menarche, menstruation cycle, menopausal status, reproductionPrz. Antropol. � Anthropol. Rev. (2001), vol. 64, pp. 57�72, Figs. 5, Tables 4. ISBN 83-86969-64-4, ISSN 0033-2003

Introduction

Unlike the decline of other physio-logical functions, including male repro-duction, the rate of senescent changes infemale reproductive capacity showsstriking abruptness (Fig. 1). A transitionfrom reproductive to non-reproductive

phase in woman�s life considerably longbefore the end of her life is a uniquefeature of human female [HILL andHURTADO 1991; PAVELKA and FEDIGAN1991; AUSTAD 1994]. In 1821 a Frenchphysician, Gardanne invented the term�menopause� to describe the phenome-non of this transition [Tilt 1870; cited

Maria Kaczmarek, Anita Szwed58

after LOCK and KAUFERT 2001]. Meno-pause signifies the permanent cessationof ovarian follicular function, thus themechanism responsible for the meno-pause refers to the continuous loss ofovarian follicles up to the point at whichmenstrual cycles are increasingly vari-able in length and finally cease.

Explanation of the origin and exis-tence of menopause can be classifiedamong adaptive or non-adaptive theo-ries. The adaptive theories of meno-pause assume that age-specific increas-ing probability of reproductive failureand risk of death in childbirth make itevolutionally more advantageous toredirect reproductive energies towardsmaternal investment in existing progenyby increasing their survivorship andpotential fertility (the mother hypothe-sis) [WILLIAMS 1957; GAULIN 1980;LANCASTER and LANCASTER 1983;PECCEI 1995a,b] or by caring directlyfor grandchildren (the grandmother hy-pothesis) [PAVELKA and FEDIGAN 1991;AUSTAD 1994; PECCEI 2001]. The non-adaptive theories of menopause consider

this phenomenon as a recent artifactcreated by the relatively sudden length-ening of human life over the past severalcenturies. Implications of either scenarioare significant for the understanding ofthe role the menopause is likely to playin women�s aging, quality of life, anddifferential susceptibility to disease.

Menopause is defined as a complexbiosocial and biocultural phenomenon,which occurs in a woman�s mid-lifephase. Although biological changesassociated with this phase of life areessentially universal, the subjective ex-periences of individual women varyaccording to sociocultural settings.Cross-cultural studies of multiple per-spectives demonstrate a continuousfeedback relationship of biological andsociocultural variables [LOCK andKAUFERT 2001]. Therefore, in order toprovide an insight into the universalnature of the menopause and individualexperiences of women undergoing me-nopause, a population study of Polishwomen was undertaken. This paper, partof a larger work, is focused on the distri-bution of the age at menopause in rela-tion to the reproductive determinants.

MethodologySurvey and sample

Investigations were carried out in theyears 1998 to 2001 in the framework ofthe project �Biological aspects of humanaging�. In the years 1998-2000, a pilotstudy was undertaken in Poznań and inthe Wielkopolska region. A sample ofmore than 2400 women were inquiredwith the use of a menopause-specificquestionnaire. Anthropometric measure-ments and body composition were exa-mined at the same time. Results of this

0

20

40

60

80

100

20 30 40 50 60 70 80

Age (years)

1

2

3

4

5

Fig. 1. Rate of human senescent changes in repro-ductive capacity and selected physiological functions:(1) female reproductive capacity; (2) male reproductivecapacity; (3) max. breathing capacity; (4) vital capacity;(5) basic metabolism [Kinsey et al. 1948; Shock 1961;Wood 1990, cited after Austad 1994].

Age at menopause and reproductive determinants in Polish women 59

investigation were presented in: KACZ-MAREK and SZWED [2000, 2001], andSZWED [2001]. The experience gainedfrom the pilot study led the authors toexpand the investigation to the regionssituated in the West part of Poland. Lo-cations of samples examined for meno-pause are presented in Fig. 2.

The menopause-specific question-naire, used in the present study, wasaddressed to women older than 35 yearsof age, and focused mainly on womenaged 45-54 years, as it is known that agreat majority of women experiencenatural menopause within this agerange. Of the 5500 respondents, replieswere obtained from 4354. This gives a79.2% final response rate. The figuremeets the requirements of sample qual-ity. The reliability of replies was esti-mated by the repeated �test-retest� pro-cedure [FERGUSON and TAKANE 1999].After a one-year interval, 100 women ofpost-menopausal status were selected atrandom and asked to fill in the samemenopause-specific questionnaire. Thediscrepancies in replies between the two

inquiries were tested according to thecharacter of variables. Statistical signifi-cance of the differences between quan-titative variables was tested with the useof the t-test and between qualitativevariables with the use of the chi-squaretest. In general, a high consistence wasfound between the replies in the twoinquiries, except for the statisticallysignificant difference in the replicates ofthe age at menarche [SZWED 2001].

The age of the women who partici-pated in the survey varied from 27.4 to93.0 years with a mean of 47.6 and me-dian of 47.2 years. The age was distrib-uted into 5-year interval cohorts. Thelargest groups are made of women aged45-49 years (28.3%) and 50-54 years(21.7%). Subjects older than 50 yearsconstitute 33.7% of the total sample.Distribution of women according to theage category is shown in Fig. 3.

The socioeconomic stratification, de-scribed in the present paper in terms ofthe degree of urbanization and educa-tional level, is shown in Fig. 4. Majorityof subjects in the sample comes fromurban settings: 30.3% from small townsand 25.4% from large cities. Those fromvillages constitute 18.3% of the total

Fig. 2. Location of samples examined for the meno-pause in 1998-2001.

Fp3y

30

ig. 3. Percentage age distribution in sample of womenarticipating in menopausal survey: (1) x-34 years; (2)5-39 years; (3) 40-44 years; (4) 45-49 years; (5) 50-54ears; (6) 55-59 years; (7) 60-x years.

0

5

10

15

20

25

1 2 3 4 5 6 7

Maria Kaczmarek, Anita Szwed60

follicular activity. Following the rec-ommendation of the WHO, we use thisterm to describe the spontaneous cessa-tion of menstruation preceded by theoccurrence of amenorrhea for 12 months[Report... 1981, 1996]. Women whosemenses stopped as a result of surgicalbilateral oophorectomy, with or withouthysterectomy, or X-ray or radium treat-ment were sampled separately and ex-cluded from subsequent analysis. Thenumber of excluded subjects was 313;which is 7.1% of all the subjects underinvestigation.

villag

esm

all to

wn

town

large

city

primary

vocationalsecondary

academic

0

2

4

6

8

10

12

14

16

Fig. 4. Percentage distribution of women accordingto their education levels and dwelling places.

sample. The level of education in thesample under study reflects the generalpattern of women�s education in Poland.The largest group consists of womenwith secondary education (45.5%) and isfollowed by the group with primary andvocational levels of education (29.4%).Women with academic education make -23.5% of the total sample. A clear gradient of education and the degree ofur-banization was found. The higher(lower) the degree of urbanization, thehigher (lower) the educational level (seeFig. 4).

The distribution of marital status inthe sample under examination, with80.6% of married women versus only0.6% of women living in concubinage,reflects a norm current in Polish society,in which marriage is a prevailing formof the male-female relationship.

Menopausal status of womenThe term �menopause�, used in the

present study, is defined according to itsderivation as the last menstrual period(LMP) as a result of the loss of ovarian

Considering the above given defini-tion, the term �post-menopause� denot-ing the cessation of the menses for aperiod of 12 months, was applied to1300 subjects � 29.8% of the total sam-ple. Out of this number, 1033 womenhad undergone natural menopause, i.e.,menopause resulting from spontaneousloss of ovarian activity. 267 of post-menopausal women declared that in theperiod of perimenopause they had un-dergone the hormonal replacement ther-apy in order to alleviate symptoms andcomplaints accompanying the beginningof the menopause. It is known that fac-tors used to treat the accompanyingsymptoms in the period prior to themenopause can have various effects onthe cessation of reproductive functions.Therefore, those women were sampledin a separate group and labeled as hav-ing gone through �hormonally con-trolled menopause�.

The status of 2410 women (55.3% ofall the subjects under investigation) whohad menstruated within the last 3months was defined as premenopausal.331 women, whose menses had termi-nated in the last year, but whose periodof amenorrhea had not exceeded 12

Age at menopause and reproductive determinants in Polish women 61

months, were defined as being at themenopause.

Considering the biological dimensionof the menopause, further analysis wasundertaken in relation to those womenwhose status was defined as undergoingthe natural menopause. Thus, the totalnumber of subjects used for statisticalcalculations was 1033.

Age at the menopause – study’s designThe individual age at menopause was

obtained by asking postmenopausalwomen to recall their age at the lastmenstrual period (LMP). The accuracyof the replies depends on the interval oftime since the menopause � the older thewoman, the more marked the tendencyto round off the date of the onset and theend of the menstruating period. Theaverage age at menopause was estimatedwith the use of the retrospective method,by calculating the mean and median agefrom individual data. In the status-quoapproach each age-specific prevalenceof the postmenopausal state was trans-formed to a probit scale, and median agewas calculated from these transformeddata [FINNEY 1952]. The status-quomethod is a more reliable measure, sinceit can be determined on the basis of eachwoman�s �yes� or �no� reply referring toevents that have occurred only duringthe previous year. However, the medianmenopausal age for a population, theresultant of the retrospective method,though considerably biased due to inac-curacy in reporting, is useful for pur-poses of quantitating the relationshipbetween age at menopause and its bio-logical and sociocultural determinants.Therefore, in the present paper we havereferred to the results of both methods.

Statistical computations were made withthe use of the Statistica 5.5 programpackage [STATSOFT, INC 1999, Statis-tica for Windows].

Selected characteristics of femalereproductive period

Female reproduction history is welldescribed by events such as the onsetand the end of reproductive capacity,pattern of menstruation cycles, potentialand real reproductive period, abortions,pregnancies and parity. Those charac-teristics were employed in the presentstudy, bearing in mind their biologicaland sociocultural backgrounds. Theonset of the reproductive period wasdated by the age at menarche. The timeinterval between the age at menarcheand the age at menopause was used forestimating the maximum length of po-tential reproductive period. While esti-mating the impact of reproductive activ-ity on age at menopause we used twovariables: number of pregnancies andparity. The latter is defined as a numberof live births ever given, whereas the�number of pregnancies� variable in-cluded spontaneous abortions, live andstillbirths. Factors such as length ofmenstrual cycles, regularity of the men-ses, number of pregnancies, length oflactation are supposed to delay themenopause, therefore these variableswere also included into the analysis.

It is known that the variables of thefemale reproductive period are relatedwith one another to the extent theyconstitute a complex factor. It is impor-tant to bear this in mind while inter-preting the influence of a single repro-ductive characteristic on the age atmenopause.

Maria Kaczmarek, Anita Szwed62

M

NHA

NHA

ResultsDetermining the age at menopause

It has been known from data in the lit-erature, that women living in developedcountries currently reach menopausearound the age of 51, whereas thosefrom underdeveloped countries, a fewyears earlier (Menopause Core CenterStudy, cited after PARAZZINI et al.1992]. However, individual recordsindicate a great variety in the age atmenopause. While describing a normalrange of the age distribution within apopulation, one may see its negativelyskewed shape with a wider scatter ofwomen reaching menopause at youngerrather than older ages. The descriptionof the normal range is important forestablishing the upper and lower limitsof normality while estimating the me-dian age at menopause, and, practically,in order to establish appropriate defini-tions for the terms of premature anddelayed menopause.

In the present work, mean and medianages at menopause have been calculatedwith the use of the retrospective (the ageat the last menstrual period � LMP) andthe current status of menstrual cycles(the status-quo method). Furthermore,apart from the group of women under-going natural menopause we selected a

group of women whose menopause hadoccurred under hormonal treatment.Results of descriptive statistics with theuse of retrospective and status quomethods are presented in Table 1. Fig-ures in the table revealed that meno-pause occurred considerably later inwomen undergoing hormonal replace-ment treatment. Median age at hormon-ally controlled menopause according tothe retrospective and status-quo methodswas 50 and 52.9 years, respectively. Theage at natural LMP ranges from 37.7 to57.7 years, with median age of 49.4years. While using the status-quomethod the median age at menopausewas estimated at 50.1 years of age. Theresultant differences meet statisticalsignificance as derived from the chi-square and Kruskal-Wallis�s tests. Thisfinding has convinced us that it wasnecessary to separate those two groupsof women.

In order to delimit the ages of prema-ture and delayed menopause, a purelystatistical point of view indicates 45.3years for the lower and 53.5 years forthe upper limits of normality (the me-dian diminished or increased by thequartiles range). Thus, natural meno-pause occurring before the age of 45years is defined as premature and thatoccurring after 53.5 years, as delayed.

Table 1. Descriptive statistics of the age at menopause � results of the retrospective and status-quoapproaches

Retrospective approachenopause N Mean Min Max SD Median Lower

quartileUpper

quartileRange

atural 851 48.9 37.7 57.7 3.4 49.4 46.9 51.0 4.1ormonally controlled 221 49.8 41.2 58.5 3.2 50.0 48.0 51.8 3.8rtificial 293 44.2 27.0 54.0 5.1 45.0 41.0 48.2 7.2

Probit analysisatural 1033 50.1 Var = 0.009ormonally controlled 221 52.9 Var = 0.015rtificial 313 45.1 Var = 0.031

Age at menopause and reproductive determinants in Polish women 63

CharaAge a

123

Lengt123

Regulyn

Marit12

Numb01234

Parityn1234

Reproductive determinants influencingthe age at menopause

Aiming at justifying the controversialfindings known from the literature, wehave analyzed the relationships betweenfactors causing long anovulatory periodsduring reproductive years, and age atmenopause. Results of ANOVA interms of statistics F, mean ages andstandard deviation at menopause instrata of selected menstrual and repro-ductive characteristics are given inTable 2; the graphic representation isshown in Fig. 5.

Considering the marital status of oursubjects, we did not find a significantimpact of this variable on age at meno-pause. Mean ages at menopause of nevermarried women and of those currently

married, widowed and divorced wereslightly different.

There are many contradictions con-cerning the relationship between theonset and the end of the reproductiveperiod dated by the age at menarche andthe age at menopause. In our sample theage at menarche ranges from 9 to 20years with the mean of 13.5 years andthe median of 14.0 years. In a greatmajority of subjects menarche occurredat the age category of 12-14 years(81.4%). Very few experienced men-arche earlier than at 12 years of age(5.2%). In 13.4% of subjects, menarcheoccurred after the age of 15 years. Thedistribution of age at menarche is posi-tively skewed with a wider scatter ofwomen beginning to menstruate after 14years of age. We found a significant

Table 2. ANOVA � mean age at menopause in strata of selected menstrual and reproductivecharacteristics: statistics F and level of significance (p)

cteristics N Mean age SD F and pt menarche:. less than 12 years of age. 12-14 years of age. after 15 years of age

42661109

49.3248.7250.18

3.53.43.3

F = 8.88p < 0.001

h of menstruation cycles:. less than 28 days. 28-32 days. longer than 32 days

26846499

47.5149.3251.35

3.53.12.7

F = 59.86p < 0.001

arity of menses:eso

682140

48.9149.02

3.33.6

F = 0.14p = 0.71

al status:. never married. married, widowed, divorced

43800

48.8249.01

2.93.4

F = 0.13p = 0.72

er of pregnancies:

and more

50127334193147

49.0648.8548.8248.8749.61

3.33.43.43.13.7

F = 1.55p = 0.18

:ulliparous 61 48.94 3.7 F = 0.36

child children children and more children

141371176101

49.0448.9248.8949.34

3.43.33.23.7

p = 0.83

Maria Kaczmarek, Anita Szwed64

correlation between the age at menarcheand the age at menopause (linear regres-sion analysis indicates r = 0.11). Theearlier (later) age at menarche, the ear-lier (later) age at menopause.

Similarly, we have found a significantcorrelation between the age at meno-pause and the length of menstrual pe-riod. The longer the menstrual periodlasts, the later the menopause occurs.

While considering both variables � thelength of the menstrual cycles and theage at menarche, results of MANOVAconfirmed the relationships betweenthese variables and the age at meno-pause, however there is no interactionbetween these factors. The later the on-set of the menses and the longer themenstrual period lasts, the later themenopause occurs. When menarche

±SD

±SE

Mean

Age at menarche

44

46

48

50

52

54

1 2 3

±SD

±SEMean

Lenght of menstruation cycles

42

44

46

48

50

52

54

56

1 2 3

±SD

±SEMean

Regularity of menses

44

46

48

50

52

54

1 2

±SD±SE

Mean

Marital status

45

47

49

51

53

1 2

±SD

±SEMean

Number of pregnancies

44

46

48

50

52

54

1 2 3 4 5

±SD±SEMean

Parity

44

46

48

50

52

54

1 2 3 4 5

Fig. 5. Graphic representation of the results of ANOVA (for explanation see Table 2).

Age at menopause and reproductive determinants in Polish women 65

CLAAAL

omt4Wataytwsal

lpcanamsNefmmwcdIw

np

Table 3. Results of stepwise multiple regression analysis with backward elimination for reproductivedeterminants of the age at menopause

haracteristics b SE t pength of menstruation cycles (in days) 0.15 0.048 3.10 0.001ge at menarche (in years) 0.08 0.05 1.69 0.09ge at first pregnancy (in years) 0.11 0.06 1.91 0.06ge at last pregnancy (in years) 0.08 0.05 1.39 0.17ength of lactation (in months) 0.05 0.05 1.05 0.29

ccurs before 12 years of age and theenstrual cycle lasts less than 28 days

he menopause occurs on average at6.2 years of age (SD = 3.6 years).hen menarche occurs after 15 years of

ge and the menstrual cycle is longerhan 32 days, the menopause occurs onverage at 52.8 years of age (SD = 3.1ears). When long-lived irregularity ofhe menses and the age at menopauseere tested, we failed to find relation-

hips between these variables. Thus, aget menopause does not depend on theong-lived regularity of the menses.Results of regression analysis indicate

ack of correlation between age at meno-ause and age at first and last pregnan-ies (correlation coefficient r = 0.065nd p = 0.06 for the age at first preg-ancy and r = 0.068, p = 0.08 for the aget last pregnancy). This means thatenopause occurs at a definite age irre-

pective of the time of pregnancies.umber of pregnancies did not influ-

nce the age at menopause either. Weound no relationship between the age atenopause and parity. The age atenopause was similar in nulliparousomen and those having one or more

hildren. The age at menopause did notepend on the length of lactation either.t was similar irrespective of how longomen breast-fed their babies.Considering all reproductive determi-

ants, selected for the purpose of theresent study, results of the stepwise

multiple regression analysis with back-ward elimination, presented in Table 3,revealed that variables essential for theage at menopause are the following: thelength of menstruation cycles, the age atmenarche, the age at first and last preg-nancies and the duration of the lactationperiod. Some of these variables, takenseparately, as for example the age atfirst and last pregnancies or the durationof lactation period, had no significantinfluence on the age at menopause, buttheir influence intensified when theyinteracted. Altogether, they explain3.9% of the variation of the age at me-nopause, thus their magnitude in modi-fying the age at menopause is consider-able.

Discussion

The age at menopause has been con-sidered to be an indicator of populationhealth, in terms of not only individualwomen�s but also of entire population�swell-being. At the same time, the age atmenopause might be used as a marker ofincreased risk for chronic diseases, suchas cardiovascular diseases, increasedrate of bone loss or breast cancer. Thus,the estimation of the age at menopausein populations of different biologicaland sociocultural background plays animportant role in current investigationsof the health of aging women [LEIDYSIEVERT 2001].

Maria Kaczmarek, Anita Szwed66

In Poland, the investigation of themiddle-aged women has had a longtradition and it goes as far back as to theresearch by STOŁYHWOWA [1938].Majority of hitherto existing populationstudies have focused on estimating theage at menopause, its biological andsociocultural determinants, and symp-toms experienced at perimenopausal pe-riod [GÓRSKA 1949; BIELICKI andWELON 1964; PANEK 1967; BOCHEŃ-SKA 1978; BARON 1968; STOKOWSKI1968; JASICKI and WOLAŃSKI 1972;SZEMIK 1984; KONIAREK and ŻUKOW-SKI 1987; RODZIEWICZ-GRUHN 1998,KACZMAREK and SZWED 2000]. It isknown that every society has its ownunderstanding of the norms regardingthe timing of marriage, family formationand childbearing as well as cultural val-ues regarding the role of motherhood, orfitness of parents changing with age.The present study provides some quan-titative clues to the understanding of thepossible impact of reproductive deter-minants on menopausal age in Polishwomen.

There is a general agreement, as far asan average age at menopause is con-cerned, that menopause occurs at around50 years of age. However, there arediscrepancies in detailed figures. Itseems that the methodological back-ground may be an important source ofthe differences here. As McKinleyclaims, the majority of investigations arebiased by imprecise methodology[MCKINLEY et al. 1972]. Many re-searchers used to draw conclusions ongrounds of data collected in medicalcenters [BALLINGER et al. 1987;STANOSZ et al. 1993; MIZIA-STEC et al.1998] or in selected groups of women[MARIN et al. 1993; PIPLAI 1991]. There

are also discrepancies in the precisedefinition of menopause. According tothe recommendation of the WHO, thepostmenopausal status of a womanmight be defined as the age at which thelast menstrual period occurs, followedwith a 12 months� period of amenorrhea.Reviewing literature on the menopauseone may find out that there are severalresearchers who do not follow theWHO�s recommendations and use dif-ferent criteria for the estimation of theage at menopause. The period of amen-orrhea, referred to in their works, rangesfrom 6 months, as for example inLINDQUIST and BENGTSSON [1980],HOLTE [1992], KULIKOWSKI and TOMA-SZEWSKA [1995], WARENIK-SZYMAN-KIEWICZ [1996], to 9 months, as postu-lated by FROMMER [1964]. KONIAREKand ŻUKOWSKI [1987], in turn, postulatea 10 months� period of amenorrhea to beemployed for the estimation of the ageat menopause. There are some research-ers who ignore the information aboutthe precise definition of menopause[RODZIEWICZ-GRUHN 1998]. Someothers, as for example, Neugarten andKraines [cited after MCKINLEY et al.1972] rely on subjective responses ofwomen and their own understanding ofmenopause and postmenopausal status.We have to bear in mind that a mean ageat menopause is calculated for post-menopausal women, therefore the esti-mation of the age at menopause is ofgreat importance for both the range ofvariation in age distribution and thevalue of the mean age. The methodo-logical bias refers also to statistical pro-cedures such as calculation of the meanage at menopause either retrospectivelyor with the use of probit analysis. An-other source of error lays in reducing the

Age at menopause and reproductive determinants in Polish women 67

SourceBielicki & WeloBocheńska [197Baron [1968]Koniarek & ŻukPresent data

median age calculated by probit trans-formation by the period of amenorrhea,thus from the median age at menopausecalculated for the sample, a 12 months�period is subtracted [MCKINLEY et al.1972; KONIAREK and ŻUKOWSKI 1987].

As revealed by our data, in womenwho had used HRT in the perimeno-pausal period the age at menopauseoccurred significantly later. Samplingthose women together with those whohad gone through natural menopausemight delay the mean age at menopauseand lead to false conclusions. Therefore,in order to investigate the relationshipbetween menopause and reproductivedeterminants we excluded from oursample all those women who had expe-rienced hormonally controlled meno-pause. Thus, average age at menopausein our sample, estimated by probits was50.1 years. Comparative data for Polishwomen, presented in Table 4, indicateminor discrepancies, which might becaused by differences in the methodolo-gies used. In contrast to the trend to-wards earlier menarche noted during thelast decades in Poland and associatedwith improved nutrition, health andsocioeconomic conditions [MILICER,1966; ŁASKA-MIERZEJEWSKA et al.,1982; BIELICKI and WELON, 1982], nosecular trend has been observed withregard to the age at menopause (seeMALINA [1979], JOHNSTON [2001], andTable 4). The absence of a secular trendin the age at menopause could be ex-

plained by the length of the time intervalbetween changes in socioeconomic con-ditions in childhood and physiologicalmanifestations in adulthood [LEIDYSIEVERT 2001: 430].

Our findings provide quantitative evi-dence that the majority of the reproduc-tive characteristics selected for the pur-pose of the present study are not corre-lated with the age at menopause. Severalobservers have reported later ages atnatural menopause among married orwidowed women compared to nevermarried or divorced women [MCKINLEYet al. 1972; STANFORD et al. 1987]. Anearlier menopause in unmarried womenwas reported by JASZMAN et al. [1969],MCKINLEY et al. [1972], LEIDYSIEVERT et al. [2001]. These are contro-versial findings, however explanationssuggesting an effect of marriage on theage at menopause involving pheromonalinfluence of a continuing presence of amale in the household [LEIDY SIEVERTet al. 2001] are still to be considered.Although attractive, this explanationneeds further substantiation. In our sam-ple we did not find any correlation be-tween these factors. Our findings corre-spond with the results of MedicalWomen�s Federation Research carriedout in 1933 (see BENJAMIN [1960] orMACMAHON & WORCESTER [1966]).

It is supposed that childbearing andweaning period might stop the atresion ofovum because of the anovulatory menses,and in this way delay the age at meno-

Table 4. Age at menopause in Polish women according to various authorsN Place of investigation Mean age (yrs.)

n [1964] 341 Wrocław 49.338] 564 Radom 50.1

Poland 49.2owski [1987] 1655 Wrocław 49.66

4354 Wielkopolska, West Poland 50.1

Maria Kaczmarek, Anita Szwed68

pause. The onset of the childbearing ca-pacity in women can be dated by men-arche, the beginning of the menstrualcycles. Although, the age at menarchedoes not imply an actual fecundity, as thelatter is determined by the first ovulation,it is sufficiently well suited for practicalpurposes. The end of childbearing comesfrom data on women�s menopausal statusrelative to age, although, by observation,fecundity ceases some time before themenopause [BONGAARTS and POTTER1983]. Gray estimates the differencebetween median age of infecundity andmedian age of menopause to be around 8years [GRAY 1979].

There are studies postulating that anearly age of menarche might lead to alate menopause (for example PARAZINIet al. [1992]). In our sample we havefound a significant relationship betweenthose variables, however, with the oppo-site tendency � the earlier the menarchethe earlier the menopause. This resultdoes not surprise us since the same wasstated by CRAMER et al. [1995],TREOLAR [1974], BEAL [1983], REBATO[1988], WHELAN et al. [1990].

We have found a statistically signifi-cant relationship between the length ofthe menstrual cycle and the age atmenopause. Short menstrual cycles (lessthan 28 days) are related with earliermenopause, whereas long menstrualcycles with later menopause. This find-ing corroborates the results obtained byTREOLAR [1981], WHELAN et al. [1990],CRAMER et al. [1995], Report... [1996].STANFORD et al. [1987] found thatwomen whose menses were regularbefore 25 years of age or before the firstpregnancy experienced menopausearound 2 years earlier than womenwhose menstrual cycles were irregular.

Our findings concerning the lack ofcorrelation between the age at first andlast pregnancies with the age at meno-pause are controversial in the light of thereference data. WHELAN et al. [1990]noticed that women who at first preg-nancy had been younger than 30 yearsof age experienced earlier menopause.But there are also indications of a lackof relationship between these variables,as found in our sample [STANFORD et al.1987; WALSH 1978; CRAMER et al.1995].

The correlation between parity andmenopausal age seems to be equallycontroversial. It has been suggested thatconditions causing long anovulatoryperiods during reproductive years maybe associated with delayed menopause[STANFORD et al. 1987; WHELAN et al.1990; PARAZZINI et al. 1992]. However,results of surveys may cast doubt uponthis statement. A consistent relationshipbetween the number of births and earliermenopause was found in some works(findings of the BCDDP Project; citedafter PARAZZINI et al. [1992]). In oursample we did not find a significantrelationship between these variables, thesame was stated in other studies[BRAND & LEHERT 1978; VANKEEP etal. 1979].

Conclusions

Our findings, deriving from the sam-ple, which meets the requirements ofsample quality, are conclusive for thecontroversies concerning the age atmenopause and its reproductive deter-minants and can be summarized in thefollowing statements:1. Median age at natural menopause in

Polish women is 50.1 years.

Age at menopause and reproductive determinants in Polish women 69

2. There is a significant correlationbetween the age at menopause and theage at menarche. The earlier the men-arche, the earlier the menopause.

3. There is a significant correlationbetween the age at menopause and thelength of menstrual cycles. The longerthe menstrual cycles, the later the age atmenopause.

4. Characteristics of female reproduc-tion, such as age at menarche, length ofmenstrual cycles, age at first and lastpregnancy and length of lactation pe-riod, interacting in a complex way, ac-count for the range of variability in theage at menopause.

ReferencesAUSTAD S.N., 1994, Menopause: An Evolutiona-

ry Perspective, Exp. Gerontol., 29, 255-263BALLINGER C.B., M.C.K. BROWNING, A.H.W.

SMITH, 1987, Hormone profiles and psycholo-gical symptoms in peri-menopausal women,Maturitas, 9, 235-251

BARON J., 1968, Klimakterium i senium, Pamięt-nik XVII Zjazdu Pol. Tow. Gin., Kraków, 2,15-18

BEAL C.M., 1983, Ages at menopause and men-arche in a high-altitude Himalayan population,Ann. Hum. Biol., 10, 365-370

BENJAMIN F., 1960, The age of the menarche andof the menopause in white South African wo-men and certain factors influencing these ti-mes, Afr. Med. J., 34, 316-321

BIELICKI T., Z. WELON, 1964, Próba oszacowa-nia strat rozrodczych u kobiet wiejskich, Mate-riały i Prace Antropol., 70, 211-216

BIELICKI T., Z. WELON, 1982, Growth data asindicator of social inequalities: the case ofPoland, Yrbk Phys. Anthropol., 25, 153-167

BOCHEŃSKA Z., 1978, Zmiany w rozwoju osobni-czym człowieka w świetle trendów sekularnychi różnic społecznych, 2nd Ed., Praca Monogra-ficzna AWF Kraków, 5

BONGAARTS J., R.G. POTTER, 1983, Fertility, Bio-logy and Behavior: An Analysis of the Proxi-mate Determinants, Academic Press, New York

BRAND P.C., P.H. LEHERT, 1978, A new way oflooking at environmental variables that mayaffect the age at menopause, Maturitas, 1, 121-132

CRAMER D.W., H. XU, B.L. HARLOW, 1995, Does�incessant� ovulation increase risk for earlymenopause?, Am. J. Obst. Gynaecol., 172,568-573

FERGUSON G.A., Y. TAKANE, 1999, Analizastatystyczna w psychologii i pedagogice, PWN,Warszawa.

FINNEY D.J., 1952, Probit analysis, 2nd Edition,Cambridge University Press, London

FROMMER D.J., 1964, Changing age of the meno-pause, Br. Med. J., ii, 349-351

GAULIN S.J.C., 1980, Sexual dimorphism in thehuman post-reproductive lifespan: possiblecauses, J. Hum. Evol., 9, 227-232

GÓRSKA I.I., 1949, Badania nad wiekiem dojrze-wania i długością okresu rozrodczego u kobietwiejskich, Spraw. PAU, 10, 635-638

GRAY R.H., 1979, Biological factors influencingthe reproductive lifespan, [in:] Natural Ferti-lity: Patterns and Determinants of Natural fer-tility, H. Leridon & J. Menken (eds.), Ordina,Liège, pp. 25-40

HILL K., A.M. HURTADO, 1991, The evolution ofpremature reproductive senescence and meno-pause in human females: An Evaluation of the�Grandmother Hypothesis�, Hum. Nat., 2,313-350

HOLTE A., 1992, Influences of natural menopauseon health complains: A prospective study ofhealthy Norwegian women, Maturitas, 14, 127-141

JASICKI B., N. WOLAŃSKI, 1972, Rozwój ontoge-netyczny a formowanie się gatunku Homo sa-piens jako funkcja zmian przystosowawczychdo środowiska, [in:] Czynniki rozwoju człowie-ka, N. Wolański (ed.), PWN, Warszawa, pp.21-56

JASZMAN L., N.D. VANLITH, J.C.A. ZAAT, 1969,The age at menopause in the Netherlands, Int.J. Fert., 14, 106-111

JOHNSTON S.L., 2001, Association with age atnatural menopause in Blackfeet women, Am. J.Hum. Biol., 13, 512-520

KACZMAREK M., A. SZWED, 2000, Wpływ warun-ków życia kobiet na wiek menopauzy orazprzebieg klimakterium, [in:] Auksologia a pro-mocja zdrowia, A. Jopkiewicz (ed.), KieleckieTow. Nauk., Kielce, vol. 2, 67-75

Maria Kaczmarek, Anita Szwed70

KACZMAREK M., A. SZWED, 2001, Wiek meno-pauzy a rozrodczość i styl życia kobiet, [in:]Menopauza, T. Pertyński (ed.), Wyd. AD1,Łódź, pp. 15-23

KONIAREK J., W. ŻUKOWSKI, 1987, Wiek meno-pauzy u mieszkanek niektórych rejonówLGOM, [in:] Człowiek, Populacja, Środowisko,Prace Dolnośląskiego Centrum DiagnostykiMedycznej �DOLMED� we Wrocławiu, IV,125-142

KULIKOWSKI M., I. TOMASZEWSKA, 1995, Okresprzekwitania u kobiet � hormonalna terapiazastępcza, Polski Tyg. Lek., L (44/47), 5-9

LANCASTER J.N., C.S. LANCASTER, 1983, Paren-tal investment: the hominid adaptation, [in:]How Humans adapt: a biocultural odyssy, D.Ortner (ed.), Washington DC: Smithsonian In-stitution Press, pp. 33-66

LEIDY SIEVERT L., D. WADDLE, C. CANALI, 2001,Marital status and age at natural menopause:considering pheromonal influence, Am. J.Hum. Biol., 13, 479-486

LEIDY SIEVERT L., 2001, Menopause as a measu-re of population health. An overview, Am. J.Hum. Biol., 13, 429-434

LINDQUIST O., C. BENGTSSON, 1980, Serumlipids, arterial blood pressure and body weightin relation to the menopause: results from apopulation study of women in Göteborg, Swe-den, Scand. J. Clin. Lab. Invest., 40, 629-636

LOCK M., P. KAUFERT, 2001, Menopause, localbiologies, and cultures of aging, Am. J. Hum.Biol., 43, 494-505

ŁASKA-MIERZEJEWSKA T., H. MILICER, H.PIECHACZEK, 1982, Age at menarche and itssecular trend in urban and rural girls in Po-land, Ann. Hum. Biol., 9, 227-233

MACMAHON B., J. WORCESTER, 1966, Age atmenopause, National Centre for Vital and He-alth Statistics, Ser. 11, no. 19, Washington

MALINA R.M., 1979, Secular changes in size andamturity causes and effects, Mon. Soc. Res.Child Dev., 44, (serial no. 179), 59-120

MARIN M., J.E. BLOCK, S.D. SANCHEZ, C.D.ARNAUD, Y. BEYENE, 1993, Menopause witho-ut symptoms: The endocrinology of menopauseamong rural Mayan Indians, Am. J. Gynaec.,168, 1839-1845

MCKINLEY S.M., M. JEFFERYS, B. THOMPSON,1972, An investigation of the age at menopau-se, J. Biosoc. Sci., 4, 161-173

MILICER H., 1966, The secular trend in growthand maturation as revealed by Polish data, T.Soc. Geneesk., 44, 562-568

MIZIA-STEC K., F. ZYCH, E. WIDALA, 1998,Ocena wybranych parametrów hormonalnychu chorych z mastopatią w okresie pomenopau-zalnym, Gin. Pol., 69, 56-61

PANEK S., 1967, Sekularne przyspieszenie tempawzrastania i rozwoju organizmu człowieka,Roczn. Nauk. WSWS, Kraków, 6, 223-335

PARAZZINI F., E. NEGRI, C. VECCHIA, 1992,Reproductive and general lifestyle determi-nants of age at menopause, Maturitas, 15,141-149

PAVELKA M.S.M., L.M. FEDIGAN, 1991, Meno-pause: a comparative life-history perspective,Yrbk Phys. Anthropol., 34, 13-38

PECCEI J.S., 1995a, A hypothesis for the originand evolution of menopause, Maturitas, 21,83-89

PECCEI J.S., 1995b, The origin and evolution ofmenopause: the altricality-lifespan hypothesis,Ethol. Sociobiol., 16, 425-449

PECCEI J.S., 2001, Grandmother hypotheses: oldand new, Am. J. Hum. Biol., 13, 434-452

PIPLAI C., 1991, Age at menopause of TamangWomen Tea-Labourers of Jalpaiguri District,west Bengal, India, Int. J. Anthropol., 6,233-236

REBATO E., 1988, Ages at Menarche and Meno-pause in Basque Women, Coll. Anthropol., 12,147-149

Report of a WHO Scientific Group, 1981, Rese-arch on the menopause, WHO Technical Re-port Series, 670, Geneva: World Health Orga-nisation

Report of a WHO Scientific Group, 1996, Rese-arch on the menopause in the 1990s, WHOTechnical Report Series, 866, Geneva: WorldHealth Organisation

RODZIEWICZ-GRUHN J., 1998, Genetyczne i śro-dowiskowe uwarunkowania wieku menopauzyi długości okresu rozrodczego, Studies in Hu-man Ecology, 2, 201-210

STANFORD J.L., P. HARTGE, L.A. BRINTON, R.N.HOOVER, R. BROOKMEYER, 1987, Factorsinfluencing the age at natural menopause,J. Chron. Dis., 40, 995-1002

STANOSZ S., B. LISIECKA, T. WESOŁOWSKA,K. GOERTZ, D. KULIGOWSKI, 1993, Wpływestradiolu i estriolu na zachowanie się niektó-rych frakcji lipidowych i estrów cholesteroluw surowicy u kobiet z osteoporozą w okresie

Age at menopause and reproductive determinants in Polish women 71

pomenopauzalnym, Medycyna 2000, 35/36,37-39

STOKOWSKI C., 1968, Wpływ czynników konstytu-cjonalnych na przekwitanie u kobiet. PamiętnikXVII Zjazdu Pol. Tow. Gin., Kraków, 2, 12-15

STOŁYHWOWA E., 1938, Badania nad zróżnico-waniem rasowym długości okresu rozrodczegou kobiet, Przegląd Antropologiczny, 12, 1-22

SZEMIK M., 1984, Age at menarche, duration ofpotential reproductive period and fertility,Studies in Human Ecology, 5, 133-151

SZWED A., 2001, Biologiczne i kulturowe uwa-runkowania klimakterium u kobiet, Doctoralthesis, Adam Mickiewicz University, Poznań

TREOLAR A., 1974, Menarche, Menopause andIntervening Fecundability, Hum. Biol., 46,89-107

TREOLAR A., 1981, Menstrual cyclicity and thepremenopause, Maturitas, 3, 249-364

VANKEEP P.A., P.C. BRAND, P.H. LEHERT, 1979,Factors affecting the age at menopause, J. Bio-soc. Sci., Suppl., 6, 37-55

WALSH R.J., 1978, The age of the menopausein Australian women, Med. J. Aust., 24, 181-215

WARENIK-SZYMANKIEWICZ A., 1996, Klimakte-rium, Klin. Perinatol. Gin., 16, 66-74

WHELAN E., D. SANDLER, R. MCCONNAUGHEY,C. WEINBERG, 1990, Menstrual and reproduc-tive characteristics and age at natural meno-pause, Am. J. Epidem., 131, 625-632

WILLIAMS G.C., 1957, Pleiotropy, natural selec-tion and the evolution of senescence, Evolu-tion, 11, 398-411

StreszczeniePorównanie tempa zmian zdolności rozrodczych kobiet z tempem starzenia się innych cech

fizjologicznych, w tym możliwości reprodukcyjnych mężczyzn, ukazuje gwałtowny spadekzdolności rozrodczych aż do całkowitego zakończenia manifestującego się ostatnią mie-siączką w życiu kobiety � menopauzą (rys. 1). Zakończenie okresu reprodukcyjnego u kobiet(menopauza) znacznie przed końcem życia jest zjawiskiem unikatowym, nie spotykanymu innych ssaków żyjących w warunkach naturalnych [HILL i HURTADO 1991; PAVELKAi FEDIGAN 1991; AUSTAD 1994]. Teorie dotyczące pochodzenia i utrwalenia się menopauzyklasyfikowane są jako adaptacyjne (hipoteza matki oraz hipoteza babki) lub nieadaptacyjne.Menopauza kończy okres zdolności reprodukcyjnych kobiet, zatem w pracy podejmuje siępróbę określenia, czy i w jakim stopniu historia cyklu płciowego oraz aktywność rozrodczakobiet wpływają na indywidualne zróżnicowanie wieku wystąpienia menopauzy u kobiet.

W pracy analizowano dane ankietowe i antropometryczne 4354 kobiet z Wielkopolskii zachodniej Polski, w wieku 35-65 lat, zebrane podczas badań przekrojowych w latach1998-2001, w ramach projektu �Biologiczne aspekty starzenia się człowieka�. Lokalizacjęmiejsc badań przedstawia rys. 2. Strukturę demograficzną i społeczno-ekonomiczną bada-nych kobiet ilustrują rys. 3-4. Zgodnie z wytycznymi WHO [Report... 1981, 1996] wiekmenopauzy określono jako wiek ostatniej miesiączki w życiu kobiety, po której następuje 12miesięczny okres amenorrhea. Wśród badanych kobiet wydzielono grupę kobiet, które sto-sowały hormonalną terapię zastępczą w celu złagodzenia dolegliwości klimakteryjnychw okresie bezpośrednio poprzedzającym menopauzę. Te kobiety, określone jako przecho-dzące menopauzę pod kontrolą hormonalną, wydzielono z analiz statystycznych. Kobietyrozdzielono według statusu menopauzalnego na następujące grupy: będące przed menopau-zą, w trakcie menopauzy i po menopauzie. Status kobiet miesiączkujących w okresie po-przedzającym badanie i w ciągu 3 ostatnich miesięcy określono jako przedmenopauzalny.Stanowiły one grupę liczącą 2410 kobiet (55,3% wszystkich badanych). 331 kobiet zanoto-wało ostatnią w życiu miesiączkę ale od tego momentu do chwili badania nie upłynęło 12miesięcy. Status tych kobiet określono jako menopauzalny. Status 1300 kobiet określono jakpo-menopauzalny, z czego 1033 kobiety przeszły menopauzę naturalną.

Maria Kaczmarek, Anita Szwed72

Wyliczono przeciętny wiek menopauzy posługując się metodą retrospektywną i status quo.Wyniki statystyki opisowej oraz transformacji probitowej przedstawiono w tabeli 1. Nieza-leżnie od stosowanej metodyki, wiek menopauzy różnił się istotnie w zależności od charakte-ru menopauzy. Przeciętny wiek menopauzy naturalnej, oszacowany za pomocą analizy pro-bitowej, wynosi 50,1 lat.

Wyniki analizy wariancji jednokierunkowej, zawarte w tabeli 2 oraz na rys. 5 wskazują, iżwiek menopauzy jest istotnie zróżnicowany ze względu na wiek menarche oraz długośćcyklu płciowego. Kobiety wcześniej dojrzewające i mające krótszy cykl płciowy przechodząmenopauzę wcześniej w porównaniu z kobietami, u których menarche wystąpiła późno a cyklpłciowy trwał dłużej. Wyniki analizy regresji krokowej z eliminacją wsteczną (tabela 3)pozwoliły wskazać zmienne, które wyjaśniają 3,96% całkowitej zmienności wieku meno-pauzy i wzajemnie współdziałając jako kompleksowy czynnik są ważne w analizie tegozjawiska. Są to: długość cyklu miesiączkowego, wiek menarche, wiek pierwszej i ostatniejciąży oraz długość okresów laktacji.

Dane porównawcze na temat przeciętnego wieku menopauzy kobiet polskich zestawionow tabeli 4. Wskazują one iż przeciętny wiek menopauzy, w przeciwieństwie do wieku me-narche, nie wykazuje trendu sekularnego. Źródeł niewielkich różnic w przeciętnym wiekumenopauzy należy upatrywać w zastosowanych metodach badań.