Lounana Joseph1, Medelli Jean1, Sozina Katuli2, Chukwueke Ihuoma3, W Lawrence Beeson3,4, Zaida Cordero-MacIntyre3,4
1Exercise Biology and Sport Medicine Unit, University-Hospital-Centre, 80054 AMIENS Cedex 1, France.
2Andrews University. Department of Physical Therapy, Berrien Springs, MI
3Center for Nutrition, Healthy Lifestyle and Disease Prevention, School of Public Health, Loma Linda University, Loma Linda, CA
4Center for Health Disparities and Molecular Medicine, School of Medicine, Loma Linda University, Loma Linda, CA
Objective: To assess the effects of family structure and physical activity on late menarche. Research Methodsand and Procedures: Retrospectivelyagesatmenarche, familysizes, birthordersandothervariables were studied in 238 Congolese women including 90 pre-menarche-trained athletes, 18 post-menarche- trained athletes and 130non-athletes. Results: Mean age at menarche was later (p<0.001) in pre-menarche (15.36 + 1.04 yrs) compared to post- menarche athletes (13.93 + 1.03 yrs) or to non-athletes (14.03 + 1.22 yrs), but similar (p>0.4) betweenthese two last groups. Mean birth order or birth order within a family size was higher (p<0.01) in athletes than in non-athletes. Mean family size or sex ratio did not differ between groups. Controlling for the effect of duration of training before menarche abolished family size or birth order slight effect (p<0.05) on age at menarche, while the duration of training substantial effect (p<0.001) on age at menarche persisted after controlling for family size and birth order effects. Regression analysis confirmed that duration of training was a better predictor of age at menarche than family size or birth order in pre-menarche athletes.
Conclusion: Physical activity before puberty seems to be a more important determinant of late menarche than family structure.
Key words: puberty, athletes, Africa, family size, populations.
Compared to non-athletes and the general population, female athletes present a later menarche (Malina et al.,1979; Frisch et al., 1980; Sudhi and Grewal, 1980; Mathur and Toriola 1982; Mesaki et al., 1984; Sharma and Shukla, 1992; Baxterjones et al., 1994, Lounana et al., 2002). Late age at menarche in athletes is generally associated with bone age retardation (Warren, 1980, Weimann et al., 2000) and delay in pubertal elevationofthe luteinizing hormone : follicle-stimulating hormone (LH:FSH) ratio (Weimannetal., 2000). Several factors relatedtogenetics, physical training, body composition, hormonal profile, nutritional or socio- economic status, geographical or socio- cultural environment as well asfamilyand psychosocial environment have been implicated in the onset of menarche. Studies of mother/daughter or sister/sister pairs has shown significant correlation with the onset of menarche (Stager et al., 1984; Brooksgun and Warren, 1988; Malina etal.,1994). Regression analysis indicatedanassociation between age at menarche and family size, birth order (Malina et al., 1997)orvarious conditions of social environment ( Lounana et al., 2002). The explanation of the late menarche among female athletes remains unclear for some authors (Malina, 1983) while other authors (Weimann et al, 2000) implicate intensive physical training, stress due to training and competition, weight loss leading to a low fat mass, hormonal disorders, and high-energy expenditure coupled with inadequatedietary intake.
To date few studies haveseparately evaluated the effects of training and family structure (Mota and Silva, 1999, Bagley et al., 2006). In addition, most of the population studies were Caucasian (European or American), showing exclusive family and specific socio- cultural characteristics (Wronka and Pawlińska-Chmara, 2005, Anh et al. 2013, Mota and Silva 1999). Our study focused at examining various determinants of age at menarche in a Black African population from Congo-Brazzaville and to evaluate the effects of factors linked to the presence or absence of a regular physical activity and its duration before menarche, simultaneously with the family structure or composition.
2. Materials and methods
Participants. This survey was carried out in May/June 1998 with 238 premenopausal African women (mean age 23.9 ± 4.5 years) from Congo-Brazzaville. All participants were students or young workers whose birth, childhood and puberty occurred during a relatively peaceful period of the country’s history, i.e. without war, starvation and migrations that could influence the onset of menarche (Prebeg and Bralic, 2000). The sample included 90 female athletes who started their physical training before menarche (pre-menarche athletes), 18 athletes who started their physical training after menarche (post-menarche athletes) and 130 subjects having no regular physical activity or athletic training (non-athletes).
The recruited 108 athletes came from six sports: 50% from handball (HB), 18% from running (RUN), 15% from soccer (FB), 8% from volleyball (VB), 7% from basketball (BB) and 2% from gymnastics For eighty percent of the individuals (HB, BB, FB and VB), training often included warm-up exercises, endurance and power exercises, interval training, jumps, stretching and specific resistance exercises. However, most of the training time was dedicated to specific exercises (i.e. ball control, dribbling, passing, shooting, defence, placement, etc.) and above all to a match lasting about 2 hours per training session. All the RUN athleteswere middle-distance or distance runners. All athletes trained about 20 to 30 hours per week with at least five out of seven days of training. The gymnasts complied with Olympic gymnastics program.
The study was undertaken with the personal consent of all the participants. This survey was approved bythe administrative authorities of the “Permanent Research Institutions” of the “HigherInstituteof Physical Education and Sport” (ISEPS) and under the care of Marien Ngouabi University (UNMG), of the Ministry ofEducationand the Ministry of Youth and SportsofCongo- Brazzaville.
Inclusion criteria. Participants hadtobe: “Bantu” ethnic or cultural origin, with childhood and teenage years in Brazzaville or its surroundings, from a non-broken monogamous family structure notably with the father at home, and satisfying physical and mental health criteria. To avoid a possible bias in the counting ofnon-pubescentgirls, the individuals were selected at the date of the study with a civil age of over 18.5 years, which is the maximum age at menarche generally observed among the Congolese population (Louisiana et al., 2002). Other criteria were the possession of a sport licence, an intensive training of 20 to 30 hours per week with at least 5 out of 7 days of training, a high level of national or international competition certified by an individual’s list of achievements or by a competition program of a sport club. To avoid any bias linked to a possible concomitance between date of entry into sport and date at menarche, and to increase the separation between groups of participants, we selected only the pre- menarche athletes who started training atleast
1.0 year before menarche and the post- menarche athletes who started training at least
1.0 year after menarche. Non-athletes were those with total absence of regularphysical training of any nature and duration.
Exclusion criteria. Not satisfying the inclusion requirements, detection of a serious pathological, nutritional or familial medical history, inability to memorize dates or miscellaneous events of family and personal life, use of contraceptives, belonging to a polygamous family or to an ethnic group practising traditional rites of puberty initiation and having less than 1.0 year of training from the date of menarche.
Variables and measurements. Retrospective method with direct interviews based on a pre- tested questionnaire with an estimate of 90 % of exactitude of recall (Lounana et al. 2002). Each participant was requested to give information as precise as possible concerning: date of birth, date of menarche, date of beginning of regular sport training, number of children in the family (family size), number of boys and girls in the family, birth order among siblings, family’ standard of living and participant’ school level. The sex-ratio was calculated in each family as number of boys divided by number of girls.
The family size was examined either as a continuous variable or was divided into five classic groups: size A (1-2 children), size B (3-4 children), size C (5-6 children), size D (7-8 children) and size E (over 8 children). The subject’s school level was taken into account to minimize a possible bias in the ability of understanding questions or memorizing dates and events. Six stages described this parameter: school non- attendance (stage 1), half and end of primary school (stages 2 and 3), half and end of secondary school (stages 4 and 5), at least one year at university (stage 6). To control any bias related to the social status, the family’s standard of living was especially focused on the childhood and teenage years of the individuals and then assessed by a classification drawn from the Warner socio- metric scale (DeLandsheere, 1972). This classification was based on the examination of parameters obtained during the interview (father’s and mother’s occupations, type and location of dwelling place, living comforts, type of school, means of travel, neighbourhood). These data were classified into six levels as; the relatively rich and prosperous class (levels 5 and 6), the middle class (levels 3 and 4) and the lower class (levels 1 and 2). Civilian age, age at menarche and age at beginning of regular sport training were calculated from the reported dates. An exact date of menarche was reported by 59% of the participants. For 30% of the participants in whom the date was not precise within a quarter, the middle of the period was considered as the annual fraction. In the particular case whose age at menarche was only reported between two consecutive birthday dates (11% of the individuals), it was put at the date of the 1st birthday plus 0.5 years (Livson, 1962).Individual,free, recreational sport is not popular yetamong
women of Congo-Brazzaville. Though the recruiting criteria to clubs are not known, it is observed that the first selection of a sport is generally definitive. “Age at beginning of regular sport participation” means“ageat which formal and specialized athletic training is initiated”. The reported datesatbeginning of regular sport training wereeasilychecked against sport licences and clubs registrations. These dates were obtained with 100% accuracy in all athletic participants (athletes without sport licence have been excluded as indicated above). For analysis purpose, the “duration of training before menarche” variable included values observed in pre- menarche athletes whereas a value of zero was attributed to the post-menarche athletes and the non-athletes.
Statistical analysis. Analysis of variance (ANOVA) was used for thecomparisonof more than two group means after testing for normality assumptions using Kolmogorov- Smirnov test and Wald-Wolfowitz test respectively. Binary comparisons were made using the Games-Howel post-hoc test. Multiple andbinarynon-parametric comparisons of the means ofthenon-paired groups were performed with the Kruskall- Wallis H and Mann-Whitney U tests. To study the effect of the variables, we carried out simple or multipleregressionand covariance (ANCOVA) analyses based on the General Linear Modelling procedures, as well as an analysis of logistic regression. Correlations were evaluated using a Pearson classical or a Spearman non-parametric test. All analyses were carried out with statistical analysis software packages (SPSS for Windows, v23, SPSS Inc., and STATVIEW 5, SAS Institute Inc. USA). Statistical significance was set at p < 0.05.
Figure 1 shows the age at menarche distributions. Table 1 summarizes the subsamples characteristics. The 3 groups of subjects did not differ for any of the variables studied except for age at menarche, birth order and birth order within a family size. The mean birth order or birth order within a family size was lower (p < 0.01) for the non- athletes compared to the two groups of athletes who were statistically similar on this point. The mean age at menarche reported by the 90 pre-menarche athletes (15.36 + 1.04 years) was significantly later (p<0.001) compared to the 18 post-menarche athletes (13.93 + 1.03 years) or the 130 non-athletes (14.03 + 1.22 years), while no difference (p > 0.4) was observed between these two last groups (Table 1).
Using an ANCOVA model and introducing duration of training variable significantly lowered the difference between groups for age at menarche (F=1.98 vs. 38.7) at a non- significant level (p > 0.1 vs. p < 0.001), whereas no effect was observed by introducing any other variable in the model. In pre-menarche athletes, there was no relation between duration of training before menarche and age at beginning of regular sport training (p > 0.9). In addition, the duration of training significant effect on age at menarche did not change even if duration of training was adjusted for age at beginning of regular sport training.
Tables 2 and 3 summarize the correlation analysis. Age at menarche was strongly (p<0.001) correlated with duration of training in pre-menarche athletes (r = 0.38). Birth order was significantly correlated with family size in pre-menarche athletes (r = 0.53), post- menarche athletes (r = 0.85) and non-athletes (r = 0.47).
Table 4 summarizes the results of linear regression analysis for the prediction of age at menarche. The main effect of duration of training on age at menarche wassignificant (p<0.001) in the pre-menarche athletes (slope β = 0.24) and in the total sample (β = 0.38). The coefficient of 0.38, which takes into account the 0 value of duration of training in the general population, indicates that menarche is late by 4.56 months for each year of training before the menarche. The individual regression effect of family size or birth order on age at menarche was not significant. However, there was a slight but significant effect (p<0.05) of family size when controlling statistically for birth order (β = -0.08), and also a slight effect of birth order when controlling for family size (b = 0.10) with similar magnitudes in pre- menarche athletes and in the total sample. There was a significant effect (p<0.001) of duration of training on age at menarche when controlling for family size or/and birth order with similar magnitudes in the pre-menarche athletes (b = 0.27) as well as in the total sample (b = 0.38).
In the population studied, the upper limit of variation including 95% of confidence interval for the average age at menarche was calculated as 14.73 years. Table 5 shows the subjects distribution according to this threshold value: 73% of pre-menarche athletes, 39% of post-menarche athletes and 35% of non-athletes had a late menarche (>14.73 years). The logistic regression analysis indicated that “presence” (but not “absence”) of training before menarche is linked to a higher risk of late menarche (Odds-ratio = 0.228; p<0.001).
Menarche is an event that is generally well remembered by women. Its recall precision frequently goes beyond 80% in developed countries (Koprowski et al., 2001; Must et al., 2002). It remains unproven if a similar level of recall can be observed in less developed countries. We therefore recognize that there is an important limit in the retrospective nature of the information used in this study. However, several potential biasfactorswere taken into consideration in the analysisofthe results (tables 2 and 3) as well as in the methodological approach.
Variability of age at menarche in African Black populations
Black Africa is constituted of about forty countries still subjected to several calamities (famine, war, poverty, malnutrition) as well as endemic diseases such as malaria or sickle cell illness, which are possible factors interfering with the normal growth and maturation, and then with the occurring of the menarche in female teenagers.
The menarcheal age in Black populations of Africa has been studied in relatively large samples (see Table 6), which unfortunately have been seldom exploited for a simultaneous analysis of determinants. Compared to Western Caucasian populations (Frisch et al., 1980; Baxterjones et al., 1994; Malina et al., 1994; Malina et al., 1997), the mean menarcheal age in African Black populations generally appears higher by 1.5to
2 years. This difference is linked to the conditions of life and nutrition, and to the socio-economical and cultural status of the populations (Morabia et al., 1998; Thomas et al., 2001). In most of Black African countries, the menarche occurs at a mean age varying between 13 and 16 years. The lowest values of this range are generally found in countries that are relatively more industrialized (i.e. South Africa, Nigeria), while higher values are found in less developed countries (see table 6). A trend towards lower values has been described in some countries but this trend seems to be totally undetectable in the poorest countries (Padez, 2003).
In African populations the mean menarchal age is higher for rural or suburban compared to urban communities (+0.6 to +1.5 years) (Pasquet et al. 1999; Pawloski 2002; Oduntan et al. 1976), for the poorest socio-economic strata compared to richest (~ +0.8 years) (Attallah, et al. 1983), for certain ethnic groups (Adadevoh et al., 1989), especially for Black compared to White South Africans (~
+1.0 year) (Henneberg and Luow,1995; Cameron and Getz, 1997), and for girls suffering from homozygous sickle cell disease compared to controls (~ +1.2 year) (Modebe, 1987; Frisch et al. 1981; M’Pemba-Loufoua et al. 2001). The relation between physical activity and menarche has not been extensively studied in Sub-Saharan Africa. The mean menarcheal age value observed in the current study in the non-athletes population of Congo-Brazzaville (14.0 ± 1.2 years) is similar to a previously reported value (13.92 ± 1.4 years) from the same population Lounana et al., 2002). It is also comparable to data of general populations from Black African countries of similar living standards (see table 6).
Late menarche in athletes and non athletes Our first finding is that a later menarche commonly characterise female athletes in comparison with non-athletes. This phenomenon is particularly observed in female athletes who started training before puberty, while female athletes who started training after puberty exhibit a pubertal chronology identical to the non-athletes or the general population. This result is in total agreement with many other studies (Malina et al., 1979; Frisch et al., 1980; Mesaki et al., 1984; Sharma and Shukla, 1992; Baxterjones et al., 1994; Lounana et al., 2002; Frisch et al., 1981; Vadocz et al., 2002). Since it is not known whether menarche is delayed or just late the term “late menarche” is suitable to use (Stager et al, 1984. This pubertal dysfunction has been noted in individual sports as well as in team sports. Female swimmers are also vulnerable to a later puberty associated with exercise training (Baxterjones et al., 1994; Frisch 1981; Constantini and Warren, 1995; Astrand et al, 1963).
In the Congolese population, each group is affected by cases of late menarche (>14.73 years). We can consider that the general population normally include about 1/3 of cases of late menarche (table 5). It is also suggested that they could be a result of combination of these two factors since the effect of thinness on late menarche doesn’t appear alone, only together with the effect of intensive physical activity (Vandenbrouke, 1982).
The effect of selection would be associated with the nature of sport (Malina 1983). Girls with similar morphology and physical abilities have a tendency to converge to the same type of sport. In some sports like gymnastics or ballet dance, recruitment at an early age could favor girls with physical characteristics, such as thinness or low body mass and fat, associated with late menarche (Beunen et al., 1994). Attempting to minimize an effect possibly due to the nature of sport on age at menarche, our samples of athletes were taken from sports whose initial recruitment methods were well defined and not suspect of bias (Lounana et al., 2002; Lounana et al., 2000). Our results did not show any difference for the mean age at menarche according to the nature of sport (p>0.17).
Effect of duration of training before menarche In the current study, the effect of training duration on age at menarche appeared to be unaffected by the effect of age at beginning of regular sport training. Notably, the estimation of a 4.56 months late menarche for each training year before menarche is in agreement with the 5 months late previously reported both for Congolese population (Lounana, 2002) and American population (Frisch et al., 1981). In addition, we noted that 1/3 of the pre-menarche athletes who did not have a late menarche (>14.73 years) displayed a mean training time significantly lower compared to the 2/3 of the individuals whose menarche was late (2.3 + 0.93 vs. 3.5 + 1.8 years; p<0.01). This is in agreement with another study (Mesaki et al, 1984) which found that intensive physical activity (and not moderate) is commonly associated with a late menarche in girls who started training before puberty. These results suggest a dose-response characteristic of the training effect associated with late menarche.
Hard physical exercise and high level of competition result in training stress that is generally associated with changes in fat/leanness ratio and disturbances in metabolic and hormonal levels (Weimann et al., 2000; Toriola, 1988; Loucks, 1990). In female gymnasts, a close relationship has been found between the “entry age” into high intensity training and alterations in peri- pubertal development, such as a delay of LH:FSH ratio of 2.3 years, a bone age retardation of 2.5 years, as well as a lateness in age at menarche of 2.5 years (Weimann et al., 2000).
Birth order and family size values
In the Congolese as well as in the American population (Malina et al., 1997) women atheletes exhibit a higher average birth order compared to non-athletes. However in the Congolese population, the average birth order does not differ between pre-menarche and post-menarche athletes. Furthermore, birth order is positively correlated to duration of training (table 4). Thus in the Congolese population, birth order would appear as a selection factor for sports activity independent of puberty maturation. The higher birth order among siblings might increase chances of a girl to be more independent and free in the family, since a particularly important barrier against physical activity for women is time constraints (Tappe et al., 1989). Time dedicated to family chores by young African girls seems to be significant.
The mean values of family size in Congo- Brazzaville population are inagreementwith the official statistics between 1965 and 1998 (United Nations, 2002). Congolese population and Western reference populations display, for family size, several pointsofdivergence which could explain the differencesseenin the effect of this variable on age at menarche. We observed higher family size among Congolese population comparedtoEuropean or American populations. Family size values were identical between athletes and non- athletes in Congolese population.
Family size and birth order effects on age at menarche
The effect on age at menarche of birth order adjusted for family size appeared relatively null (b = 0.01 to 0.04) in the American non- athlete university students (Malina et al., 1997; Wellens et al., 1992). This is also observed in the Congolese non-athletes and Post-menarche athletes (b = -0.02 to 0.02; tables 1 and 4). On the contrary in athletes, this effect is significant and almost of similar (Pillai, 1995) magnitude among the different populations. However, in Congolese Pre- menarche athletes this effect is positive (b = 0.10), whereas it is mostly negative (b = -0.07 to –0.19) in English university athletes (Malina et al., 1979) or American athletes (Malina et al., 1997).
There was a weaker inverse relationship between family size and age at menarche adjusted for birth order in the Congolese pre- menarche athletes and the total sample (b = -0.05 to –0.08) compared to American university student and Olympic athletes (b = 0.16 to 0.22; (Malina et al. 1997; Wellens, 1992). The family size effect observed more often is generally in a positive direction, and its magnitude turns out to be weaker for the non-athletes (b = 0.08 to 0.15) than for athletes (b = 0.12 to 0.22) from America, Belgium (Malina et al., 1994; Wellens et al., 1992), Romania (Stukovsk et al., 1967), Italy (Gallo, 1977) or United Kingdom (Malina et al., 1979).
How family size and birth order may generate positive or negative effects on age at menarche remain to be elucidated. One typical hypothesis of the probable mechanism of the positive effect of family size leading to a late menarche is based on nutrition, health, and the assumption that the more the family size increases, the larger the number of mouths to feed, and thus the higher the risk of disease and malnutrition possibly leading a girl to remain below the physique necessary for the onset of menarche. Another hypothesis (Malina et al., 1997) is that the effect of increase in family size can show-up earlier, related to the frequency of births, the reduction and repetition of which would have unfavourable cumulative effects on the mother/fetus relationship as well as on the conditions of fetal development. These processes can affect the initial programming of maturation during the sensitive period of fetal development.
The inverse relation of family size and age at menarche could be linked to specificities of the Congolese population family structure and socio-cultural context and could be considered as an indirect negative effect generated in the family life by an increase in family size associated with an amplification of family stress factors, such as family conflicts or unstable home environments, which are capable of generating an early menarche (Graber et al, 1995; Malina, 1983).
Duration of training and family size - birth order relationships
Duration of training appears to be a confounding factor for the effects of family size and birth order on ageatmenarche (tables 2 and 4). However,whencontrolled for duration of training, the effect of family size or birth order on age atmenarchewas removed (table 4, figure 2), while the duration of training effect on age at menarche persisted after controlling for family size and birth order effects (table 4). Furthermore, in agreement with other reports (Malina et al, 1997), this study reveals that the effects of family size and birth order,oncemutually adjusted, are systematically present in athletes but neither in non-athletes nor in the general population. These facts suggest that family size and birth order effects would be associated with sports activity, whichseems to be itself a prime factor.
The maximum level of age at menarche variability predicted by duration of training effect (R² = 0.25) is around 8 times higher than the negligible one by family size or birth order (R² = 0.03). A longitudinal survey of genetic behaviour carried out with 1,338 American parental couples (Doughty and Rodgers, 2000) indicated that around half of the variability of the age at menarche could be attributed to genetic factors (R² = 0.54) and almost half to non-shared environmental influences. However in this study the 25 % rate of prediction due to duration of training would account for around half of the age at menarche variability predictedby environmental factors. Thus the effect of physical activity before menarche appears to be significant.
The relationship observed between duration of training and family size-birth order complex is a new finding. This relation is consistent with the hypothesis that freedom of action allowing a girl to escape more often from family chores for personal activities could increase as the family size increases, the elder daughter belonging to a larger family having possibly the highest freedom.
Due to limitations during enrolment the study ended up with uneven sample size for the two groups compared. This may pose some problems as these may end up with unequal variance which violates some of the assumptions of the tests used, particularly the regression analysis. However for the t-test the uneven variance is accounted for in the SPSS since there is an option for equal variance not assumed.
Conclusion. This study showed that, in Congolese female athletes, late age at menarche is associated with intensive physical activity initiated before puberty. The effects of family size and birth order on age at menarche appear to be unsteady, unlike the clearly apparent effect of duration of physical activity. Family size and birth order are not good predictors of the menarche in the Congolese population. Surprisingly, these two variables appear to be associated with the sports activity variables. It is suggested that the effect of physical activity should be taken into account when examining the relative contribution of familial and environmental factors implicated in the late menarche of female athletes.
The authors would like to thank all the Congolese women who kindly participated in this study. Special thanks for help in data collection go to Lambert Bakala, Bernard Batambika, Pierre Bazolo and François Mbemba, lecturers at ISEPS (Marien Ngouabi University of Congo–Brazzaville), Frederic Campion for his assistance and Emmanuel Ojo for editing.
• Abioye-Kuteyi EA, Ojofeitimi EO, Aina OI, Kio F, Aluko Y, Mosuro O. The influence of socioeconomic and nutritional status on menarche in Nigerian school girls. Nutr Health. 1997;11:185-195.
• Adadevoh SW, Agble TK, Hobbs C, Elkins TE. Menarcheal age in Ghanaian school girls. Int J Gynaecol Obstet. 1989;30:63-68.
• Astrand PO, Eriksson BO, Karlberg P, Saltin B, Nylander I, Engstrom L, et al. Girl Swimmers - with Special Reference to Respiratory and Circulatory Adaptation and Gynaecological and and Psychiatric Aspects. Acta Paediatrica. 1963;Suppl 147 :1-75.
• Attallah NL, Matta WM, Elmankoushi M. Age at Menarche of Schoolgirls in Khartoum. Annals of Human Biology. 1983;10:185-188.
• Ahn JH, Lim SW, Song BS, Seo J, Lee JA, Kim DH, Lim JS. Age at menarche in the Korean female: secular trends and relationship to adulthood body mass index. Ann Pediatr Endocrinol Metab. 2013 18:60-4.
• Bagley S, Salmon, J. Crawford D, Family structure and children's television viewing and physical activity. Medicine and Science in Sports and Exercise. 2006; 38:910-918.
• Baxterjones ADG, Helms P, Bainespreece J, Preece M. Menarche in Intensively Trained Gymnasts, Swimmers and Tennis Players. Annals of Human Biology. 1994;21:407-415.
• Beunen GP, Malina RM, Lefevre JA, Claessens AL, Renson R, Vanreusel B. Adiposity and Biological Maturity in Girls 6-16 Years of Age. International Journal of Obesity. 1994;18:542-546.
• Brooksgunn J, Warren MP. Mother Daughter Differences in Menarcheal Age in Adolescent Girls Attending National Dance-Company Schools and Non- Dancers. Annals of Human Biology. 1988;15:35-44.
• Cameron N, Getz B. Sex differences in the prevalence of obesity in rural African adolescents. International Journal of Obesity. 1997;21:775-1782.
• Constantini NW, Warren MP. Menstrual Dysfunction in Swimmers - a Distinct Entity. Journal of Clinical Endocrinology & Metabolism. 1995;80:2740-2744.
• DeLandsheere G. Introduction a la recherche en education.: Armand Colin- Bourrelier, Paris; 1972.
• Doughty D, Rodgers JL. Behavior genetic modeling of menarche in U.S. females. In : Genetic influences on human fertility and sexuality. Boston: Kluwer Academic Publishers, 2000.
• Frisch RE, Gotzwelbergen AV, Mcarthur JW, Albright T, Witschi J, Bullen B, et al. Delayed Menarche and Amenorrhea of College Athletes in Relation to Age of Onset of Training. Journal of the American Medical Association. 1981;246:1559-1563.
• Frisch RE, Wyshak G, Vincent L. Delayed Menarche and Amenorrhea in Ballet Dancers - Reply. New England Journal of Medicine. 1980;303:1125-1126.
• Gallo PG. Age at Menarche in Somalia. Annals of Human Biology. 1975;197-200.
• Gallo PG. Age at Menarche in Some Populations of Veneto, North Italy. Annals of Human Biology. 1977;4:179-81.
• Garnier D, Ndiaye G, Benefice E. [Influence of urban migration on physical activity, nutritional status and growth of Senegalese adolescents of rural origin]. Bull Soc Pathol Exot. 2003;96:223-227.
• Gillett-Netting R, Meloy M, Campbell BC. Catch-up reproductive maturation in rural Tonga girls, Zambia? Am J Hum Biol. 2004;16:658-669.
• Graber JA, Brooks-Gunn J, Warren MP. The antecedents of menarcheal age: heredity, family environment, and stressful life events. Child Dev. 1995;66:346-359.
• Henneberg M, Louw GJ. Average menarcheal age of higher socioeconomic- status urban Cape colored girls assessed by means of status-quo and recall methods. Amer J Physical Anthro. 1995;96:1-5.
• Ihekwaba FN. Benign breast disease in Nigerian women: a study of 657 patients. J R Coll Surg Edinb. 1994;39:280-283.
• Koprowski C, Coates RJ, Bernstein L. Ability of young women to recall past body size and age at menarche. Obesity Research. 2001;9:478-485.
• Leenstra T, Petersen LT, Kariuki SK, Oloo AJ, Kager PA, ter Kuile FO. Prevalence and severity of malnutrition and age at menarche; cross-sectional studies in adolescent schoolgirls in western Kenya. European Journal of Clinical Nutrition. 2005;59:41-48.
• Livson N, Mc ND. The accuracy of recalled age of menarche. Hum Biol. 1962;34:218-221.
• Loucks AB. Effects of exercise training on the menstrual cycle: existence and mechanisms. Med Sci Sports Exerc. 1990;22:275-80.
• Lounana J, Bantsimba M, Silou J, Packa-Tchissambou B, Medelli J. Effect of intensive and precocious sport activity on the age at menarche in Congolese women population. Science & Sports. 2002;17:17- 25.
• Lounana J, Medelli J. [The child and competitive sports]. Soins Pediatr Pueric. 2000;195:28-31.
• M'Pemba-Loufoua AB, Nzingoula S, Moubouh-Akouala F, Oba A. [Pubertal development in girls with homozygote sickle cell disease. Apropos of 72 cases]. Bull Soc Pathol Exot. 2001;94:326-329.
• Malina RM, Bouchard C, Shoup RF, Demirjian A, Lariviere G. Age at Menarche, Family-Size, and Birth-Order in Athletes at the Montreal-Olympic-Games, 1976. Medicine and Science in Sports and Exercise. 1979;11:354-358.
• Malina RM, Katzmarzyk PT, Bonci CM, Ryan RC, Wellens RE. Family size and age at menarche in athletes. Medicine and Science in Sports and Exercise. 1997;29:99-106.
• Malina RM, Ryan RC, Bonci CM. Age at Menarche in Athletes and Their Mothers and Sisters. Annals of Human Biology. 1994;21:417-422.
• Malina RM. Menarche in Athletes - a Synthesis and Hypothesis. Annals of Human Biology. 1983;10:1-24.
• Mathur DN, Toriola AL. Age at menarche in Nigerian athletes. Br J Sports Med. 1982;16:250-252.
• Mbizvo MT, Kasule J, Gupta V, Rusakaniko S, Gumbo J, Kinoti SN, et al. Reproductive biology knowledge, and behaviour of teenagers in East, Central and Southern Africa: the Zimbabwe case study. Cent Afr J Med. 1995;41:346-354.
• Mesaki N, Sasaki J, Shoji M, Iwasaki H. [Delayed menarche following early onset of athletic sports training]. Nippon Sanka Fujinka Gakkai Zasshi. 1984;36:49- 56.
• Modebe O. The Effect of Homozygous Sickle-Cell Disease on the Age at Menarche in Nigerian Schoolgirls. Annals of Human Biology. 1987;14(2):181-185.
• Mota J, Silva G. Adolescent's physical activity: association with socio-economic status and parental participation among a portuguese sample,Sport, Education and Society 1999;4:193–199.
• Morabia A, Costanza MC, Neoplasia WHOCS. International variability in ages at menarche, first livebirth, and menopause. American Journal of Epidemiology. 1998;148:1195-1205.
• Must A, Phillips SM, Naumova EN, Blum M, Harris S, Dawson-Hughes B, et al. Recall of early menstrual history and menarcheal body size: after 30 years, how well do women remember? Am J Epidemiol. 2002;155:672-679.
• Nilses C, Lindmark G, Munjanja S, Nystrom L. Trends in fertility patterns of women in rural Zimbabwe. Health Care Women Int. 1997;18:369-382.
• Oduntan SO, Ayeni O, Kale OO. Age of Menarche in Nigerian Girls. Annals of Human Biology. 1976;3:269-274.
• Padez C. Age at menarche of schoolgirls in Maputo, Mozambique. Annals of Human Biology. 2003;30:487- 495.
• Pasquet P, Biyong AMD, Rikong-Adie H, Befidi-Mengue R, Garba MT, Froment A. Age at menarche and urbanization in Cameroon: current status and secular trends. Annals of Human Biology. 1999;26:89-97.
• Pawloski LR. Growth and development of adolescent girls from the Segou Region of Mali (West Africa). American Journal of Physical Anthropology. 2002;117:364- 372.
• Pillai VK. Age at menarche among adolescent females in Zambia: implication for family formation. Int J Sociol Fam. 1995;25:33-38.
• Prebeg Z, Bralic I. Changes in menarcheal age in girls exposed to war conditions. American Journal of Human Biology. 2000;12:503-508.
• Rogo KO, Oniango RK, Muruli LA. Menarche in African Girls in Some Post- Secondary Institutions in Kenya. East African Medical Journal. 1987;64:745-50.
• Sharma SS, Shukla NB. Menarcheal age among Indian sportswomen. Br J Sports Med. 1992;26:129-131.
• Sidhu LS, Grewal R. Age of menarche in various categories of Indian sportswomen. Br J Sports Med. 1980;14:199-203.
• Simondon KB, Simon I, Simondon F. Nutritional status and age at menarche of Senegalese adolescents. Annals of Human Biology. 1997;24:521-32.
• Stager JM, Robertshaw D, Miescher E. Delayed Menarche in Swimmers in Relation to Age at Onset of Training and Athletic Performance. Medicine and Science in Sports and Exercise. 1984;16:550-555.
• Stukovsk R, Valsik JA, Bulaisti M. Family size and menarcheal age in Constanza, Romania. Human Biology. 1967;3:277.
• Tappe MK, Duda JL, Ehrnwald PM. Perceived Barriers to Exercise among Adolescents. Journal of School Health. 1989;59:153-155.
• Thomas F, Renaud F, Benefice E, De Meeus T, Guegan JF. International variability of ages at menarche and menopause: Patterns and main determinants. Human Biology. 2001;73:271-290.
• Thomas KD, Okonofua FE, Chiboka O. A Study of the Menstrual Patterns of Adolescents in Ile-Ife, Nigeria. International Journal of Gynecology & Obstetrics. 1990;33:31-34.
• Toriola AL. Survey of Menstrual Function in Young Nigerian Athletes. International Journal of Sports Medicine. 1988;9:29-34.
• United Nations. Growth rate of population, total fertility rate and sex ratio of Congo Brazzaville. 2002.
• Vadocz EA, Siegel SR, Malina RM. Age at menarche in competitive figure skaters: variation by competency and discipline. Journal of Sports Sciences. 2002;20:93-100.
• Vandenbroucke JP, Vanlaar A, Valkenburg HA. Synergy between Thinness and Intensive Sports Activity in Delaying Menarche. British Medical Journal. 1982;284:1907-1908.
• Wellens R, Malina RM, Roche AF, Chumlea WC, Guo S, Siervogel RM. Body Size and Fatness in Young-Adults in Relation to Age at Menarche. American Journal of Human Biology. 1992;783-787.
• Warren MP. The Effects of Exercise on Pubertal Progression and Reproductive Function in Girls. Journal of Clinical Endocrinology & Metabolism. 1980;51:1150-1157.
• Weimann E, Witzel C, Schwidergall S, Bohles HJ. Peripubertal perturbations in elite gymnasts caused by sport specific training regimes and inadequate nutritional intake. Int J Sports Med. 2000;21:210-215.
• Wronka I, Pawlińska-Chmara, R. Menarcheal age and socio-economic factors in Poland, Annals of Human Biology. 2005;32(5):630-638.