Has Age at Menarche Changed? Results from the National Health and Nutrition Examination Survey (NHANES) 1999–2004

Article Outline

• Abstract
• Methods
  • Sample description
  • Covariates
  • Statistical analysis
• Results
• Discussion
• Acknowledgment
• References
• Copyright

Abstract 

Purpose

To examine self-reported age at menarche in U.S. adults and the associations between age at menarche and race/ethnicity.

Methods

Data from 1999–2004 National Health and Nutrition Examination Survey (NHANES) for 6788 females 20 years and over were analyzed. Self-reported age at first menses (in years) by birth year groups is reported overall and for Mexican Americans, non-Hispanic whites, and non-Hispanic blacks.

Results

Mean age at menarche in the United States declined over time from 13.3 years (95% CI: 13.2–13.5) in the oldest age group, those born prior to 1920, to 12.4 years (95% C.I. 12.2–12.5 years) in the youngest group, born between 1980 and 1984. Declines in age at menarche were observed for all race/ethnicity groups. Non-Hispanic black females had the largest decline in mean age at menarche from 13.6 years (95% CI: 13.1–14.1) in women born prior to 1920, to 12.2 years (95% CI: 11.8–12.6) in the 1980–84 birth cohort. Mean age at menarche among non-Hispanic white females declined from 13.3 years (13.1–13.6) in the pre-1920 birth cohort to 12.5 years (12.3–12.8) in the 1980–84 birth cohort.

Conclusions

Significant declines in the mean age of menarche for U.S. females occurred overall and for all race/ethnic groups examined. Mean age of menarche declined by .9 year overall in women born before 1920 compared to women born in 1980–84; the declines in the mean age at menarche ranged from .7 to 1.4 years depending on the race/ethnicity group.

Keywords: Females, Menarche, NHANES, Race, Ethnicity

The onset of menarche is an important milestone in female sexual maturation. Earlier reports documented declines in the age of menarche in developed countries. Menarche occurred on average at 14 years in the early 1900s and declined to approximately 12.8 years by the late 1940s [1], [2]. Fredriks et al [3] reported declines in median age at menarche in Dutch girls of approximately one month per decade. Recent reports have noted that the declines in the age of onset of menses among U.S. females may be slowing [4], [5]. A decline of approximately 2.5 months in the mean age of menarche was reported for U.S. females between 1963–70 and 1988–1994 [4], and a further decline of approximately 2.4 months between 1988–94 and 1999–2002 [6].

Our interest in examining data on the onset of menarche in the U.S. population stems from a growing literature on health outcomes related to sexual maturation age, demographic changes in the U.S. population, and recent trends in overweight among U.S. children and adolescents. The increasing prevalence of adolescent overweight in the United States [7], [8], [9] appears to coincide with secular trends in pubertal timing [10], [11]. Increased adiposity may trigger estrogen production, leading to earlier onset of puberty in girls [3], [10], [11], [12], [13], [14]. This report examines age and race/ethnicity trends in age at menarche for U.S. adults using data from the 1999–2004 National Health and Nutrition Examination Survey (NHANES).

Back to Article Outline

Methods 

The NHANES is conducted to assess the health and nutritional status of the civilian, non-institutionalized U.S. population [15]. NHANES data are obtained using standardized household interview and health examination methods [15], [16]. NHANES were conducted on a periodic basis between 1971 and 1994. A special study of three Hispanic subgroups living in the United States was conducted during the Hispanic HANES (HHANES), 1982–84; Mexican Americans living in the southwest United States, Cubans in Dade County, Florida, and Puerto Ricans living in the New York City metropolitan area were included. The HHANES data are not nationally representative, however. NHANES became a continuous survey in 1999 and nationally representative samples of Mexican Americans were included in the survey samples. Continuous NHANES public data release files are released in 2-year survey cycles. We combined data from the first three continuous NHANES cycles, 1999–2000, 2001–2002, and 2003–2004 to examine trends in the age at menarche.

Sample description 

For each 2-year survey, nationally representative samples of the civilian, noninstitutionalized U.S. household population were selected using a complex, stratified, multistage probability cluster sampling design [16]. Selected population subgroups were over sampled in NHANES to improve the reliability of statistical estimates derived from the surveys. The over sampled subgroups in NHANES 1999–2004 included Mexican Americans, non-Hispanic blacks, adolescents 12–19 years of age, pregnant women, and persons with low income. Descriptions of the sample design, data collection methods, and protocols are provided elsewhere [16], [17]. Following a home interview, subjects were examined in mobile examination centers (MECs); the reproductive history questionnaire was administered during a private health interview at the MEC. The NCHS Ethics Review Board approved the survey protocols and informed consent was obtained for all subjects.

Covariates 

Year of birth was used to create birth cohort groups for this analysis. The years of birth were grouped into eight birth cohorts. The oldest birth cohort was labeled “before 1920” and the youngest group was labeled “1980–84”; six other birth year cohorts were grouped by decade. The birth cohorts used for this analysis were: before 1920, 1920–29, 1930–39, 1940–49, 1950–59, 1960–69, 1970–79, and 1980–84.

Age at menarche was ascertained for 6788 subjects. The reproductive health questionnaire included a question that is worded as follows: “How old were you when you had your first menstrual period?” Responses to the question were given in whole years. If respondents reported that they did not know the exact age when their periods began, they were asked to specify an age range. The age range choices were: younger than 10 years, 10–12 years, 13–15 years, or 16 years or older. A total of 183 subjects reported a range value for age at menarche. For the analysis, we used the midpoint of the age range for the age at menarche, a value of 9 years if age range was <10 years, and a value of 16 years if the age range 16 years or older was reported. Seventeen respondents did not know their age or age range, and data for one 58-year-old respondent indicated that she had not started her menstrual cycle yet. These eighteen respondents were deleted from the analysis.

Race/ethnicity was self-reported using the recommended federal statistical reporting guidelines for race/ethnicity [18]. The NHANES 1999–2004 data were reported for Mexican Americans, non-Hispanic whites, and non-Hispanic blacks. The results for the “other races” group, including “multiple race” and “other Hispanic” were included in the total sample analyses, but were not reported separately because of insufficient sample sizes and the heterogeneity of the “other races” group.

Statistical analysis 

Statistical analysis was carried out using SAS [19] and SAS-callable SUDAAN [20]. All estimates were weighted using the examination sampling weights provided with the NHANES data files. The sampling weights account for the unequal probabilities of selection, person-level non-response, and a post-stratification adjustment to the estimated U.S. population [16]. Six-year examination sample weights were derived using an approach described in the NHANES analytic guidelines [16].

Means and 95% confidence intervals (CI) by birth cohorts overall and by race/ethnicity category were calculated using SUDAAN (Table 1). Linear regression was used to test for a linear trend in age at menarche with the birth cohorts coded as 1 for before 1920, to 8 for 1980–1984. The F-statistic with Satterthwaite correction for degrees of freedom was used to test statistical significance. A p-value of <.05 was used to define statistical significance. The cumulative percent achieving menarche by age was calculated for each birth-year group.

Table 1. Age at menarche for U.S females 20 years and older by year of birth and race/ethnicity group: United States 1999–2004

Year of birth	Totala		Mexican American		Non-Hispanic White		Non-Hispanic Black
	n	Mean (95% CI)	n	Mean (95% CI)	n	Mean (95% CI)	n	Mean (95% CI)
Before 1920	403	13.3(13.1–13.5)	29	13.2(12.6–13.7)	305	13.3(13.1–13.5)	43	13.6(13.1–14.1)
1920–1929	714	13.2(13.1–13.4)	121	13.2(12.8–13.7)	453	13.1(13.0–13.3)	107	13.4(13.1–13.8)
1930–1939	975	12.8(12.7–13.0)	262	13.0(12.8–13.2)	463	12.8(12.6–12.9)	182	13.1(12.8–13.4)
1940–1949	900	12.7(12.5–12.8)	191	12.5(12.1–12.9)	462	12.6(12.4–12.8)	161	12.9(12.5–13.2)
1950–1959	1075	12.7(12.6–12.8)	248	12.6(12.5–12.8)	490	12.7(12.5–12.8)	246	12.4(12.2–12.7)
1960–1969	1143	12.7(12.6–12.8)	250	12.6(12.4–12.9)	536	12.8(12.6–12.9)	250	12.7(12.4–12.9)
1970–1979	1302	12.5(12.4–12.6)	342	12.4(12.2–12.7)	611	12.6(12.5–12.8)	218	12.2(11.9–12.5)
1980–1984	276	12.4(12.2–12.6)	72	12.2(11.5–12.9)	111	12.5(12.2–12.8)	67	12.2(11.8–12.6)
p-valueb		<.001		.005		<.001		<.001

Back to Article Outline

Results 

Sample sizes, means, and 95% confidence intervals (CI) are shown by birth-year group and race/ethnicity (Table 1). A significant decline in the mean age at onset from 13.3 (CI: 13.2, 13.5) years in the pre-1920 birth cohort to 12.4 years (CI: 12.2, 12.5) in the 1980–84 birth cohort occurred. Significant declines occurred in the mean age at onset for all three race/ethnicity groups as well. Mean ages for the 1980–84 cohort were one year lower for Mexican Americans, .8 years lower for non-Hispanic whites, and 1.4 years lower for non-Hispanic black females. In the 1980–84 cohort, the mean age of onset was the same for non-Hispanic Blacks and Mexican Americans (12.2 years) while the mean age at onset for non-Hispanic whites was 12.5 years. The cumulative percentages of age at menarche for females by birth year group are shown in Figure 1. For simplicity, four of the eight birth-year groups examined in the analysis are shown to illustrate the declines in menarcheal age over time.

• 
  • View Large Image
  • Download to PowerPoint

• Figure 1. 

  Cumulative percentages of age at menarche by birth-year group: U.S. females 20 years and older, 1999–2004.

Back to Article Outline

Discussion 

The NHANES provides cross-sectional “snapshots” of population health characteristics, including reproductive health milestones such as the age of menarche. NHANES estimates reflect changes in the demographic and health characteristics of the U.S. population. This report examined trends in age at menarche in a national probability sample of adults. The analysis revealed statistically significant declines in mean age at menarche overall, and among three race/ethnicity groups. Our findings are consistent with the declines in age at menarche reported by other researchers using different populations and different methods [1], [3], [4], [21], [22], [23]. We compared our findings to results reported by researchers in the United States and Europe who used prospective [21], status quo [4], and retrospective reporting methods [23] to examine trends in the age at menarche.

The prospective inquiry method entails asking questions of non-menstruating samples of girls at regular intervals of time until menstruation occurs. This method was used in the National Heart, Lung, and Blood Institute Growth and Health Study (GHS) reported by Kimm et al [21] to study the development of obesity in black and white girls during adolescence. Girls between the ages of 9 and 10 years of age (comparable to the 1970–79 NHANES birth cohort) were recruited in 1987. The mean age at menarche among white and black GHS participants was 12.7 years and 12.0 years, respectively [21]. The NHANES 1999–2004 mean ages (95% CI) at onset for the 1970–1979 birth cohort were 12.6 years (CI: 12.5, 12.8) and 12.3 years (CI: 12.0, 12.6), for non-Hispanic white and black females, respectively.

The second approach, the status quo method, was used by Anderson et al [4] to report results based on National Health Examination Survey (NHES) Cycles II (1963–65) and III (1966-70) and NHANES III, 1988–94. This approach uses data obtained by querying females of reproductive age to determine whether their first menstrual period occurred and obtaining their exact age. Results for girls between the ages of 10 and 15 years were included in the analytic sample, and cumulative percentages of subjects achieving menarche are computed by year of age. Probit analysis was used to determine the mean age at menarche during both survey periods. Mean age (95% CI) at menarche declined from 12.75 (95% CI: 12.69, 12.82) in NHES to 12.54 years (95% CI: 12.44, 12.64) in NHANES III. The NHANES 1999–2004 total sample mean was 12.7 years for the 1940–49 and 1950–59 birth cohorts, which are the birth cohorts that are closest to the NHES subjects. The NHANES III females 10–15 years would have been born between 1973 and 1984; the mean age at menarche for NHANES 1999–2004 women born between 1970 and 1989 was 12.4–12.5 years (95% CI: 12.2, 12.6), which is also consistent with results reported by Anderson et al [4]. A recent report that was based on a comparison of NHANES III and NHANES 1999–2002 data reported that the mean age at menarche for U.S. girls 9–15 years declined 2.3 months from 12.53 years (95% CI: 12.43, 12.63) to 12.34 years (95% CI: 12.24 to 12.45 years) between 1988–94 and 1999–2002 [6].

Our analysis used a third method that is based on retrospective reporting. Women who are past the menarcheal age are queried so as not to bias mean menarcheal age downward if women who have not begun menstruating are excluded. This method was used in a study with British females aged 20 years and older for the European Prospective Investigation into Cancer and Nutrition (EPIC-Oxford) Study conducted between 1993 and 1999 [23]. The EPIC-Oxford study sample was comprised of women who were born between 1940 and 1970. The mean age at menarche reported for the study was 12.9 years (standard deviation: 1.5 year). The mean menarcheal age for NHANES 1999–2004 non-Hispanic white females in the 1940–1969 birth cohorts ranged from 12.6 to 12.7 years (CI: 12.3, 12.9).

The limitations of our analysis include the use of retrospective information on age at menarche, especially for older women, and the cross-sectional nature of the NHANES data. Overall, 93% of women reported age at menarche in years. Higher percentages of women in the two oldest birth cohorts, however, reported age at menarche using age ranges compared to the younger birth cohorts. The validity of long-term recall of physiological events, including year of menarche, has been assessed by others [24], [25], [26], [27]. A 30-year follow-up of adults who participated in the Newton Girls Study reported that subjects’ recall of age at menarche was .08 years earlier, on average, than their actual age at menarche. Mean absolute error in recalled age at menarche (recalled-original) was .62 years, on average, and varied nonlinearly across menarcheal ages [24]. Results from a longitudinal growth study reported that 84% of women 50 years of age recalled their age at menarche within 1 year of the actual event [28]. Nichols et al examined secular trends in age at menarche, age at menopause, and reproductive life span within a population-based cohort of U.S. women who resided in three U.S. states [29]. In this study, mean age at menarche decreased by 6 months for women born between 1910 and 1949 (13.1 vs. 12.7 years; p < .001), but increased to 13.0 years among women born between 1960 and 1969 (p < .001), a reversal of the downward trend of age at menarche observed within developed countries. Although substudy results revealed that responses to questions about age at menarche and age at menopause were highly reproducible, the authors acknowledged that the reported variation in self-reported age at menarche and menopause increases with increasing time since menopause [29].

Another limitation is that the temporal sequence in the events that may contribute to the observed declines in age at menarche cannot be established with these cross-sectional data. The role of overweight and obesity on menarcheal onset could not be determined because there is no information on NHANES subjects’ body mass index status during adolescence. Other prospective studies have reported that sexual maturation is associated with body size [22], [25], [26], ethnicity [10], [26], [27], and exposure to environmental toxicants including polychlorinated aromatic hydrocarbons (PCAHs) [30], polychlorinated biphenyls (PCBs) and polybrominated biphenyls (PBBs) [31], and dioxins [32]. Additionally, menarcheal age may be an etiologic factor for chronic diseases such as osteoporosis [33], cancer [34], [35], and cardiovascular disease [36]. This analysis used a unique approach to examine menarcheal trends in U.S. adults and provides national data on three race/ethnic groups, including Mexican American women 60 years of age and over.

Back to Article Outline

Acknowledgment 

The findings and conclusions in this report are those of the authors and do not necessarily represent the views of the funding agency.

Back to Article Outline

References 

1. Zacharias L, Wurtman RJ. Age at menarche: Genetic and environmental influences. NEJM. 1969;280:868–875
   • View In Article
2. Wyshak G, Frisch RE. Evidence for secular trend in age of menarche. NEJM. 1982;306:1033–1035
   • View In Article
3. Fredriks AM, van Buuren S, Burgmeijer RJ, et al. Continuing positive secular growth change in the Netherlands 1955–97. Pediatr Res. 2000;47:316–323
   • View In Article
   • MEDLINE
   • CrossRef
4. Anderson SE, Dallal GE, Must A. Relative weight and race influence average age at menarche: Results from two nationally representative surveys of U.S. girls studied 25 years apart. Pediatrics. 2003;111:844–850
   • View In Article
   • CrossRef
5. Chumlea W, Schubert C, Roche A, et al. Age at menarche and racial comparisons. Pediatrics. 2003;111:110–113
   • View In Article
   • CrossRef
6. Anderson SE, Must A. Interpreting the continued decline in the average age at menarche: Results from two nationally representative surveys of U.S. girls studied 10 years apart. J Pediatr. 2005;147:753–760
   • View In Article
   • Abstract
   • Full Text
   • Full-Text PDF (245 KB)
   • CrossRef
7. Troiano RP, Flegal KM. Overweight children and adolescents: Description, epidemiology, and demographics. Pediatrics. 1998;101(3 Pt 2):497–504
   • View In Article
8. Ogden CL, Flegal KM, Carroll MD, et al. Prevalence and trends in overweight among US children and adolescents, 1999–2000. JAMA. 2002;288:1728–1732
   • View In Article
   • MEDLINE
   • CrossRef
9. Hedley AA, Ogden CL, Johnson CL, et al. Prevalence of overweight and obesity among US children, adolescents and adults, 1999–2004. JAMA. 2004;291:2847–2850
   • View In Article
   • CrossRef
10. Wattigney WA, Srinivasan SR, Chen W, et al. Secular trend of earlier onset of menarche with increasing obesity in black and white girls: The Bogalusa Heart Study. Ethn Dis. 1999;9:181–189
    • View In Article
    • MEDLINE
11. Davison KK, Susman EJ, Birch LL. Percent body fat at age 5 predicts earlier pubertal development among girls at age 9. Pediatrics. 2004;111:815–821
    • View In Article
    • CrossRef
12. Garn SM, LaVelle M, Pilkington JJ. Comparisons of fatness in premenarcheal and postmenarcheal girls of the same age. Adolesc Med. 1983;103:328–331
    • View In Article
13. Dorn LD, Dahl RE, Williamson DE, et al. Developmental markers in adolescence: Implications for studies of pubertal processes. J Youth Adol. 2002;32:315–324
    • View In Article
14. Himes JH, Obarzanek E, Baranowski T, et al. Early sexual maturation, body composition, and obesity in African American girls. Obes Res. 2004;12:64S–72S
    • View In Article
15. CDC, National Center for Health Statistics. Introduction to NHANES. Available at: http://www.cdc.gov/nchs/about/major/nhanes/intro_mec.htm. Accessed March 1, 2006.
    • View In Article
16. CDC, National Center for Health Statistics. NHANES Analytic Guidelines. Available at: http://www.cdc.gov/nchs/about/major/nhanes/1nhanes2003-2004/analytical_guidelines.htm. Accessed March 1, 2006.
    • View In Article
17. CDC, National Center for Health Statistics. NHANES Datasets and Related Documentation. Available at: http://www.cdc.gov/nchs/about/major/nhanes/datalink.htm. Accessed March 1, 2006.
    • View In Article
18. Office of Management and Budget. Race and Ethnic Standards for Federal Statistics and Administrative Reporting. OMB Directive 15. 1977;Adopted on May 12
    • View In Article
19. SAS Institute. SAS 9.1. Cary, NC: SAS Institute; 2004;
    • View In Article
20. Research Triangle Institute. SUDAAN 9.0 software and Language Manual Release 9.0. Research Triangle Park, NC: Research Triangle Institute; 2004;
    • View In Article
21. Kimm SYS, Barton BA, Obarzanek E, et al. racial divergence in adiposity during adolescence: The NHLBI Growth and Health Study. Pediatrics. 2001;107:E34
    • View In Article
    • CrossRef
22. Biro FM, McMahon RP, Striegel-Moore R, et al. Impact of pubertal maturation on growth in black and white female adolescents: The National Heart, Lung, and Blood Institute Growth and Health Study. J Pediatr. 2001;138:636–643
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (141 KB)
    • CrossRef
23. Rosell M, Appleby P, Key T. Height, age at menarche, body weight and body mass index in lifelong vegetarians. Public Health Nutr. 2005;8:870–875
    • View In Article
    • MEDLINE
24. Must A, Phillips SM, Naumova EN, et al. Recall of early menstrual history and menarchal body size: After 30 years how well do women remember?. Am J Epidemiol. 2002;155:672–679
    • View In Article
    • MEDLINE
    • CrossRef
25. Maclure M, Travis LB, Willett WC, et al. A prospective cohort study of nutrient intake and age at menarche. Am J Clin Nutr. 1991;54:649–656
    • View In Article
    • MEDLINE
26. Koprowski C, Ross RK, Mack WJ, et al. Diet, body size and menarche in a multiethnic cohort. Br J Cancer. 1999;79:1907–1911
    • View In Article
    • MEDLINE
    • CrossRef
27. Freedman DS, Khan LK, Serdula MK, et al. Relation of age at menarche to race (Time period, and anthropometric dimensions: The Bogalusa Heart Study). Pediatrics. 2002;110:e43
    • View In Article
    • CrossRef
28. Casey VA, Dwer JT, Coleman KA, et al. Accuracy of recall by middle-aged participants in a longitudinal study of their body size and indices of maturation earlier in life. Ann Hum Biol. 1991;18:155–177
    • View In Article
    • MEDLINE
    • CrossRef
29. Nichols HB, Trentham-Dietz A, Hampton JM, et al. From menarche to menopause: Trends among US women born from 1912 to 1969. Am J Epidemiol. 2006;Aug 23;Epub
    • View In Article
30. Den Hond E, Roels HA, Hoppenbrouwers K, et al. Sexual maturation in relation to polychlorinated aromatic hydrocarbons: Sharpe and Skakkebaek’s hypothesis revisited. Envron Health Perspect. 2002;110:771–776
    • View In Article
31. Blanck HM, Marcus M, Tolbert PE, et al. Age at menarche and Tanner stage in girls exposed in utero and postnatally to polybromated biphenyl. Epidemiology. 2000;11:641–647
    • View In Article
    • MEDLINE
    • CrossRef
32. Brouwer A, Ahlborg UG, Vendenberg M, et al. Functional aspects of development toxicity of polyhalogenated aromatic-hydrocarbons in experimental animals and human infants. Eur J Pharmacol Environ Toxicol Pharmacol. 1995;293:1–40
    • View In Article
33. Ribot C, Pouilles JM, Bonnea M, et al. Assessment of the risk of post-menopausal osteoporosis using clinical factors. Clin Endocrinol (Oxf). 1992;36:225–228
    • View In Article
    • MEDLINE
    • CrossRef
34. Tavani A, Negri E, Francheschi S, et al. Risk factors for epithelial ovarian cancer in women under age 45. Eur J Cancer. 1993;29A:1297–1301
    • View In Article
    • MEDLINE
35. Stoll BA, Vatten LJ, Kvinnsland S. Does early physical maturity influence breast cancer risk?. Acta Oncol. 1994;33:171–176
    • View In Article
    • MEDLINE
    • CrossRef
36. La Rosa JC. Lipids and cardiovascular disease: Do findings and therapy apply equally to men and women?. Women’s Health Issues. 1992;2:102–111
    • View In Article