- P H Whincup (), professor of cardiovascular epidemiologya,
- J A Gilg, research statisticiana,
- K Odoki, clinical research fellowa,
- S J C Taylor, senior clinical lecturerb,
- D G Cook, professor of epidemiologya
- a Department of Public Health Sciences, St George's Hospital Medical School, London SW17 0RE
- b Department of General Practice and Primary Care, Medical Sciences, Queen Mary and Westfield College, London E1 4NS
- Correspondence to: P H Whincup
The possibility that puberty is occurring earlier in Britain than previously has caused great interest.1 Despite the importance of menarcheal age as an indicator of puberty,2 there is little information on menarcheal age in contemporary teenagers to compare with data on girls born in the 1950s and 1960s. We report on the distribution of menarcheal age in a survey of British girls born between 1982 and 1986.
Participants, methods, and results
In 1998-9 we studied the cardiovascular health of secondary school children aged 12-16 in schools in 10 British towns: five in southern England (Esher, Leatherhead, Chelmsford, Bath, Tunbridge Wells), three in north west England (Wigan, Burnley, Rochdale), and two in south Wales (Port Talbot, Rhondda). We approached those secondary schools corresponding to a stratified random sample of primary schools in our earlier study3; 62 of 65 (95%) with female pupils participated. In each school we invited girls from the earlier study to participate, with a supplementary random sample of pupils from the same classes.3 During screening the girls completed a confidential self administered questionnaire on date of birth, whether they had started their periods and, if so, their age (years and months) at the first period. Social class was based on parental occupation (using the Registrar General's 1990 classification). Ethnicity was based on appearance and cross checked with surname and parental self assessment. We used SAS (version 6.12) for the statistical analysis. We determined the median age of menarche and confidence intervals using survival analysis with PROC LIFETEST. We included girls who had not yet had their first period (88 participants); for girls providing only year of menarche (231), the month was imputed using the mean value for other girls of the same age in years. Probit transformation2 of the percentage of affirmative responses at each age gave almost identical results. In all, 1166 girls aged 12-16 years (1068 European, 79 South Asian, 19 other) reported their menarcheal age (response 66%). The median menarcheal age was 12 years 11 months (95% confidence interval 12 years 10 months to 13 years 1 month). The percentages of girls who reported having had their first period by their 10th, 11th, or 12th birthdays were 0.8, 3.6 and 21.7, respectively; 11.8% had their first period before leaving primary school. Median ages of menarche were similar in different regions (table) and did not differ by social class or ethnic group (see table on website). Non-responders closely resembled responders in age and geographical location.
The median age of menarche in contemporary British teenagers is around 13 years. In our study geographical, social, and ethnic variations were small, suggesting that non-response bias in menarcheal age was likely to be limited. Comparison with British girls born between 1950 and 1965 (table) suggests that the median menarcheal age reported here is close to or slightly below the earlier findings. Two points emerge clearly from the results. Firstly, any decrease in average menarcheal age during the past 20-30 years has been small (almost certainly less than six months), particularly when compared with the reduction of a year or more that occurred in many European countries (including Britain) between the late 19th and mid 20th centuries. 2 4 Secondly, even though no appreciable recent decrease in menarcheal age has occurred, almost one girl in eight reaches menarche while still at primary school. This needs to be taken into account when providing sanitary facilities and health information for female pupils in primary school.
We thank all the participants for their help.
Contributors: PHW and DGC had the idea for the paper and designed the study with support from KO and SJCT. JAG carried out the analyses. PHW drafted the paper, with contributions from all authors. PHW and DGC will act as guarantors for the paper.
Funding This study was funded by the Wellcome Trust (grant 051187/Z/97/A).
Competing interests None declared.
A figure showing age at menarche, a longer version of the table, and details of the previous studies appear on the BMJ's website