Season of birth and recalled age at menarche
Boldsen, J.L.
Journal of Biosocial Science 24(2): 167-173
1992
ISSN/ISBN: 0021-9320 PMID: 1583031 DOI: 10.1017/s0021932000019702Document Number: 319106
This study examines seasonality of age of menarche among 908 schoolgirls 14-16 years in 1987 from Odense, Denmark and seasonality of birth. A medial examination was provided and information received from 69% (611) on the month of menstruation, 16.9% (150) on the season, and 9.7% (86) who were premenarcheal. The GLM, 3.77, statistical package was used to estimate the mean and variance of data which were treated as grouped observations from a normal distribution. The results show seasonality for time at menarche (x2=40.99, df=3,p.005), and confirms Hungarian findings. Winter and summer are the expected seasons of menarche. There was no seasonality to the birthrate (x2=2.85, df=3,p.4) even when subdivided by type of district, and no interaction between season of birth and season of menarche (x2=10.05,db=9,p.3). Month of birth and month of menarche were also unrelated 9x2=1.06,df=1,p.03). In a Fourier method of analysis of the means of age at menarche, the fit of the curve shows that the mean age at menarche is low for girls in November, December, May, June, and July, and higher in February, March, August, September, and October. The major new findings is that the cause of menarche seasonality is the different mean age of menarche of girls born in different seasons. The results also do not appear to be a consequence of outliers in the peak months (March-September). 2 way analysis of variance for variances and mean was used to analyze months of menarche by 4 seasons (Feb-Apr, May-July, Aug-Oct, Nov-Jan) and district for possible social effects. The results were a statistically significant effect of seasonal differences and season/district interaction on mean age at menarche. Variances are significantly different due to larger variance for age at menarche among girls attending private school (var=1.70,db=91) than public schools (var=1.241, db=783). This reflects the diversity in the private population which is less well off and ethnically more diverse and less exposed to changing weather conditions. It appears that there are ecological differences between living in a suburb and in an inner city area that alters the biological response to it.