A comparative analysis of estimation of age at menarche by various methods in women participating in the Krakow Longitudinal Growth Study, Poland.

The purpose of this study was to see whether menarcheal age assessment by means of the most frequently used methods that were conducted every time on the same group of girls would yield the same results. One hundred and one Polish girls, whose ages at menarche were recorded in a longitudinal study between 1976 and 1990, were asked to recall the age of menarche in 2004. The mean menarcheal ages of those women were calculated by means of the probit (PA), prospective (AA), retrospective method without age correction (RA), and retrospective method with the recall age corrected by 0.5 year (RcA). The PA, AA, and retrospective methods: RA and RcA revealed results: 13.14 +/- 1.1; 13.10 +/- 1.1; 13.12 +/- 1.36, and 14.39 +/- 1.34 years, respectively. The menarcheal AA was insignificantly different from the PA (95% CI) and RA. The RcA was significantly higher than the AA and RA (P = 0.05). The correlation coefficient (r) between AA and RA was 0.70. Only 16% of the interviewed women accurately remembered the date of their menarche, 63% of them missed their menarche time by about 1 year, whereas 22% were wrong by 2 and more years. (1) The PA and the AA method yield comparable results when estimating menarcheal age. (2) The menarcheal age determined by the retrospective methods is not very reliable and the application of age correction overestimates the results.