Fig 1.
Data selection procedures and sample sizes.
Fig 2.
Confirmation of Gaussian transformation of AMH values by probability plot.
The vertical axis represents the cumulative frequency (P %), transformed as log [(100 − P)/P]. The horizontal axis represents AMH test results. This axis was transformed by use of Box–Cox formula using p = 0.4 as an optimal power. The linearity of the segment between 10 and 90% indicates successful Gaussian transformation by use of the formula.
Fig 3.
Data distribution of AMH before (left) and after transformation where the scale is adjusted to fit the AMH power transformation (right).
Fig 4.
Scatterplot of AMH versus age without (left) / with adjustment of the scale of AMH axis to fit the AMH power transformation (right).
Fig 5.
Nomogram of age-related profile of Antimullerian hormone (AMH).
The scale is adjusted to fit the AMH power transformation. The central lines and the outer margins of the colour shades represent the median and the fifth and 95th centiles for the year-by-year moving vertical 10-year segment.
Table 1.
Reference intervals of AMH determined for every 5 years of age from 16 to 49.