Fig. 6

Plot of the individual sMM p-values (y-axis) against the number of cycles recorded (x-axis). The left panel contains the women without a \(\frac {2}{3}\)MRM diagnosis, the right those with a \(\frac {2}{3}\)MRM diagnosis. The lighter the dot is, the lower is her estimated migraine frequency. We see that amongst the women with an \(\frac {2}{3}\)MRM diagnosis, the p-values appear to be low when many cycles were observed, while the p-values for women with few recorded cycles are more dispersed (along the y-axis). The women with relatively low p-values in the left panel represent women with low migraine frequency, and more cycles recorded