Intended for healthcare professionals

Rapid response to:


The timing of the “fertile window” in the menstrual cycle: day specific estimates from a prospective study

BMJ 2000; 321 doi: (Published 18 November 2000) Cite this as: BMJ 2000;321:1259

Rapid Response:

The luteal phase may be more constant than you think

The authors conclude that the length of the luteal phase is highly
variable (and therefore cannot be used in family planning). How does this
fit with the work by Lenton et al 1984 (Br J Obstet Gynaecol, 1984,
91[7]:685-9), which showed for a large sample of ovulating women that the
length of the luteal phase is normally-distributed with a mean of 14.13 +/
- 1.41 days? Lenton et al's result suggests that the
luteal phase is constant to within a couple of days, not only for a given
woman, but amongst women in general. (The authors noted a tail, comprising
5% of the sample, of short luteal phases that they associated with
abnormal cycles.)

I would like to have seen in Wilcox et al's paper a plot showing
length of luteal phase versus length of cycle. If the authors are correct,
this plot will be a scatter diagram. If on the other hand Lenton et al are
right, the data points will fall in a narrow, horizontal band centred on
14.1 days. Disregarding the abnormal cycles, 68% of the points will lie
within 12.7 and 15.5 days, and 95% within 11.3 and 16.9 days.

This range is adequately covered by the guidelines for natural family
planning, which tell couples to avoid intercourse for +/- 5 days around
their predicted ovulation date, i.e., the expected date of the woman's
next period minus 14 days. (In fact most users employ the much more
reliable temperature and mucus indicators in addition to the date

Unless I missed something, I also didn't see a discussion of the
uncertainty intrinsic to the method Wilcox et al used to determine the
day of ovulation.

Competing interests: No competing interests

26 January 2002
Alison Campbell
Honorary Lecturer
St Andrews, KY16 9SS