Normal probability plots were used to assess the homogeneity of a population of 327 luteal phases from apparently ovulatory menstrual cycles. The length of the luteal phase was defined as the interval (in days) following but not including, the luteinizing hormone peak, up to and including the day before onset of menstruation. A small sub-set of the population consisted of cycles with abnormally short luteal phases but the majority of the data followed a normal frequency distribution which gave a mean (+ SD) for normal luteal phase length of 14.13 (k 1.41) days. It was estimated that all cycles with a luteal phase < 9 days were abnormal, and that 74%, 22% and 2% respectively of cycles with luteal phases of 10, 11 and 12 days were also abnormal. The total incidence of short luteal phases defined as above was 5.2%.

Luteinizing hormone (LH), follicle-stimulating hormone (FSH), oestradiol and progesterone concentrations in plasma were obtained daily throughout the menstrual cycles of 94 regularly cycling women, aged between 24 and 50 years. Although mean LH concentrations changed little with advancing age, mean FSH concentrations were significantly (P less than 0.001) elevated from the age of 39 years. FSH concentrations in the oldest women studied (48-50 years) were approximately 3-fold greater than in the younger controls (women aged 23-35 years). LH concentrations rose slightly (P less than 0.05) during the last 5 years only. The increase in FSH concentration was not, however, uniform across the cycle, but was confined predominantly to the mid-follicular and post-ovulatory phases (i.e. those times in the normal menstrual cycle when circulating inhibin concentrations appear to be minimal). Despite the clear increases in FSH concentration, there was little alteration in the mean steroid profiles which remained within the normal fertile range throughout the last decade of reproductive life. The only exception to this was a small, transient, but significant (P less than 0.05) decrease in preovulatory oestradiol concentration between the ages of 36 and 38 years, which was followed by a transient increase (P less than 0.01) in oestradiol concentration between 39 and 44 years. However, no corresponding significant changes in mean progesterone concentrations were observed.

Normal probability plots were used to analyse the distribution of follicular phase length in a population of 293 apparently ovulatory menstrual cycles from women aged between 18 and 39 years. The length of the follicular phase was defined as the interval (in days) from the onset of menstruation up to, but not including, the day of the LH peak. Follicular phase length appeared to be log-normally distributed and graphical inspection of the probability plot suggested that the geometric mean (and 95% confidence limits) of follicular phase length in this study group was 12.9 (10.3 to 16.3) days. There was a significant decrease (P

No abstract