Ovulation and Spinal Bone Mineral Densitye

Moira A. Petit, Christine L. Hitchcock, Jerilynn C. Prior, Susan I. Barr, Yvette M. Vigna and Heather A. McKay

University of British Columbia Vancouver, BC Canada Karim M. Khan University of Melbourne Melbourne NSW, Australia

De Souza et al. (1) have asserted that luteal phase abnormalities and decreased ovarian progesterone production do not affect "bone health." The investigators have done an excellent job of carefully documenting ovulation, attempting prospective monitoring of flow and menstrual cycles, measuring morphometric parameters, and observing activity through a daily exercise log (1). However, we believe that a better research design and larger sample size are necessary to support their conclusions.

Our prospective 1-yr study in ovulatory, normal-weight women of varying exercise habits documented that mean luteal phase length explained 22% of the variance in bone mineral density (BMD) change by quantitative computed tomography (QCT) (2). Further, QCT change significantly related to progesterone but not to estrogen (2). Because replication is essential for scientific knowledge, De Souza’s study (1) is welcome. We believe, however, that four methodological issues explain the apparent discrepancies between their study (1) and ours (2):

  1. De Souza et al. used a cross-sectional design, and ours was longitudinal.
  2. De Souza et al. assessed BMD by dual energy x-ray absorptiometry (DXA), and we measured cancellous bone by QCT.
  3. De Souza et al’s study was not powered to allow detection of luteal phase length effects on BMD.
  4. De Souza documented ovulatory characteristics of no more than 3 cycles versus our continuous 1-yr assessment using quantitative basal temperature.

1. De Souza et al. report no effect of luteal length on BMD by DXA measured at one time point and therefore confirm our cross-sectional assessments of BMD by QCT (2) and by DXA after 5 yr (3). Longitudinal research designs remove much of the individual variation in absolute BMD by looking at changes within individuals over time; this reduces variability, thus increasing statistical power to detect a real difference.

2. An important difference between our study and that of De Souza et al. is the type of bone assessed. DXA estimates the mineral in the whole bone (which includes both cortical and cancellous components), while QCT measures only cancellous bone within the vertebral body. In a recent prospective study in which measurements were made by both QCT and DXA in the same individuals, Friedlander et al. (4) found that a statistically significant (-3.0%) loss of cancellous bone using QCT corresponded to a nonsignificant change (-0.2%) in whole bone using DXA. This illustrates the increased responsiveness to change of cancellous bone measured by QCT.

3. Despite De Souza’s assertion that there is no effect of short luteal phase cycles on BMD, in their study exercising women with short luteal phase lengths (1) had whole-body BMD 4% lower than in exercising, ovulatory women. They indicate that they had statistical power to detect a 1 SD (13%) difference, which is an effect size of 0.47.1Power guidelines for one-way ANOVA suggest that an effect size of 0.10 is small, of 0.25 is moderate, and 0.40 is large (5). In order to obtain 80% power to detect the 4% difference they report, De Souza et al. would have needed 222 participants. With the sample sizes they used, they had a 16% power to detect the 4% difference they observed.

It is regrettable that, in their Fig. 3, De Souza et al. report statistical tests with P > 0.05 as "significant correlations...." We feel it is inappropriate to state a nonsignificant result as a primary conclusion. We also believe it is a false dichotomy to argue that, if estrogen has an important BMD effect, progesterone cannot. Our work suggests that progesterone is synergistic with and complements the necessary actions of estrogen.

4. The validity of any data relating BMD change to changes in progesterone exposure depends on accurate identification of women with ovulatory disturbances. De Souza (1) argues that we misclassified ovulatory status by use of basal temperature (BBT). BBT as used in the past has been justifiably criticized: 1) because the interpretation of temperature shift has been subjective (by eye) and thus reproducibility and accuracy have been low; 2) because the date of the BBT rise follows ovulation, rather than preceding it, BBT rise is a poor predictor of the date of ovulation; 3) because a full cycle of temperatures is required before the date of temperature shift can be determined, it is a poor choice for infertility or contraceptive interventions; and 4) because life events such as late sleeping or illness alter the morning temperature.

For all of the above reasons we have developed a quantitative analysis system and attempted to eliminate the problems of the old BBT method (6). We use a computer algorithm (Maximina; J.C. Prior, Vancouver, BC), which has been validated against the day of the LH peak (6) and confirmed by assays of serum progesterone (2, and our unpublished data). Women in our study also recorded temperatures as a list along with a record of daily events that may have influenced the temperature reading including fever, poor sleep, waking early or late. Therefore, the method we use for basal temperature documentation of ovulation and luteal phase length is valid, accurate, noninvasive, inexpensive, and accessible to women. In physiological studies a quantitative BBT method with occasional hormonal confirmation is less burdensome for women than other available methods, and it allows for long-term, prospective data collection. Because it is clear that ovulatory characteristics do not remain constant over time (3, 7) and that they change cycle by cycle even within women (2, 3, 7, 8, 9), continuous or long-term monitoring is critical.

We agree with De Souza et al. that ideally, progesterone and estrogen should be directly assayed, rather than inferred from BBT-documented ovulatory status. We do not agree that the 3 months of intensive hormone sampling they performed gives a better picture of ovulatory status and experience than 12 months of continuous, quantitatively assessed BBT data.

In summary, we are grateful that colleagues such as B. L. Lasley, De Souza, Waller, and others have focused scientific attention on the variability of menstrual cycle and ovulatory characteristics and their influences on health and on BMD. However, we believe that the conclusive study should be adequately powered, should assess changes in BMD over several years utilizing both DXA and QCT technologies, and should document ovulatory and cycle characteristics continuously, or at least for six cycles per year. A method also needs to be developed that will allow accurate, noninvasive assessment of ovulation and hormonal characteristics that is acceptable to women for long-term data collection and that will allow epidemiological investigation of the relationships between ovulation and changes in cancellous and cortical bone.

Footnotes

Address correspondence to: Jerilynn C. Prior, Department of Medicine, The University of British Columbia, Vancouver Hospital and Health Sciences Center, 855 West 12th Avenue, Vancouver, BC Canada V5Z 1M9.

1 We estimated power using GPOWER (11) and a pooled estimate of standard deviation of 0.115, calculated from Table 3. We assumed a one-way ANOVA design, with three groups of equal size. The outlying value for the SEM in the sedentary ovulatory group was assumed to be a typographical error; if this is not the case, or if the groups have unequal sizes, then the required sample size would be higher. Back

Received April 7, 1998.

References

  1. De Souza MJ, Miller BE, Sequenzia LC, et al. 1997 Bone health is not affected by luteal phase abnormalities and decreased ovarian progesterone production in female runners. J Clin Endocrinol Metab. 82:2867–2876.[Abstract/Free Full Text]
  2. Prior JC, Vigna YM, Schechter MT, Burgess AE. 1990 Spinal bone loss and ovulatory disturbances. N Engl J Med. 323:1221–1227.[Abstract]
  3. Prior JC, Vigna YM, Barr SI, Kennedy S, Schulzer M, Li DKB. 1996 Ovulatory premenopausal women lose cancellous spinal bone: A 5-year prospective study. Bone. 18:261–267.[CrossRef][Medline]
  4. Friedlander AL, Genant HK, Sadowsky S, Byl NN, Glüer C. 1995 A 2-year program of aerobics and weight training enhances bone mineral density of young women. J Bone Miner Res. 10:574–585.[Medline]
  5. Erdfelder E, Faul F, Buchner A. 1996 GPower: A general power analysis program. Behav Res Methods Instrum Comput. 28:1–11.
  6. Prior JC, Vigna YM, Schulzer M, Hall JE, Bonen A. 1990 Determination of luteal phase length by quantitative basal temperature methods: Validation against the midcycle LH peak. Clin Invest Med. 13:123–131.[Medline]
  7. Prior JC, Vigna YM, Barr SI, Rexworthy C, Lentle BC. 1994 Cyclic medroxyprogesterone treatment increases bone density: A controlled trial in active women with menstrual cycle disturbances. Am J Med. 96:521–530.[Medline]
  8. Bullen BA, Skrinar GS, Beitins IZ, Von Mering G, Turnbull BA, McArthur JW. 1985 Induction of menstrual cycle disorders by strenuous exercise in untrained women. N Engl J Med. 312:1349–1353.[Abstract]
  9. Prior JC, Ho-Yuen B, Clement P, Bowie L, Thomas J. 1982 Reversible luteal phase changes and infertility associated with marathon training. Lancet. 1:269–270.




This Article
Full Text (PDF)
Submit a related Letter to the Editor
Purchase Article
View Shopping Cart
Alert me when this article is cited
Alert me when eLetters are posted
Alert me if a correction is posted
Services
Email this article to a friend
Similar articles in this journal
Similar articles in PubMed
Alert me to new issues of the journal
Download to citation manager
Request Copyright Permission
Google Scholar
Articles by Petit, M. A.
Articles by McKay, H. A.
Articles citing this Article
PubMed
PubMed Citation
Articles by Petit, M. A.
Articles by McKay, H. A.


HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Endocrinology Endocrine Reviews J. Clin. End. & Metab.
Molecular Endocrinology Recent Prog. Horm. Res. All Endocrine Journals