- Email: [email protected]
Oral contraceptives and bone mineral density: A population-based study
Oral contraceptives and bone mineral density: A populationbased study Julie A. Pasco, PhD,a Mark A. Kotowicz, MB, BS,a Margaret J. Henry, BSc,a Soheila Panahi, BSc,a Ego Seeman, MD,b and Geoffrey C. Nicholson, MB, BS, PhDa Geelong, Victoria, Australia OBJECTIVE: We sought to test the hypothesis that exposure to oral contraceptives protects the skeleton. STUDY DESIGN: Multiple regression techniques were used to analyze data for a random sample of 710 Australian women (age range, 20-69 years). Bone mineral density was measured at the lumbar spine, proximal femur, whole body, and distal forearm. Oral contraceptive exposure was assessed by a questionnaire. RESULTS: Women exposed to oral contraceptives had a 3.3% greater mean bone mineral density adjusted for body mass index and age at the lumbar spine (partial r 2 = 0.009; P = .014). Adjusted mean vertebral bone mineral density was 3.3% greater for premenopausal women (partial r 2 = 0.008; P < .05), but the effect did not reach significance among postmenopausal women. Higher bone mineral density was associated with increased duration of exposure, with a mean increase of 3.2% associated with the first 5 years and a further 0.2% with ≥5 years of exposure. No association was detected at other sites. CONCLUSION: Exposure to oral contraceptives may be associated with higher lumbar spine bone mineral density. (Am J Obstet Gynecol 2000;182:265-9.)
Key words: Oral contraceptives, bone mineral density, epidemiology
Whereas treatment of postmenopausal women with exogenous estrogen can reverse the adverse effects of ovarian failure on bone mass,1 it is unclear whether bone mineral density can be modified by the premenopausal administration of exogenous estrogen associated with oral contraceptives. Several,2-8 but not all,9-13 studies suggest that exposure to oral contraceptives may preserve or even increase bone mineral density. There is evidence that premenopausal bone loss starts soon after attainment of peak bone mass in late adolescence.14, 15 Vertebral bone mineral density decreases at approximately 1% each year in premenopausal women,16, 17 whereas the rates of loss may be less at the proximal femur18 and distal forearm.19 Pharmacologic doses of estrogen used to suppress ovulation may exert a positive influence on bone, preventing or slowing bone loss, particularly at sites rich in cancellous bone, which is more responsive to hormonal stimuli because of its greater surface area.22 From the Departments of Medicine, St Vincent’s and The Geelonga and Austin & Repatriationb Hospitals, the University of Melbourne. Supported by the Victorian Health Promotion Foundation and the Geelong Region Medical Research Foundation. Received for publication March 15, 1999; revised July 12, 1999; accepted September 23, 1999. Reprint requests: Julie A. Pasco, PhD, The University of Melbourne, Department of Medicine, The Geelong Hospital, PO Box 281, Geelong 3220, Australia. Copyright © 2000 by Mosby, Inc. 0002-9378/2000 $12.00 + 0 6/1/103215
We hypothesized that exposure to oral contraceptives protects the skeleton in apparently healthy women. The aim of this population-based study was to determine whether exposure to oral contraceptives was associated with greater bone mineral density. Material and methods Subjects. An age-stratified sample of women was selected at random from electoral rolls for the Barwon Statistical Division, a geographically distinct area surrounding Geelong in southeastern Australia. Of the 1602 eligible women (age range, 20-92 years), 1291 agreed to participate. For this analysis, we excluded women aged >69 years (n = 220) and those exposed to diseases and other drugs known to influence bone mass (n = 361). Thus 710 women (age range, 20-69 years; mean age ± SD, 41.1 ± 13.8 years; 511 premenopausal, 172 postmenopausal, and 27 in whom menopausal status was indeterminate) were included in the analysis. Written informed consent was obtained from all participants. The study was approved by The Geelong Hospital Human Research and Ethics Advisory Committee. Measurements. Dual-energy x-ray absorptiometry was performed by a Lunar DPX-L densitometer, and the data were analyzed by Lunar DPX-L software version 1.31. Bone mineral density was measured at the spine (L2-4) in the posterior-anterior projection, proximal femur, whole body, ultradistal forearm, and midforearm sites. Short-term precision in vivo was 0.6% for the spine, 1.6% 265
266 Pasco et al
February 2000 Am J Obstet Gynecol
Table I. Characteristics and unadjusted bone mineral density (mean ± SD) of women exposed and unexposed to oral contraceptives All subjects Exposed Age (y) 38.9 ± 12.3* Weight (kg) 68.0 ± 14.0 Height (cm) 162.8 ± 6.1* Body mass index (kg/m2) 25.7 ± 5.3* Calcium (mg/d) 652 ± 353 Parity 2.0 ± 1.5 Lactation (mo) 10.2 ± 14.9 Alcohol (standard drinks/wk) 4.0 ± 6.0* Smoking (pack-years) 5.5 ± 11.5 Menopausal age (y) 46.8 ± 5.7* Active (No. and %) 493 (85%)* Bone mineral density (g/cm2) Spine 1.221 ± 0.157* (n = 579) Proximal femur Femoral neck 0.972 ± 0.139* (n = 577) Ward’s triangle 0.863 ± 0.164* (n = 577) Trochanter 0.827 ± 0.135* (n = 577) Whole body 1.164 ± 0.088* (n = 577) Ultradistal forearm 0.329 ± 0.051* (n = 573) Midforearm 0.703 ± 0.067* (n = 572)
Premenopausal subjects Unexposed
Exposed
Unexposed
50.6 ± 16.2* 69.9 ± 15.2 160.8 ± 6.3* 27.1 ± 6.1* 705 ± 380 2.3 ± 2.2 10.3 ± 14.8 2.5 ± 4.8* 4.4 ± 12.1 48.6 ± 5.2* 97 (74%)*
34.4 ± 8.2 67.4 ± 13.4 163.3 ± 5.9 25.3 ± 5.0 650 ± 337 1.7 ± 1.4 10.3 ± 15.0* 3.9 ± 5.1* 4.3 ± 8.7 — 405 (88%)
32.2 ± 10.5 70.3 ± 18.8 163.6 ± 5.5 26.3 ± 6.9 648 ± 347 1.3 ± 1.7 6.7 ± 11.5* 2.3 ± 3.9* 3.3 ± 7.4 — 41 (84%)
1.119 ± 0.189* (n = 131)
1.245 ± 0.134 (n = 462)
1.213 ± 0.164 (n = 49)
0.924 ± 0.162* (n = 131) 0.803 ± 0.198* (n = 131) 0.813 ± 0.153* (n = 131) 1.124 ± 0.110* (n = 131) 0.311 ± 0.061* (n = 131) 0.666 ± 0.085* (n = 131)
0.991 ± 0.126 (n = 461) 0.889 ± 0.148 (n = 461) 0.835 ± 0.125 (n = 461) 1.175 ± 0.075 (n = 461) 0.333 ± 0.046 (n = 459) 0.714 ± 0.055 (n = 458)
1.019 ± 0.151 (n = 49) 0.933 ± 0.187 (n = 49) 0.849 ± 0.161 (n = 49) 1.177 ± 0.092 (n = 49) 0.330 ± 0.038 (n = 49) 0.704 ± 0.054 (n = 49)
*P < .05 (significant difference).
for the femoral neck, 2.1% for Ward’s triangle, 1.6% for the trochanter, 0.4% for the whole body, 2.1% for the ultradistal forearm, and 1.1% for the midforearm sites. Height and weight were measured to the nearest 0.1 cm and 0.1 kg, respectively, and the body mass index was calculated in kilograms per square meter. Details of reproductive history and lifestyle factors were documented by questionnaire during an interview. Habitual dietary and supplemental calcium intake was estimated by using a calcium-specific food-frequency questionnaire. Cigarette smoking was calculated in pack-years (1 pack-year is equivalent to 20 cigarettes per day per year), whereas alcohol consumption was recorded as the number of standard drinks per week (1 standard drink is equivalent to approximately 10 g of alcohol). Current activity levels were classified as active if exercise was performed regularly; otherwise they were designated as sedentary. For the purposes of this analysis, women who had undergone hysterectomy and retained one or both ovaries were not included in the postmenopausal group. Statistics. Two-sample t tests were used to test for differences between women exposed and unexposed to oral contraceptives for continuous variables, and the χ2 test of independence was used to test for differences in discrete variables. Analysis of variance, followed by the Tukey test of multiple comparisons, was used to test for differences in unadjusted bone mineral density with duration of exposure to oral contraceptives. By use of bone mineral density as the dependent variable, covariates were identified by best-fitting regression techniques21 for each site.
Analysis was performed for the whole group and for premenopausal and postmenopausal women separately. Best models included adjustments for age, weight, height, or body mass index, with some sites requiring higher than linear adjustments for age, centered about the mean to avoid collinearity. Menopause status was not included for the whole group because its contribution was negligible. The influence on bone mineral density of discrete variables was determined by analysis of covariance, and continuous variables were analyzed by multiple regression techniques. All analyses were performed by the Minitab (release 11) software package. Results Analysis of the whole group. Women exposed to oral contraceptives (n = 579) were younger and taller, had a lower body mass index, and consumed more alcohol than unexposed women (n = 131). Reflecting the age distribution of the groups, unadjusted bone mineral density was greater for women exposed to oral contraceptives at all sites except the trochanter (Table I). There was a positive relationship between exposure to oral contraceptives and bone mineral density at the spine adjusted for body mass index and age (partial r2 = 0.009; P = .014), with women exposed to oral contraceptives having a 3.3% greater mean adjusted bone mineral density. Ages for first exposure to oral contraceptives ranged from 11 to 48 years, with a mean and SD of 21.1 ± 5.7 years. Age at first exposure to oral contraceptives had no detectable effect on bone mineral density. Among women exposed to oral contraceptives, 44% (n = 257)
Pasco et al 267
Volume 182, Number 2 Am J Obstet Gynecol
Postmenopausal subjects Exposed
Unexposed
59.7 ± 6.8* 70.8 ± 15.5 160.1 ± 6.1 27.6 ± 5.7 699 ± 418 3.2 ± 1.8 10.2 ± 14.9 3.1 ± 5.7 11.6 ± 19.7* 46.8 ± 5.7* 72 (77%)
62.1 ± 4.8* 69.3 ± 12.4 159.2 ± 6.3 27.5 ± 5.5 755 ± 396 3.0 ± 2.4 13.0 ± 16.4 2.8 ± 5.4 5.3 ± 14.5* 48.6 ± 5.2* 54 (69%)
1.094 ± 0.181 (n = 94)
1.060 ± 0.180 (n = 78)
0.877 ± 0.159 (n = 93) 0.739 ± 0.181 (n = 93) 0.785 ± 0.162 (n = 93) 1.103 ± 0.114 (n = 93) 0.300 ± 0.063 (n = 93) 0.644 ± 0.088 (n = 93)
0.867 ± 0.142 (n = 78) 0.723 ± 0.162 (n = 78) 0.789 ± 0.146 (n = 78) 1.089 ± 0.108 (n = 78) 0.299 ± 0.070 (n = 78) 0.642 ± 0.094 (n = 78)
had used them for <5 years, and 56% (n = 322) had used them for ≥5 years. There was a pattern of increasing bone mineral density (unadjusted) at the spine with increasing duration of exposure (Fig 1), which failed to reach significance with each increment. However, after we adjusted for age and body mass index, a comparison with unexposed women showed a 3.2% higher mean bone mineral density associated with exposure up to 5 years and a further modest but significant increase of 0.2% with ≥5 years of exposure (P < .05). There was no detectable influence of age at menarche, calcium or alcohol intake, smoking, activity levels, parity, or lactation on the association between exposure to oral contraceptives and bone mineral density. No association was detected between exposure to oral contraceptives and bone mineral density at other sites. Separate analyses of premenopausal and postmenopausal women. There was a longer period of lactation and a greater alcohol intake among premenopausal women exposed to oral contraceptives, but there were no differences in physical characteristics between the groups (Table I). No differences in unadjusted bone mineral density existed between exposed and unexposed premenopausal women. There was a positive association between exposure to oral contraceptives and bone mineral density adjusted for weight and height at the spine (partial r2 = 0.008; P < .05), with exposed women having a 3.3% greater adjusted mean bone mineral density. Thirtyseven percent (n = 170) of the premenopausal women exposed to oral contraceptives reported current use; however, current or past use of oral contraceptives was not a
Fig 1. Unadjusted bone mineral density (mean ± SE) at lumbar spine (L2-4) with increasing duration of exposure to oral contraceptives (asterisk, significant difference, P < .05). BMD, Bone mineral density.
significant predictor of bone mineral density. Within this subgroup, the protective effect associated with oral contraceptive exposure was detected at the spine alone. Among postmenopausal women, those exposed to oral contraceptives were younger, had a greater exposure to cigarette smoking, and had an earlier age at menopause (Table I). The association detected at the lumbar spine for the whole group failed to reach significance for both unadjusted (Table I) and adjusted bone mineral density within this subgroup. Comment We report results from a population-based study that indicate that women who have been exposed to oral contraceptives have a greater bone mineral density at the lumbar spine. Age at menarche, calcium intake, parity, lactation, habitual exercise, smoking, alcohol consumption, and age at first exposure had no influence on the association between exposure to oral contraceptives and bone mineral density. There was a pattern of increasing bone mineral density with increasing duration of exposure, with the greatest effects occurring during the first 5 years of exposure. Our regression models predict that greater bone mineral density may be sustained throughout the postmenopausal years. However, the association demonstrated in premenopausal women did not reach significance among postmenopausal women and may indicate that the positive effect of oral contraceptives diminishes with an increase in the number of years after cessation. Positive associations have been reported between oral contraceptive use and bone mineral density at the lumbar spine,3, 5, 6, 8 distal radius,2, 4, 6 proximal femur,5, 8 and the whole body.7 Lindsay et al3 observed among premenopausal women that exposure to oral contraceptives was associated with a 12% higher mean bone mineral
268 Pasco et al
density, whereas among postmenopausal women the difference failed to reach significance. They reported a linear relationship, with 1% higher mean vertebral bone mineral density associated with each year of exposure. Kleerekoper et al6 also suggested that the degree of protection to the skeleton was related to duration of exposure. Hreshchyshyn et al9 reported no difference in bone mineral density at the spine and hip with oral contraceptive use; however, exclusion of women with a history of fractures or oophorectomy may have excluded women with predominantly low bone mineral densities. In a prospective study of premenopausal women, Mais et al13 reported a nonsignificant trend for higher bone mineral density associated with the use of low-dose oral contraceptives. A small sample size and a short period of exposure may have limited the ability to detect small differences in this and other studies.10, 22 Hartard et al23 studied the effects of exercise in combination with exposure to oral contraceptives in a group of premenopausal women and found the highest bone mineral density in the group characterized by long-term exercise and shortterm exposure to oral contraceptives. No effects were noted for long-term exposure, irrespective of activity levels. We were unable to detect any influence of lifestyle factors on the association between oral contraceptive use and bone mineral density. In our study alcohol consumption and smoking were self-reported and possibly underestimated. Abstainers numbered 30% for alcohol and 57% for smoking. We observed a trend for greater exposure to smoking among older age strata in our cohort. However, smoking was not a significant confounder, despite a greater exposure among our postmenopausal women who had used oral contraceptives. Community-use estimates for Australia indicate that patterns of prescribing oral contraceptives have changed. Between 1990 and 1997, the proportion of fixed combinations (progestogen-estrogen) changed from 38% to 48%, sequential preparations from 54% to 42%, and progestogen-only preparations from 8% to 10%. Current data indicate that 10% of prescriptions for combined oral contraceptives in Australia are for high (50 µg) and 90% for low (30-40 µg) ethinyl estradiol contents (Commonwealth Department of Health and Family Services, Drug Utilisation Sub-Committee database, 1990-1997). In this study the use of oral contraceptives was self-reported, and although there may be good agreement between medical records and self-reported past medication use,24 43% of our subjects could not recall the brand of any oral contraceptives used. We were unable to identify women exposed to higher doses of estrogen and the small proportion (<10%) who had taken progestogen-only preparations. Therefore exposure in this study is broad-based, and the dose remains unspeci-
February 2000 Am J Obstet Gynecol
fied. Similar limitations may be responsible for some of the discrepancies in the literature because different dose responses in the skeleton have been suggested.25 In conclusion, in women exposed to oral contraceptives, we detected a modest but clinically significant higher bone mineral density, which would be expected to reduce fracture risk by approximately 20%.26 Whether the higher bone mineral density is the result of either higher peak bone mineral density or a reduction in bone loss before menopause or both of these will require further study. We thank B. Skoric for her assistance with scanning, interviewing subjects, and completing questionnaires and P. McManus and J. Dudley from the Commonwealth Department of Health and Family Services for supplying the community use estimates for oral contraceptives. REFERENCES
1. Felson DT, Zhang Y, Hannan MT, Kiel DP, Wilson PWF, Anderson JJ. The effect of postmenopausal estrogen therapy on bone density in elderly women. N Engl J Med 1993;329:1141-6. 2. Goldsmith NF, Johnston JO. Bone mineral: effects of oral contraceptives, pregnancy, and lactation. J Bone Joint Surg Am 1975;57:657-68. 3. Lindsay R, Tohme J, Kanders B. The effect of oral contraceptive use on vertebral bone mass in pre- and post-menopausal women. Contraception 1986;34:333-40. 4. Enzelsberger H, Metka M, Heytmanek G, Schurz B, Kurz Ch, Kusztrich M. Influence of oral contraceptive use on bone density in climacteric women. Maturitas 1988;9:375-8. 5. Stevenson JC, Lees B, Devenport M, Cust MP, Ganger KF. Determinants of bone density in normal women: risk factors for future osteoporosis? BMJ 1989;298:924-8. 6. Kleerekoper M, Brienza RS, Schultz LR, Johnson CC. Oral contraceptive use may protect against low bone mass. Arch Intern Med 1991;151:1971-6. 7. Recker RR, Davies KM, Hinders SM, Heaney RP, Stegrnan MR, Kimmel DB. Bone gain in young adult women. JAMA 1992;268:2403-8. 8. Kritz-Silverstein D, Barrett-Connor E. Bone mineral density in postmenopausal women as determined by prior oral contraceptive use. Am J Public Health 1993;83:100-2. 9. Hreshchyshyn MM, Hopkins A, Zylstra S, Anbar M. Associations of parity, breast-feeding, and birth control pills with lumbar spine and femoral neck bone densities. Am J Obstet Gynecol 1988;159:318-22. 10. Lloyd T, Buchanan JR, Ursino GR, Myers C, Woodward G, Halbert DR. Long-term oral contraceptive use does not affect trabecular bone density. Am J Obstet Gynecol 1989;160:402-4. 11. Mazess RB, Barden HS. Bone density in premenopausal women: effects of age, dietary intake, physical activity, smoking, and birth-control pills. Am J Clin Nutr 1991;53:132-42. 12. Murphy S, Khaw KT, Compston JE. Lack of relationship between hip and spine bone mineral density and oral contraceptive use. Eur J Clin Invest 1993;23:108-11. 13. Mais V, Fruzzetti F, Ajossa S, Paoletti AM, Guerriero S, Melis GB. Bone metabolism in young women taking a monophasic pill containing 20 mcg ethinylestradiol: a prospective study. Contraception 1993;48:445-52. 14. Haapasalo H, Kannus P, Sievanen H, Pasanen M, Uusi-Rasi K, Heinonen A, et al. Development of mass, density, and estimated mechanical characteristics of bones in Caucasian females. J Bone Miner Res 1996;11:1751-60. 15. Matkovic V, Jelic T, Wardlaw GM, Ilich JZ, Goel PK, Wright JK, et al. Timing of peak bone mass in Caucasian females and its im-
Volume 182, Number 2 Am J Obstet Gynecol
16.
17.
18.
19. 20.
plication for the prevention of osteoporosis. Inference from a cross-sectional model. J Clin Invest 1994;93:799-808. Riggs BL, Wahner HW, Melton LJ III, Richelson LS, Judd HL, Offord KP. Rates of bone loss in the appendicular and axial skeletons of women. Evidence of substantial vertebral bone loss before menopause. J Clin Invest 1986;77:1487-91. Baran D, Sorensen A, Grimes J, Lew R, Karellas A, Johnson B, et al. Dietary modification with dairy products for preventing vertebral bone loss in premenopausal women: a three-year prospective study. J Clin Endocrinol Metab 1990;70:264-70. Hedlund LR, Gallagher JC. The effect of age and menopause on bone mineral density of the proximal femur. J Bone Miner Res 1989;4:639-42. Duppe H, Gardsell P, Johnell O, Nilsson BE. Bone mineral content in women: trends of change. Osteoporos Int 1992;2:262-5. Seeman E, Wahner HW, Offord KP, Kumar R, Johnson WJ, Riggs BL. Differential effects of endocrine dysfunction on the axial and the appendicular skeleton. J Clin Invest 1982;69:1302-9.
Pasco et al 269
21. Ott R. An introduction to statistical methods and data analysis. 4th ed. California: Wadsworth, Inc; 1993. p. 647-733. 22. Sowers MR, Wallace RB, Lemke JH. Correlates of mid-radius bone density among postmenopausal women: a community study. Am J Clin Nutr 1985;41:1045-53. 23. Hartard M, Bottermann P, Bartenstein P, Jeschke D, Schwaiger M. Effects on bone mineral density of low-dosed oral contraceptives compared to and combined with physical activity. Contraception 1997;55:87-90. 24. Harlow SD, Linet MS. Agreement between questionnaire data and medical records. The evidence for accuracy of recall. Am J Epidemiol 1989;129:233-48. 25. Horsman A, Jones M, Francis R, Nordin C. The effect of estrogen dose on postmenopausal bone loss. N Engl J Med 1983;309:1405-7. 26. Melton LJ III, Atkinson EJ, O’Fallon WM, Wahner HW, Riggs BL. Long-term fracture prediction by bone mineral assessed at different skeletal sites. J Bone Miner Res 1993;8:1227-33.
Bound volumes available to subscribers Bound volumes of the American Journal of Obstetrics and Gynecology are available to subscribers (only) for the 2000 issues from the publisher, at a cost of $122.00 for domestic, $156.22 for Canada, and $146.00 for international for Vol. 182 (January-June) and Vol. 183 (July-December). Shipping charges are included. Each bound volume contains a subject and author index, and all advertising is removed. Copies are shipped within 60 days after publication of the last issue in the volume. The binding is durable buckram with the Journal name, volume number, and year stamped in gold on the spine. Payment must accompany all orders. Contact Mosby, Inc, Subscription Services, 11830 Westline Industrial Dr, St. Louis, MO 63146-3318, USA; phone (800)453-4351 or (314)453-4351. Subscriptions must be in force to qualify. Bound volumes are not available in place of a regular Journal subscription.
Key words: Oral contraceptives, bone mineral density, epidemiology
Whereas treatment of postmenopausal women with exogenous estrogen can reverse the adverse effects of ovarian failure on bone mass,1 it is unclear whether bone mineral density can be modified by the premenopausal administration of exogenous estrogen associated with oral contraceptives. Several,2-8 but not all,9-13 studies suggest that exposure to oral contraceptives may preserve or even increase bone mineral density. There is evidence that premenopausal bone loss starts soon after attainment of peak bone mass in late adolescence.14, 15 Vertebral bone mineral density decreases at approximately 1% each year in premenopausal women,16, 17 whereas the rates of loss may be less at the proximal femur18 and distal forearm.19 Pharmacologic doses of estrogen used to suppress ovulation may exert a positive influence on bone, preventing or slowing bone loss, particularly at sites rich in cancellous bone, which is more responsive to hormonal stimuli because of its greater surface area.22 From the Departments of Medicine, St Vincent’s and The Geelonga and Austin & Repatriationb Hospitals, the University of Melbourne. Supported by the Victorian Health Promotion Foundation and the Geelong Region Medical Research Foundation. Received for publication March 15, 1999; revised July 12, 1999; accepted September 23, 1999. Reprint requests: Julie A. Pasco, PhD, The University of Melbourne, Department of Medicine, The Geelong Hospital, PO Box 281, Geelong 3220, Australia. Copyright © 2000 by Mosby, Inc. 0002-9378/2000 $12.00 + 0 6/1/103215
We hypothesized that exposure to oral contraceptives protects the skeleton in apparently healthy women. The aim of this population-based study was to determine whether exposure to oral contraceptives was associated with greater bone mineral density. Material and methods Subjects. An age-stratified sample of women was selected at random from electoral rolls for the Barwon Statistical Division, a geographically distinct area surrounding Geelong in southeastern Australia. Of the 1602 eligible women (age range, 20-92 years), 1291 agreed to participate. For this analysis, we excluded women aged >69 years (n = 220) and those exposed to diseases and other drugs known to influence bone mass (n = 361). Thus 710 women (age range, 20-69 years; mean age ± SD, 41.1 ± 13.8 years; 511 premenopausal, 172 postmenopausal, and 27 in whom menopausal status was indeterminate) were included in the analysis. Written informed consent was obtained from all participants. The study was approved by The Geelong Hospital Human Research and Ethics Advisory Committee. Measurements. Dual-energy x-ray absorptiometry was performed by a Lunar DPX-L densitometer, and the data were analyzed by Lunar DPX-L software version 1.31. Bone mineral density was measured at the spine (L2-4) in the posterior-anterior projection, proximal femur, whole body, ultradistal forearm, and midforearm sites. Short-term precision in vivo was 0.6% for the spine, 1.6% 265
266 Pasco et al
February 2000 Am J Obstet Gynecol
Table I. Characteristics and unadjusted bone mineral density (mean ± SD) of women exposed and unexposed to oral contraceptives All subjects Exposed Age (y) 38.9 ± 12.3* Weight (kg) 68.0 ± 14.0 Height (cm) 162.8 ± 6.1* Body mass index (kg/m2) 25.7 ± 5.3* Calcium (mg/d) 652 ± 353 Parity 2.0 ± 1.5 Lactation (mo) 10.2 ± 14.9 Alcohol (standard drinks/wk) 4.0 ± 6.0* Smoking (pack-years) 5.5 ± 11.5 Menopausal age (y) 46.8 ± 5.7* Active (No. and %) 493 (85%)* Bone mineral density (g/cm2) Spine 1.221 ± 0.157* (n = 579) Proximal femur Femoral neck 0.972 ± 0.139* (n = 577) Ward’s triangle 0.863 ± 0.164* (n = 577) Trochanter 0.827 ± 0.135* (n = 577) Whole body 1.164 ± 0.088* (n = 577) Ultradistal forearm 0.329 ± 0.051* (n = 573) Midforearm 0.703 ± 0.067* (n = 572)
Premenopausal subjects Unexposed
Exposed
Unexposed
50.6 ± 16.2* 69.9 ± 15.2 160.8 ± 6.3* 27.1 ± 6.1* 705 ± 380 2.3 ± 2.2 10.3 ± 14.8 2.5 ± 4.8* 4.4 ± 12.1 48.6 ± 5.2* 97 (74%)*
34.4 ± 8.2 67.4 ± 13.4 163.3 ± 5.9 25.3 ± 5.0 650 ± 337 1.7 ± 1.4 10.3 ± 15.0* 3.9 ± 5.1* 4.3 ± 8.7 — 405 (88%)
32.2 ± 10.5 70.3 ± 18.8 163.6 ± 5.5 26.3 ± 6.9 648 ± 347 1.3 ± 1.7 6.7 ± 11.5* 2.3 ± 3.9* 3.3 ± 7.4 — 41 (84%)
1.119 ± 0.189* (n = 131)
1.245 ± 0.134 (n = 462)
1.213 ± 0.164 (n = 49)
0.924 ± 0.162* (n = 131) 0.803 ± 0.198* (n = 131) 0.813 ± 0.153* (n = 131) 1.124 ± 0.110* (n = 131) 0.311 ± 0.061* (n = 131) 0.666 ± 0.085* (n = 131)
0.991 ± 0.126 (n = 461) 0.889 ± 0.148 (n = 461) 0.835 ± 0.125 (n = 461) 1.175 ± 0.075 (n = 461) 0.333 ± 0.046 (n = 459) 0.714 ± 0.055 (n = 458)
1.019 ± 0.151 (n = 49) 0.933 ± 0.187 (n = 49) 0.849 ± 0.161 (n = 49) 1.177 ± 0.092 (n = 49) 0.330 ± 0.038 (n = 49) 0.704 ± 0.054 (n = 49)
*P < .05 (significant difference).
for the femoral neck, 2.1% for Ward’s triangle, 1.6% for the trochanter, 0.4% for the whole body, 2.1% for the ultradistal forearm, and 1.1% for the midforearm sites. Height and weight were measured to the nearest 0.1 cm and 0.1 kg, respectively, and the body mass index was calculated in kilograms per square meter. Details of reproductive history and lifestyle factors were documented by questionnaire during an interview. Habitual dietary and supplemental calcium intake was estimated by using a calcium-specific food-frequency questionnaire. Cigarette smoking was calculated in pack-years (1 pack-year is equivalent to 20 cigarettes per day per year), whereas alcohol consumption was recorded as the number of standard drinks per week (1 standard drink is equivalent to approximately 10 g of alcohol). Current activity levels were classified as active if exercise was performed regularly; otherwise they were designated as sedentary. For the purposes of this analysis, women who had undergone hysterectomy and retained one or both ovaries were not included in the postmenopausal group. Statistics. Two-sample t tests were used to test for differences between women exposed and unexposed to oral contraceptives for continuous variables, and the χ2 test of independence was used to test for differences in discrete variables. Analysis of variance, followed by the Tukey test of multiple comparisons, was used to test for differences in unadjusted bone mineral density with duration of exposure to oral contraceptives. By use of bone mineral density as the dependent variable, covariates were identified by best-fitting regression techniques21 for each site.
Analysis was performed for the whole group and for premenopausal and postmenopausal women separately. Best models included adjustments for age, weight, height, or body mass index, with some sites requiring higher than linear adjustments for age, centered about the mean to avoid collinearity. Menopause status was not included for the whole group because its contribution was negligible. The influence on bone mineral density of discrete variables was determined by analysis of covariance, and continuous variables were analyzed by multiple regression techniques. All analyses were performed by the Minitab (release 11) software package. Results Analysis of the whole group. Women exposed to oral contraceptives (n = 579) were younger and taller, had a lower body mass index, and consumed more alcohol than unexposed women (n = 131). Reflecting the age distribution of the groups, unadjusted bone mineral density was greater for women exposed to oral contraceptives at all sites except the trochanter (Table I). There was a positive relationship between exposure to oral contraceptives and bone mineral density at the spine adjusted for body mass index and age (partial r2 = 0.009; P = .014), with women exposed to oral contraceptives having a 3.3% greater mean adjusted bone mineral density. Ages for first exposure to oral contraceptives ranged from 11 to 48 years, with a mean and SD of 21.1 ± 5.7 years. Age at first exposure to oral contraceptives had no detectable effect on bone mineral density. Among women exposed to oral contraceptives, 44% (n = 257)
Pasco et al 267
Volume 182, Number 2 Am J Obstet Gynecol
Postmenopausal subjects Exposed
Unexposed
59.7 ± 6.8* 70.8 ± 15.5 160.1 ± 6.1 27.6 ± 5.7 699 ± 418 3.2 ± 1.8 10.2 ± 14.9 3.1 ± 5.7 11.6 ± 19.7* 46.8 ± 5.7* 72 (77%)
62.1 ± 4.8* 69.3 ± 12.4 159.2 ± 6.3 27.5 ± 5.5 755 ± 396 3.0 ± 2.4 13.0 ± 16.4 2.8 ± 5.4 5.3 ± 14.5* 48.6 ± 5.2* 54 (69%)
1.094 ± 0.181 (n = 94)
1.060 ± 0.180 (n = 78)
0.877 ± 0.159 (n = 93) 0.739 ± 0.181 (n = 93) 0.785 ± 0.162 (n = 93) 1.103 ± 0.114 (n = 93) 0.300 ± 0.063 (n = 93) 0.644 ± 0.088 (n = 93)
0.867 ± 0.142 (n = 78) 0.723 ± 0.162 (n = 78) 0.789 ± 0.146 (n = 78) 1.089 ± 0.108 (n = 78) 0.299 ± 0.070 (n = 78) 0.642 ± 0.094 (n = 78)
had used them for <5 years, and 56% (n = 322) had used them for ≥5 years. There was a pattern of increasing bone mineral density (unadjusted) at the spine with increasing duration of exposure (Fig 1), which failed to reach significance with each increment. However, after we adjusted for age and body mass index, a comparison with unexposed women showed a 3.2% higher mean bone mineral density associated with exposure up to 5 years and a further modest but significant increase of 0.2% with ≥5 years of exposure (P < .05). There was no detectable influence of age at menarche, calcium or alcohol intake, smoking, activity levels, parity, or lactation on the association between exposure to oral contraceptives and bone mineral density. No association was detected between exposure to oral contraceptives and bone mineral density at other sites. Separate analyses of premenopausal and postmenopausal women. There was a longer period of lactation and a greater alcohol intake among premenopausal women exposed to oral contraceptives, but there were no differences in physical characteristics between the groups (Table I). No differences in unadjusted bone mineral density existed between exposed and unexposed premenopausal women. There was a positive association between exposure to oral contraceptives and bone mineral density adjusted for weight and height at the spine (partial r2 = 0.008; P < .05), with exposed women having a 3.3% greater adjusted mean bone mineral density. Thirtyseven percent (n = 170) of the premenopausal women exposed to oral contraceptives reported current use; however, current or past use of oral contraceptives was not a
Fig 1. Unadjusted bone mineral density (mean ± SE) at lumbar spine (L2-4) with increasing duration of exposure to oral contraceptives (asterisk, significant difference, P < .05). BMD, Bone mineral density.
significant predictor of bone mineral density. Within this subgroup, the protective effect associated with oral contraceptive exposure was detected at the spine alone. Among postmenopausal women, those exposed to oral contraceptives were younger, had a greater exposure to cigarette smoking, and had an earlier age at menopause (Table I). The association detected at the lumbar spine for the whole group failed to reach significance for both unadjusted (Table I) and adjusted bone mineral density within this subgroup. Comment We report results from a population-based study that indicate that women who have been exposed to oral contraceptives have a greater bone mineral density at the lumbar spine. Age at menarche, calcium intake, parity, lactation, habitual exercise, smoking, alcohol consumption, and age at first exposure had no influence on the association between exposure to oral contraceptives and bone mineral density. There was a pattern of increasing bone mineral density with increasing duration of exposure, with the greatest effects occurring during the first 5 years of exposure. Our regression models predict that greater bone mineral density may be sustained throughout the postmenopausal years. However, the association demonstrated in premenopausal women did not reach significance among postmenopausal women and may indicate that the positive effect of oral contraceptives diminishes with an increase in the number of years after cessation. Positive associations have been reported between oral contraceptive use and bone mineral density at the lumbar spine,3, 5, 6, 8 distal radius,2, 4, 6 proximal femur,5, 8 and the whole body.7 Lindsay et al3 observed among premenopausal women that exposure to oral contraceptives was associated with a 12% higher mean bone mineral
268 Pasco et al
density, whereas among postmenopausal women the difference failed to reach significance. They reported a linear relationship, with 1% higher mean vertebral bone mineral density associated with each year of exposure. Kleerekoper et al6 also suggested that the degree of protection to the skeleton was related to duration of exposure. Hreshchyshyn et al9 reported no difference in bone mineral density at the spine and hip with oral contraceptive use; however, exclusion of women with a history of fractures or oophorectomy may have excluded women with predominantly low bone mineral densities. In a prospective study of premenopausal women, Mais et al13 reported a nonsignificant trend for higher bone mineral density associated with the use of low-dose oral contraceptives. A small sample size and a short period of exposure may have limited the ability to detect small differences in this and other studies.10, 22 Hartard et al23 studied the effects of exercise in combination with exposure to oral contraceptives in a group of premenopausal women and found the highest bone mineral density in the group characterized by long-term exercise and shortterm exposure to oral contraceptives. No effects were noted for long-term exposure, irrespective of activity levels. We were unable to detect any influence of lifestyle factors on the association between oral contraceptive use and bone mineral density. In our study alcohol consumption and smoking were self-reported and possibly underestimated. Abstainers numbered 30% for alcohol and 57% for smoking. We observed a trend for greater exposure to smoking among older age strata in our cohort. However, smoking was not a significant confounder, despite a greater exposure among our postmenopausal women who had used oral contraceptives. Community-use estimates for Australia indicate that patterns of prescribing oral contraceptives have changed. Between 1990 and 1997, the proportion of fixed combinations (progestogen-estrogen) changed from 38% to 48%, sequential preparations from 54% to 42%, and progestogen-only preparations from 8% to 10%. Current data indicate that 10% of prescriptions for combined oral contraceptives in Australia are for high (50 µg) and 90% for low (30-40 µg) ethinyl estradiol contents (Commonwealth Department of Health and Family Services, Drug Utilisation Sub-Committee database, 1990-1997). In this study the use of oral contraceptives was self-reported, and although there may be good agreement between medical records and self-reported past medication use,24 43% of our subjects could not recall the brand of any oral contraceptives used. We were unable to identify women exposed to higher doses of estrogen and the small proportion (<10%) who had taken progestogen-only preparations. Therefore exposure in this study is broad-based, and the dose remains unspeci-
February 2000 Am J Obstet Gynecol
fied. Similar limitations may be responsible for some of the discrepancies in the literature because different dose responses in the skeleton have been suggested.25 In conclusion, in women exposed to oral contraceptives, we detected a modest but clinically significant higher bone mineral density, which would be expected to reduce fracture risk by approximately 20%.26 Whether the higher bone mineral density is the result of either higher peak bone mineral density or a reduction in bone loss before menopause or both of these will require further study. We thank B. Skoric for her assistance with scanning, interviewing subjects, and completing questionnaires and P. McManus and J. Dudley from the Commonwealth Department of Health and Family Services for supplying the community use estimates for oral contraceptives. REFERENCES
1. Felson DT, Zhang Y, Hannan MT, Kiel DP, Wilson PWF, Anderson JJ. The effect of postmenopausal estrogen therapy on bone density in elderly women. N Engl J Med 1993;329:1141-6. 2. Goldsmith NF, Johnston JO. Bone mineral: effects of oral contraceptives, pregnancy, and lactation. J Bone Joint Surg Am 1975;57:657-68. 3. Lindsay R, Tohme J, Kanders B. The effect of oral contraceptive use on vertebral bone mass in pre- and post-menopausal women. Contraception 1986;34:333-40. 4. Enzelsberger H, Metka M, Heytmanek G, Schurz B, Kurz Ch, Kusztrich M. Influence of oral contraceptive use on bone density in climacteric women. Maturitas 1988;9:375-8. 5. Stevenson JC, Lees B, Devenport M, Cust MP, Ganger KF. Determinants of bone density in normal women: risk factors for future osteoporosis? BMJ 1989;298:924-8. 6. Kleerekoper M, Brienza RS, Schultz LR, Johnson CC. Oral contraceptive use may protect against low bone mass. Arch Intern Med 1991;151:1971-6. 7. Recker RR, Davies KM, Hinders SM, Heaney RP, Stegrnan MR, Kimmel DB. Bone gain in young adult women. JAMA 1992;268:2403-8. 8. Kritz-Silverstein D, Barrett-Connor E. Bone mineral density in postmenopausal women as determined by prior oral contraceptive use. Am J Public Health 1993;83:100-2. 9. Hreshchyshyn MM, Hopkins A, Zylstra S, Anbar M. Associations of parity, breast-feeding, and birth control pills with lumbar spine and femoral neck bone densities. Am J Obstet Gynecol 1988;159:318-22. 10. Lloyd T, Buchanan JR, Ursino GR, Myers C, Woodward G, Halbert DR. Long-term oral contraceptive use does not affect trabecular bone density. Am J Obstet Gynecol 1989;160:402-4. 11. Mazess RB, Barden HS. Bone density in premenopausal women: effects of age, dietary intake, physical activity, smoking, and birth-control pills. Am J Clin Nutr 1991;53:132-42. 12. Murphy S, Khaw KT, Compston JE. Lack of relationship between hip and spine bone mineral density and oral contraceptive use. Eur J Clin Invest 1993;23:108-11. 13. Mais V, Fruzzetti F, Ajossa S, Paoletti AM, Guerriero S, Melis GB. Bone metabolism in young women taking a monophasic pill containing 20 mcg ethinylestradiol: a prospective study. Contraception 1993;48:445-52. 14. Haapasalo H, Kannus P, Sievanen H, Pasanen M, Uusi-Rasi K, Heinonen A, et al. Development of mass, density, and estimated mechanical characteristics of bones in Caucasian females. J Bone Miner Res 1996;11:1751-60. 15. Matkovic V, Jelic T, Wardlaw GM, Ilich JZ, Goel PK, Wright JK, et al. Timing of peak bone mass in Caucasian females and its im-
Volume 182, Number 2 Am J Obstet Gynecol
16.
17.
18.
19. 20.
plication for the prevention of osteoporosis. Inference from a cross-sectional model. J Clin Invest 1994;93:799-808. Riggs BL, Wahner HW, Melton LJ III, Richelson LS, Judd HL, Offord KP. Rates of bone loss in the appendicular and axial skeletons of women. Evidence of substantial vertebral bone loss before menopause. J Clin Invest 1986;77:1487-91. Baran D, Sorensen A, Grimes J, Lew R, Karellas A, Johnson B, et al. Dietary modification with dairy products for preventing vertebral bone loss in premenopausal women: a three-year prospective study. J Clin Endocrinol Metab 1990;70:264-70. Hedlund LR, Gallagher JC. The effect of age and menopause on bone mineral density of the proximal femur. J Bone Miner Res 1989;4:639-42. Duppe H, Gardsell P, Johnell O, Nilsson BE. Bone mineral content in women: trends of change. Osteoporos Int 1992;2:262-5. Seeman E, Wahner HW, Offord KP, Kumar R, Johnson WJ, Riggs BL. Differential effects of endocrine dysfunction on the axial and the appendicular skeleton. J Clin Invest 1982;69:1302-9.
Pasco et al 269
21. Ott R. An introduction to statistical methods and data analysis. 4th ed. California: Wadsworth, Inc; 1993. p. 647-733. 22. Sowers MR, Wallace RB, Lemke JH. Correlates of mid-radius bone density among postmenopausal women: a community study. Am J Clin Nutr 1985;41:1045-53. 23. Hartard M, Bottermann P, Bartenstein P, Jeschke D, Schwaiger M. Effects on bone mineral density of low-dosed oral contraceptives compared to and combined with physical activity. Contraception 1997;55:87-90. 24. Harlow SD, Linet MS. Agreement between questionnaire data and medical records. The evidence for accuracy of recall. Am J Epidemiol 1989;129:233-48. 25. Horsman A, Jones M, Francis R, Nordin C. The effect of estrogen dose on postmenopausal bone loss. N Engl J Med 1983;309:1405-7. 26. Melton LJ III, Atkinson EJ, O’Fallon WM, Wahner HW, Riggs BL. Long-term fracture prediction by bone mineral assessed at different skeletal sites. J Bone Miner Res 1993;8:1227-33.
Bound volumes available to subscribers Bound volumes of the American Journal of Obstetrics and Gynecology are available to subscribers (only) for the 2000 issues from the publisher, at a cost of $122.00 for domestic, $156.22 for Canada, and $146.00 for international for Vol. 182 (January-June) and Vol. 183 (July-December). Shipping charges are included. Each bound volume contains a subject and author index, and all advertising is removed. Copies are shipped within 60 days after publication of the last issue in the volume. The binding is durable buckram with the Journal name, volume number, and year stamped in gold on the spine. Payment must accompany all orders. Contact Mosby, Inc, Subscription Services, 11830 Westline Industrial Dr, St. Louis, MO 63146-3318, USA; phone (800)453-4351 or (314)453-4351. Subscriptions must be in force to qualify. Bound volumes are not available in place of a regular Journal subscription.