- Email: [email protected]
Leisure-time physical activity and rate of bone loss among peri- and postmenopausal women: a longitudinal study
Bone Vol. 29, No. 5 November 2001:442– 446
Leisure-time Physical Activity and Rate of Bone Loss Among Peri- and Postmenopausal Women: A Longitudinal Study ¨ GER, T. LAKKA, M. TUPPURAINEN, J. JURVELIN, and R. HONKANEN E. PUNTILA, H. KRO Research Institute of Public Health, University of Kuopio, Kuopio, Finland Departments of Surgery, Obstetrics and Gynecology, and Clinical Physiology and Nuclear Medicine, Kuopio University Hospital, Kuopio, Finland
(BMD) are mainly genetically determined.5,20,21,24 Peak bone mass is normally reached during the third decade. Subsequently, there is an inevitable loss of bone, which can lead to osteoporosis and osteoporotic fractures in later life.6 Several drugs have been developed to reduce bone loss.7,12,14,16,17 However, life style factors such as high dietary calcium intake and physical activity have also been found to prevent bone loss and thus osteoporosis.15,18,22,26 The positive effect of physical activity on the change in BMD and BMC among middle-aged and older women has been demonstrated in training trials2,3,11,26 and also in some cross-sectional studies.8 However, in most of these trials the study populations have been small and the physical loading relatively high. Therefore, we used a population-based study to investigate whether regular, moderate, leisure-time physical activity is associated with diminished axial bone loss among peri- and postmenopausal women.
We examined the association between continuous leisuretime physical activity and the change in bone mineral density (BMD) and bone mineral content (BMC) in a populationbased random sample of 1873 peri- and postmenopausal women. Leisure-time physical activities were registered with self-administered questionnaires in 1989 and 1994, and with an assisted questionnaire in 1995–1997. BMD and BMC were measured from lumbar vertebrae L2– 4 and left femoral neck using dual-energy X-ray absorptiometry (DXA) in 1989 – 1991 and 1994 –1997. During the average 5.6 year follow-up, annual loss of lumbar BMC was 124 mg (311 vs. 435 mg, p ⴝ 0.036) and annual loss of lumbar BMD was 1.22 mg/cm2 (4.15 vs. 5.37 mg/cm2, p ⴝ 0.21) smaller among women with regular (at least 1 h each week) weight-bearing leisure-time exercise compared with sedentary women. The advantage was even larger in women with walking or jogging as their only regular weight-bearing leisure-time exercise; that is, their annual loss of lumbar BMC was 180 mg (272 vs. 452 mg, p ⴝ 0.022), and annual loss of lumbar BMD was 2.78 mg/cm2 (2.96 vs. 5.74 mg/cm2, p ⴝ 0.029) smaller than in sedentary women. Continuous leisure-time physical activity did not have any association with loss of BMC or BMD in the femoral neck Physical activity during 12 months before the last bone densitometry was not associated with loss of BMC or BMD at any site. Our results suggest that regular weightbearing exercise diminishes lumbar bone loss, but might be ineffective in the prevention of femoral osteoporosis among peri- and early postmenopausal women. (Bone 29:442– 446; 2001) © 2001 by Elsevier Science Inc. All rights reserved.
Materials and Methods Subjects A postal enquiry with a question about willingness to participate in BMD measurement was mailed to 14,200 women, aged 47–56 years, living in the Kuopio province of eastern Finland, in May 1989. A total of 13,100 (92.8%) women responded to the enquiry, and 11,055 (84.4%) women reported their willingness to undergo BMD measurements. Of these women, 2025 were randomly selected for this study. The final study population consisted of 1873 women who participated in both baseline and follow-up measurements. Bone Densitometry
Key Words: Bone mineral content (BMC); Bone mineral density (BMD); Exercise; Leisure time; Menopause; Osteoporosis.
Bone mineral densitometry measurements were performed by trained personnel at the Kuopio University Hospital using dualenergy X-ray absorptiometry (DXA; Lunar DPX, Madison, WI). The long-term stability (coefficient of variation) of the instrument, as assisted by the regular phantom measurements (n ⫽ 60), during the study period, was 0.4% for BMD and 0.8% for BMC. The short-term repeatability of this method has been shown to be 0.9% for spine and 1.5% for femoral neck BMD measurements.12 BMD (grams per square centimeter) and BMC (grams) were measured from the lumbar spine (L2– 4) and from the left femoral neck during 1989 –1991 and 1994 –1997. In all, 1873 women underwent both bone densitometry study. A total of 247 (13.2%) lumbar and 91 (4.9%) femoral BMD measurement pairs were excluded because of marked osteophytes, spinal deformations, other artifacts, or incorrect measurement. In addition, 166 (8.1%) lumbar BMC measurement pairs were lost due to a a computer hardware error in the storage of baseline data.
Introduction Osteoporosis is an increasing health problem in all aging populations. In addition to human suffering, osteoporosis imposes high monetary costs on society, mainly in the form of hip fractures.4,19,27,28 This problem is especially notable among women whose longer life and rapid loss of bone after menopause make them more vulnerable to osteoporosis. Axial bone mineral content (BMC) and bone mineral density Address for correspondence and reprints: Dr. Eero Puntila, Osteoporosis Prevention Study Project, University of Kuopio, P.O. Box 1627, FIN70211 Kuopio, Finland. E-mail: [email protected] © 2001 by Elsevier Science Inc. All rights reserved.
442
8756-3282/01/$20.00 PII S8756-3282(01)00597-X
Bone Vol. 29, No. 5 November 2001:442– 446
E. Puntila et al. Leisure activity and bone loss in women
443
Table 1. Characteristics of the study populationa Regular leisure-time weight-bearing physical activity
Weight (kg) at baseline Height (cm) at baseline Age (years) Chronic diseases Years of HRT during follow-up Years since menopause to end of follow-up L2–4 BMD (mg/cm2) at baseline L2–4 BMC (g) at baseline Femoral neck BMD (mg/cm2) at baseline Femoral neck BMC (g) at baseline Use of HRT during the follow-up Smoking
Physically active (N ⫽ 527)
Sedentary (N ⫽ 231)
p
65.5 (64.6, 66.3) 161.4 (161.2, 162.0) 53.6 (53.3, 53.8) 65.0% 2.1 (1.9, 2.3) 9.0 (8.5, 9.6) 1112 (1104, 1128) 48.2 (47.3, 49.3) 920 (909, 931) 4.47 (4.4, 4.5) 57.3% 24.4%
72.4 (70.6, 74.2) 161.6 (160.7, 162.2) 53.3 (53.0, 53.7) 66.5% 1.4 (1.1, 1.7) 8.6 (7.8, 9.3) 1154 (1132, 1171) 50.6 (49.0, 52.3) 946 (929, 960) 4.61 (4.5, 4.7) 42.7% 21.1%
0.000 0.679 0.324 0.788 ⬍0.001 0.336 0.007 0.007 0.010 0.025 ⬍0.001 0.344
KEY: BMC, bone mineral content; BMD, bone mineral density; HRT, hormone replacement therapy. a Data presented as mean percentages (95% confidence interval).
Measurement of Leisure-time Physical Activity Self-administered questionnaires were mailed to the subjects in May 1989 and May 1994. Current (May 1989) regular conditioning leisure-time physical activity and duration (hours per week) was registered with the first questionnaire. Regular conditioning leisuretime physical activity in winter and in summer during the previous 12 months were registered with the second questionnaire in May 1994 (Appendix 1). Finally, leisure-time physical activity during the preceding 12 months was registered in detail with the Kuopio Ischemic Heart Disease Risk Factor Study Leisure-time Physical Activity Questionnaire, in connection with the second bone densitometry (Appendix 2). The validity and reliability of this questionnaire for a Finnish population have been reported earlier.13 Other Assessments Body weight, body height, menopausal status, chronic diseases, and use of hormone replacement therapy (HRT) were assessed in May 1989 and May 1994 with self-administered questionnaires. HRT and fractures during follow-up were also registered with an assisted questionnaire at the second bone densitometry. Statistical Analyses Statistical analyses were performed with SPSS software for WINDOWS 8.01. Student’s t-test was used for continuous variables and Pearson’s chi-square test for categorical background variables to test for differences between groups subdivided according to regular leisure-time physical activity. An analysis of covariance was used to test differences in changes of BMCs and BMDs between the groups subdivided according to regular leisure-time physical activity. Possible confounding factors (age, weight, time since menopause to second densitometry, duration of HRT during follow-up, smoking, and baseline BMC or BMD) were used as covariates. Statistical significance of interrelationships between continuous variables was tested using the significance test for Pearson’s correlation coefficient.
running, aerobics, gymnastics, rowing, ball games) physical activity per week during the last 12 months of follow-up (physically active women). In contrast, 231 (11.4%) women did not report any regular leisure-time physical activity in either 1989 or 1994 and ⬍1 h of leisure-time weight-bearing exercise per week at the end of follow-up (sedentary women). Among the physically active women, the mean weekly duration of weight-bearing exercise during the 12 months preceding the second bone densitometry was 2.9 h, with the most popular weight-bearing sports being walking/jogging (97.7%), skiing (52.9), gymnastics/aerobics (43.5%), and rowing (17.5%). Other characteristics of the study groups are presented in Table 1. Average follow-up time (between the bone densitometries) was 5.6 years (range 3.5– 8.0 years). The mean annual losses of BMC in lumbar vertebrae L2– 4 and left femoral neck in the whole study population were 330 mg (0.7% of the baseline BMC) and 20 mg (0.4% of the baseline BMC), respectively. The annual loss of lumbar BMC was 27% (p ⫽ 0.036) less in physically active women than in sedentary women, after adjusting for baseline age, body weight, number of chronic diseases, smoking, and baseline BMC, as well as time since menopause to May 1994 and duration of HRT during follow-up (Table 2). The annual loss of lumbar BMD was 23% (p ⫽ 0.21) less among physically active women than among sedentary women with the same adjustments (Table 2). The subgroup of physically active women who reported conditioning walking or jogging as their sole leisure-time weight-bearing exercise during the last 12 months of follow-up had an even smaller loss of adjusted lumbar BMC and BMD: 60% (p ⫽ 0.022) and 52% (p ⫽ 0.029), when compared with losses in sedentary women (Table 3). No significant difference in the loss of BMD or BMC in the femoral neck was found between physically active and sedentary women (Table 2). Furthermore, no dose-response relationship was found between loading or duration of the weight-bearing exercise during the 12 month period preceding the last bone densitometry and the loss of BMCs or BMDs at lumbar vertebrae L2– 4 or femoral neck when whole the sample of 1873 women was analyzed.
Results Altogether, 527 (26.0%) women reported at least 1 h of conditioning physical activity per week in 1989 and 1994 and at least 1 h of regular leisure-time weight-bearing (walking, jogging,
Discussion Physical loading is necessary for bone growth and maintenance although the mechanisms at the cellular level are unknown. Physical
444
E. Puntila et al. Leisure activity and bone loss in women
Bone Vol. 29, No. 5 November 2001:442– 446
Table 2. Mean adjusted annual changes (95% confidence intervals) in lumbar and femoral BMC and BMD according to regular (at least 1 h each week) weight-bearing leisure-time physical activity Regular weight-bearing leisure-time physical activity
a
Adjusted annual change in L2–4 BMC (mg) N Adjustedb annual change in L2–4 BMD (mg/cm2) N Adjusteda annual change in femoral neck BMC (mg) N Adjustedb annual change in femoral neck BMD (mg/cm2) N
Active
Sedentary
p
⫺311 (⫺371, ⫺252) 349 ⫺4.15 (⫺5.13, ⫺3.09) 444 ⫺18.3 (⫺22.9, ⫺13.1) 472 ⫺4.64 (⫺5.44, ⫺3.91) 472
⫺435 (⫺529, ⫺342) 151 ⫺5.37 (⫺7.03, ⫺3.80) 179 ⫺19.7 (⫺28.2, ⫺11.8) 185 ⫺3.92 (⫺5.07, ⫺2.74) 191
0.036 0.206 0.777 0.317
Refer to Table 1 for abbreviations. Adjusted for age, baseline body weight, time from menopause to second densitometry, chronic diseases at baseline, baseline smoking, duration of HRT during follow-up, and baseline BMC. b Adjusted for age, baseline body weight, time from menopause to second densitometry, chronic diseases at baseline, baseline smoking, duration of HRT during follow-up, and baseline BMD. a
loading seems to have the strongest inducing effect in fast-growing, prepubertal bones.10,25 Physical loading can also affect bone growth in adulthood and later life, although the effect is weaker than at prepuberty.3,26 Suitable sports for a favorable effect on bone include those that place sufficient stress on the bones, taking into consideration the natural physical limitations in different periods of life. The leisure-time physical activities assessed in our study were mainly weight-bearing. The amount of weight-bearing leisure-time exercise was relatively low and, in particular, the proportion of women who continued to participate in leisure-time physical activity throughout the study was small, which indicates the physically inactive trend in our western-style culture and community. Nonetheless, we did detect significant differences in loss of BMC in the lumbar vertebrae between the physically active and sedentary women. The lack of a dose-response relationship between weekly duration or intensity of weight-bearing leisure-time physical activity and loss of lumbar BMC was possibly due to the instability of sporting activities during the follow-up. Another explanation for our finding could be that there was a threshold for the beneficial effect of physical loading on lumbar spine. The effect of exercise on preventing femoral neck bone loss among women ⬎50 years is controversial.3,26 In one recent training study, high-impact exercise increased femoral neck BMD in premenopausal but not postmenopausal women.2 Our study population consisted of mainly postmenopausal women with quite low-impact exercise, which might be one explanation
for the small differences in loss of femoral neck BMD and BMC between physically active and sedentary women. It is also likely that continuous weight-bearing loading at a certain stable level does not have the same positive response on bone as an increase in loading from an earlier lower level up to a higher level. Previous studies concerning habitual leisure-time physical activity and bone loss have been cross-sectional or training trials. The difference or change in BMD or BMC in those trials has been at least partly due to the short-term effect of increased physical activity. Population-based follow-up studies examining the association between continuous leisure-time physical activity and bone loss have not been performed. From public health viewpoint these would be of most interest. The link between walking or jogging and diminished lumbar bone loss found in our study is important because of the popularity and suitability of these sports for most women. Although extensive overlapping of different sports activities makes it difficult to estimate an association between other specific sports and bone loss, it is possible, even probable, that some less popular and increased-load weightbearing sports such as aerobics have a more preventive effect on bone loss, perhaps also in the femoral neck, although we were unable to detect such an effect in our study. The differences in loss of lumbar BMD between physically active and sedentary women found in our study were smaller compared with those in loss of lumbar BMC. Shortening of the lumbar spine is caused mainly by loss in height of vertebrae and less
Table 3. Mean adjusted annual changes (95% confidence intervals) in lumbar and femoral BMC and BMD according to regular (at least 1 h each week) weight-bearing leisure-time physical activity (walking or jogging only) Regular leisure-time walking or jogging
Adjusteda annual change in L2–4 BMC (mg) N Adjustedb annual change in L2–4 BMD (mg/cm2) N Adjusteda annual change in femoral neck BMC (mg) N Adjustedb annual change in femoral neck BMD (mg/cm2) N
Yes
No
p
⫺272 (⫺392, ⫺149) 101 ⫺2.96 (⫺4.79, ⫺1.14) 126 ⫺19.9 (⫺30.1, ⫺9.8) 133 ⫺5.14 (⫺6.63, ⫺3.61) 133
⫺452 (⫺554, ⫺361) 151 ⫺5.74 (⫺7.32, ⫺4.18) 179 ⫺21.8 (⫺29.9, ⫺13.8) 185 ⫺4.49 (⫺5.74, ⫺3.20) 185
0.022 0.029 0.778 0.525
Refer to Table 1 for abbreviations. Adjusted for age, baseline body weight, time from menopause to second densitometry, chronic diseases at baseline, baseline smoking, duration of HRT during follow-up, and baseline BMC. b Adjusted for age, baseline body weight, time from menopause to second densitometry, chronic diseases at baseline, baseline smoking, duration of HRT during follow-up, and baseline BMD. a
Bone Vol. 29, No. 5 November 2001:442– 446
E. Puntila et al. Leisure activity and bone loss in women
by degeneration and shallowing of intervertebral disks.9,23 Due to the measuring protocol of lumbar DXA scans (intervertebral spaces included), the individual shortening of the lumbar spine and reduction of the measured anteroposterior area include a source of error in the follow-up lumbar BMD measurements, which may explain some of the aforementioned differences. Thus, due to its greater accuracy, reporting of changes in BMC instead of BMD seems to be preferable in future prospective studies investigating lumbar osteoporosis and DXA, as has been suggested earlier.1 The use of self-administered questionnaires includes a risk of misestimation that cannot be entirely avoided, even with an assisted questionnaire. With leisure-time physical activity, direct measurement is still too complicated in population studies, so the use of self-administered and assisted questionnaires in measuring leisuretime physical activity can be justified. In our study, physically active women were more slender and more often used HRT than sedentary women. Even with these adjustments, the differences indicate a possible inequality in overall lifestyle between our study groups. This type of inequality cannot be completely adjusted and thus may have had an impact on our results. The results of the present population-based follow-up study in peri- and postmenopausal women are important from the public health perspective. It was shown that as little as 1–3 h/week of leisure-time weight-bearing physical activity appears to be sufficient to diminish lumbar bone loss, but light-to-moderate physical activity seems to be ineffective in the prevention of femoral bone loss. The optimal physical activities appear to be brisk walking or jogging.
445
Therefore, regular weight-bearing physical activity may be recommended to prevent osteoporosis in peri- and postmenopausal women. Nevertheless, further studies of weight-bearing physical activity and femoral bone loss are clearly needed.
Acknowledgments: This study was partially supported by the Yrjo¨ Jahnsson Foundation, the Juho Vainio Foundation, and the Academy of Finland.
Appendix 1 Questions concerning leisure-time physical activity 1989 and 1994: May 1989: 11. Do you regularly take conditioning exercise? 1. No 2. Yes, average
hours per week
May 1994: 28. Did you regularly take conditioning leisure-time exercise during the past 12 months? 1. No hours and minutes per week in winter 2. Yes, average and hours and minutes per week in summer
Appendix 2 Leisure-time physical activity questions in connection with the second bone densitometry:
15. Which of the following exercises have you undertaken during the last 12 months?
How many times per month? Type of activity Walking to work Conditioning walking Jogging, running, orienteering Skiing Bicycling Bicycling to work Swimming Gymnastics, aerobic Ball games Gardening, snow-shoveling Hunting, picking berries, mushroom gathering Fishing Hobby crafts Rowing Tree-chopping Downhill skiing Skating Indoor keep-fit training Dance Householding, cleaning Other, what?
Jan.
Feb.
Mar.
Apr.
May
Jun.
Jul.
Aug.
Sep.
Oct.
Nov.
Dec.
Average duration per occasion (h, min)
Intensity (0 ⫽ light 1 ⫽ medium 2 ⫽ strenuous 3 ⫽ very strenuous)
446
E. Puntila et al. Leisure activity and bone loss in women
References 1. Brahm, H., Mallmin, H., Michaelsson, K., Strom, H., and Ljunghall, S. Relationships between bone mass measurements and lifetime physical activity in a Swedish population. Calcif Tissue Int 62:400 – 412; 1998. 2. Bassey, E., Rothwell, M., Littlewood, J., and Pye D. Pre- and postmenopausal women have different bone mineral density responses to the same high-impact exercise. J Bone Miner Res 13:1085–1813; 1998. 3. Be´ rard, A., Bravo, G., and Gauthier, P. Meta-analysis of the effectiveness of physical activity for the prevention of bone loss in postmenopausal women. Osteopor Int 7:331–337; 1997. 4. Brainsky, A., Glick, H., Lydick, E., Epstein, R., Fox, K. M., Hawkes, W., Kashner, M., Zimmerman, S. I. and Magaziner, J. The economic cost of hip Lifetime leisure exercise and osteoporosis. The Rancho Bernardo study. Am J Epidemiol 141:951–959; 1995. fractures in community-dwelling older adults: A prospective study. J Am Geriatr Soc 45:281–287; 1997. 5. Dequecker, J., Nijs, J., Verstraeten, A., Geusens, P., and Gevers, G. Genetic determinants of bone mineral content at the spine and radius: A twin study. Bone 8:207–209; 1987. 6. Du¨ ppe, H., Ga¨ rdsell, P., Nilsson, B., and Johnell, O. A. single bone density measurement can predict fractures over 25 years. Calcif Tissue Int 60:171–174; 1997. 7. Eastell, R. Treatment of postmenopausal osteoporosis. N Engl J Med 338:736 – 746; 1998. 8. Greendale, G., Barrett-Connor, E., Edelstein, S., Ingles, S., and Haile, R. 9. Holm, S. Pathophysiology of disc degeneration. Acta Orthop Scand 251(Suppl.):13–15; 1993. 10. Kannus, P., Haapasalo, H., Sankelo, M., Sieva¨ nen, H., Pasanen, M., Heinonen, A., Oja, P., and Vuori, I. Effect of starting age of physical activity on bone mass in the dominant arm of tennis and squash players. Ann Intern Med 123:27–31; 1995. 11. Krall, E. and Dawson-Hughes, B. Walking is related to bone density and rates of bone loss. Am J Med 96:20 –26; 1994. 12. Kro¨ ger, H., Tuppurainen, M., Honkanen, R., Alhava, E., and Saarikoski, S. Bone mineral density and risk factors for osteoporosis—A population-based study of 1600 perimenopausal women. Calcif Tissue Int 55:1–7; 1994. 13. Lakka, T. A. and Salonen, J. T. Intra-person variability of various physical activity assessments in the Kuopio Ischaemic Heart Disease Risk Factor Study. Int J Epidemiol 21:467– 472; 1992. 14. McClung, M. R. Therapy for fracture prevention. JAMA 282:687– 689; 1999. 15. Michaelsson, K., Bergstro¨ m, R., Holmberg, L., Mallmin, H., Wolk, A., and Ljunghall, S. A high dietary calcium intake is needed for a positive effect on bone density in Swedish postmenopausal women. Osteopor Int 7:155–161; 1997.
Bone Vol. 29, No. 5 November 2001:442– 446 16. Lane, J. M. Osteoporosis. Medical prevention and treatment. Spine 22(Suppl.): 32S–37S; 1997. 17. Netelenbos, C. Osteoporosis: Intervention options. Maturitas 30:235–239; 1998. 18. Prince, R. L., Smith, M., Dick, I. M., et al. Prevention of postmenopausal osteoporosis. N Engl J Med 325:1189 –1195; 1991. 19. Reginster, J. Y., Gillet, P., Ben Sedrine, W., Brands, G., Ethgen, O., De Froidmont, C., and Gosset, C. Direct costs of hip fractures inpatients over 60 years of age in Belgium. Pharmacoeconomics 15:507–514; 1999. 20. Rubin, L., Hawker, G., Peltekova, V., Fielding, L., Ridout, R., and Cole, D. Determinants of peak bone mass: Clinical and genetic analyses in a female Canadian cohort. J Bone Miner Res 14:633– 643; 1999. 21. Slemenda, W., et al. Genetic determinants of bone mass in adult women: A reevaluation of the twin model and the potential importance of gene interaction on heritability estimates. J Bone Miner Res 6:561–567; 1991. 22. Snow-Harter, C. and Marcus, R. Exercise, bone mineral density, and osteoporosis. Exerc Sport Sci Rev 19:351–388; 1991. 23. Twomey, L. T. and Taylor, J. R. Age changes in lumbar vertebrae and intervertebral discs. Clin Orthop 224:97–104; 1987. 24. Ulrich, C. M., Georgiou, C. C., Snow-Harter, C. M., and Gillis, D. E. Bone mineral density in mother– daughter pairs: Relations to lifetime exercise, lifetime milk consumption, and calcium supplements. Am J Clin Nutr 63:72– 79; 1996. 25. Welten, D. C., Kemper, H. C. G., Post, G. B., Van Mechelen, W., Twisk, J., Lips, P., and Teule, G. J. Weight-bearing activity during youth is a more important factor for peak bone mass than calcium intake. J Bone Miner Res 9:1089 –1095; 1994. 26. Wolff, I., van Croonenborg, J. J., Kemper, H. C. G., Kostense, P. J., and Twisk, J. W. R. The effect of exercise training programs on bone mass: A metaanalysis of published controlled trials in pre- and postmenopausal women. Osteopor Int 9:1–12; 1999. 27. Youm, T., Koyal, K. J., and Zuckerman, J. D. The economic impact of geriatric hip fractures. Am J Orthop 28:423– 428; 1999. ¨ hlen, G. The 28. Zethraeus, N., Stro¨ mberg, L., Jo¨ nsson, B., Svensson, O., and O cost of a hip fracture. Estimates for 1,709 patients in Sweden. Acta Orthop Scand 68:13–17; 1997.
Date Received: June 23, 2000 Date Revised: May 10, 2001 Date Accepted: May 15, 2001
Leisure-time Physical Activity and Rate of Bone Loss Among Peri- and Postmenopausal Women: A Longitudinal Study ¨ GER, T. LAKKA, M. TUPPURAINEN, J. JURVELIN, and R. HONKANEN E. PUNTILA, H. KRO Research Institute of Public Health, University of Kuopio, Kuopio, Finland Departments of Surgery, Obstetrics and Gynecology, and Clinical Physiology and Nuclear Medicine, Kuopio University Hospital, Kuopio, Finland
(BMD) are mainly genetically determined.5,20,21,24 Peak bone mass is normally reached during the third decade. Subsequently, there is an inevitable loss of bone, which can lead to osteoporosis and osteoporotic fractures in later life.6 Several drugs have been developed to reduce bone loss.7,12,14,16,17 However, life style factors such as high dietary calcium intake and physical activity have also been found to prevent bone loss and thus osteoporosis.15,18,22,26 The positive effect of physical activity on the change in BMD and BMC among middle-aged and older women has been demonstrated in training trials2,3,11,26 and also in some cross-sectional studies.8 However, in most of these trials the study populations have been small and the physical loading relatively high. Therefore, we used a population-based study to investigate whether regular, moderate, leisure-time physical activity is associated with diminished axial bone loss among peri- and postmenopausal women.
We examined the association between continuous leisuretime physical activity and the change in bone mineral density (BMD) and bone mineral content (BMC) in a populationbased random sample of 1873 peri- and postmenopausal women. Leisure-time physical activities were registered with self-administered questionnaires in 1989 and 1994, and with an assisted questionnaire in 1995–1997. BMD and BMC were measured from lumbar vertebrae L2– 4 and left femoral neck using dual-energy X-ray absorptiometry (DXA) in 1989 – 1991 and 1994 –1997. During the average 5.6 year follow-up, annual loss of lumbar BMC was 124 mg (311 vs. 435 mg, p ⴝ 0.036) and annual loss of lumbar BMD was 1.22 mg/cm2 (4.15 vs. 5.37 mg/cm2, p ⴝ 0.21) smaller among women with regular (at least 1 h each week) weight-bearing leisure-time exercise compared with sedentary women. The advantage was even larger in women with walking or jogging as their only regular weight-bearing leisure-time exercise; that is, their annual loss of lumbar BMC was 180 mg (272 vs. 452 mg, p ⴝ 0.022), and annual loss of lumbar BMD was 2.78 mg/cm2 (2.96 vs. 5.74 mg/cm2, p ⴝ 0.029) smaller than in sedentary women. Continuous leisure-time physical activity did not have any association with loss of BMC or BMD in the femoral neck Physical activity during 12 months before the last bone densitometry was not associated with loss of BMC or BMD at any site. Our results suggest that regular weightbearing exercise diminishes lumbar bone loss, but might be ineffective in the prevention of femoral osteoporosis among peri- and early postmenopausal women. (Bone 29:442– 446; 2001) © 2001 by Elsevier Science Inc. All rights reserved.
Materials and Methods Subjects A postal enquiry with a question about willingness to participate in BMD measurement was mailed to 14,200 women, aged 47–56 years, living in the Kuopio province of eastern Finland, in May 1989. A total of 13,100 (92.8%) women responded to the enquiry, and 11,055 (84.4%) women reported their willingness to undergo BMD measurements. Of these women, 2025 were randomly selected for this study. The final study population consisted of 1873 women who participated in both baseline and follow-up measurements. Bone Densitometry
Key Words: Bone mineral content (BMC); Bone mineral density (BMD); Exercise; Leisure time; Menopause; Osteoporosis.
Bone mineral densitometry measurements were performed by trained personnel at the Kuopio University Hospital using dualenergy X-ray absorptiometry (DXA; Lunar DPX, Madison, WI). The long-term stability (coefficient of variation) of the instrument, as assisted by the regular phantom measurements (n ⫽ 60), during the study period, was 0.4% for BMD and 0.8% for BMC. The short-term repeatability of this method has been shown to be 0.9% for spine and 1.5% for femoral neck BMD measurements.12 BMD (grams per square centimeter) and BMC (grams) were measured from the lumbar spine (L2– 4) and from the left femoral neck during 1989 –1991 and 1994 –1997. In all, 1873 women underwent both bone densitometry study. A total of 247 (13.2%) lumbar and 91 (4.9%) femoral BMD measurement pairs were excluded because of marked osteophytes, spinal deformations, other artifacts, or incorrect measurement. In addition, 166 (8.1%) lumbar BMC measurement pairs were lost due to a a computer hardware error in the storage of baseline data.
Introduction Osteoporosis is an increasing health problem in all aging populations. In addition to human suffering, osteoporosis imposes high monetary costs on society, mainly in the form of hip fractures.4,19,27,28 This problem is especially notable among women whose longer life and rapid loss of bone after menopause make them more vulnerable to osteoporosis. Axial bone mineral content (BMC) and bone mineral density Address for correspondence and reprints: Dr. Eero Puntila, Osteoporosis Prevention Study Project, University of Kuopio, P.O. Box 1627, FIN70211 Kuopio, Finland. E-mail: [email protected] © 2001 by Elsevier Science Inc. All rights reserved.
442
8756-3282/01/$20.00 PII S8756-3282(01)00597-X
Bone Vol. 29, No. 5 November 2001:442– 446
E. Puntila et al. Leisure activity and bone loss in women
443
Table 1. Characteristics of the study populationa Regular leisure-time weight-bearing physical activity
Weight (kg) at baseline Height (cm) at baseline Age (years) Chronic diseases Years of HRT during follow-up Years since menopause to end of follow-up L2–4 BMD (mg/cm2) at baseline L2–4 BMC (g) at baseline Femoral neck BMD (mg/cm2) at baseline Femoral neck BMC (g) at baseline Use of HRT during the follow-up Smoking
Physically active (N ⫽ 527)
Sedentary (N ⫽ 231)
p
65.5 (64.6, 66.3) 161.4 (161.2, 162.0) 53.6 (53.3, 53.8) 65.0% 2.1 (1.9, 2.3) 9.0 (8.5, 9.6) 1112 (1104, 1128) 48.2 (47.3, 49.3) 920 (909, 931) 4.47 (4.4, 4.5) 57.3% 24.4%
72.4 (70.6, 74.2) 161.6 (160.7, 162.2) 53.3 (53.0, 53.7) 66.5% 1.4 (1.1, 1.7) 8.6 (7.8, 9.3) 1154 (1132, 1171) 50.6 (49.0, 52.3) 946 (929, 960) 4.61 (4.5, 4.7) 42.7% 21.1%
0.000 0.679 0.324 0.788 ⬍0.001 0.336 0.007 0.007 0.010 0.025 ⬍0.001 0.344
KEY: BMC, bone mineral content; BMD, bone mineral density; HRT, hormone replacement therapy. a Data presented as mean percentages (95% confidence interval).
Measurement of Leisure-time Physical Activity Self-administered questionnaires were mailed to the subjects in May 1989 and May 1994. Current (May 1989) regular conditioning leisure-time physical activity and duration (hours per week) was registered with the first questionnaire. Regular conditioning leisuretime physical activity in winter and in summer during the previous 12 months were registered with the second questionnaire in May 1994 (Appendix 1). Finally, leisure-time physical activity during the preceding 12 months was registered in detail with the Kuopio Ischemic Heart Disease Risk Factor Study Leisure-time Physical Activity Questionnaire, in connection with the second bone densitometry (Appendix 2). The validity and reliability of this questionnaire for a Finnish population have been reported earlier.13 Other Assessments Body weight, body height, menopausal status, chronic diseases, and use of hormone replacement therapy (HRT) were assessed in May 1989 and May 1994 with self-administered questionnaires. HRT and fractures during follow-up were also registered with an assisted questionnaire at the second bone densitometry. Statistical Analyses Statistical analyses were performed with SPSS software for WINDOWS 8.01. Student’s t-test was used for continuous variables and Pearson’s chi-square test for categorical background variables to test for differences between groups subdivided according to regular leisure-time physical activity. An analysis of covariance was used to test differences in changes of BMCs and BMDs between the groups subdivided according to regular leisure-time physical activity. Possible confounding factors (age, weight, time since menopause to second densitometry, duration of HRT during follow-up, smoking, and baseline BMC or BMD) were used as covariates. Statistical significance of interrelationships between continuous variables was tested using the significance test for Pearson’s correlation coefficient.
running, aerobics, gymnastics, rowing, ball games) physical activity per week during the last 12 months of follow-up (physically active women). In contrast, 231 (11.4%) women did not report any regular leisure-time physical activity in either 1989 or 1994 and ⬍1 h of leisure-time weight-bearing exercise per week at the end of follow-up (sedentary women). Among the physically active women, the mean weekly duration of weight-bearing exercise during the 12 months preceding the second bone densitometry was 2.9 h, with the most popular weight-bearing sports being walking/jogging (97.7%), skiing (52.9), gymnastics/aerobics (43.5%), and rowing (17.5%). Other characteristics of the study groups are presented in Table 1. Average follow-up time (between the bone densitometries) was 5.6 years (range 3.5– 8.0 years). The mean annual losses of BMC in lumbar vertebrae L2– 4 and left femoral neck in the whole study population were 330 mg (0.7% of the baseline BMC) and 20 mg (0.4% of the baseline BMC), respectively. The annual loss of lumbar BMC was 27% (p ⫽ 0.036) less in physically active women than in sedentary women, after adjusting for baseline age, body weight, number of chronic diseases, smoking, and baseline BMC, as well as time since menopause to May 1994 and duration of HRT during follow-up (Table 2). The annual loss of lumbar BMD was 23% (p ⫽ 0.21) less among physically active women than among sedentary women with the same adjustments (Table 2). The subgroup of physically active women who reported conditioning walking or jogging as their sole leisure-time weight-bearing exercise during the last 12 months of follow-up had an even smaller loss of adjusted lumbar BMC and BMD: 60% (p ⫽ 0.022) and 52% (p ⫽ 0.029), when compared with losses in sedentary women (Table 3). No significant difference in the loss of BMD or BMC in the femoral neck was found between physically active and sedentary women (Table 2). Furthermore, no dose-response relationship was found between loading or duration of the weight-bearing exercise during the 12 month period preceding the last bone densitometry and the loss of BMCs or BMDs at lumbar vertebrae L2– 4 or femoral neck when whole the sample of 1873 women was analyzed.
Results Altogether, 527 (26.0%) women reported at least 1 h of conditioning physical activity per week in 1989 and 1994 and at least 1 h of regular leisure-time weight-bearing (walking, jogging,
Discussion Physical loading is necessary for bone growth and maintenance although the mechanisms at the cellular level are unknown. Physical
444
E. Puntila et al. Leisure activity and bone loss in women
Bone Vol. 29, No. 5 November 2001:442– 446
Table 2. Mean adjusted annual changes (95% confidence intervals) in lumbar and femoral BMC and BMD according to regular (at least 1 h each week) weight-bearing leisure-time physical activity Regular weight-bearing leisure-time physical activity
a
Adjusted annual change in L2–4 BMC (mg) N Adjustedb annual change in L2–4 BMD (mg/cm2) N Adjusteda annual change in femoral neck BMC (mg) N Adjustedb annual change in femoral neck BMD (mg/cm2) N
Active
Sedentary
p
⫺311 (⫺371, ⫺252) 349 ⫺4.15 (⫺5.13, ⫺3.09) 444 ⫺18.3 (⫺22.9, ⫺13.1) 472 ⫺4.64 (⫺5.44, ⫺3.91) 472
⫺435 (⫺529, ⫺342) 151 ⫺5.37 (⫺7.03, ⫺3.80) 179 ⫺19.7 (⫺28.2, ⫺11.8) 185 ⫺3.92 (⫺5.07, ⫺2.74) 191
0.036 0.206 0.777 0.317
Refer to Table 1 for abbreviations. Adjusted for age, baseline body weight, time from menopause to second densitometry, chronic diseases at baseline, baseline smoking, duration of HRT during follow-up, and baseline BMC. b Adjusted for age, baseline body weight, time from menopause to second densitometry, chronic diseases at baseline, baseline smoking, duration of HRT during follow-up, and baseline BMD. a
loading seems to have the strongest inducing effect in fast-growing, prepubertal bones.10,25 Physical loading can also affect bone growth in adulthood and later life, although the effect is weaker than at prepuberty.3,26 Suitable sports for a favorable effect on bone include those that place sufficient stress on the bones, taking into consideration the natural physical limitations in different periods of life. The leisure-time physical activities assessed in our study were mainly weight-bearing. The amount of weight-bearing leisure-time exercise was relatively low and, in particular, the proportion of women who continued to participate in leisure-time physical activity throughout the study was small, which indicates the physically inactive trend in our western-style culture and community. Nonetheless, we did detect significant differences in loss of BMC in the lumbar vertebrae between the physically active and sedentary women. The lack of a dose-response relationship between weekly duration or intensity of weight-bearing leisure-time physical activity and loss of lumbar BMC was possibly due to the instability of sporting activities during the follow-up. Another explanation for our finding could be that there was a threshold for the beneficial effect of physical loading on lumbar spine. The effect of exercise on preventing femoral neck bone loss among women ⬎50 years is controversial.3,26 In one recent training study, high-impact exercise increased femoral neck BMD in premenopausal but not postmenopausal women.2 Our study population consisted of mainly postmenopausal women with quite low-impact exercise, which might be one explanation
for the small differences in loss of femoral neck BMD and BMC between physically active and sedentary women. It is also likely that continuous weight-bearing loading at a certain stable level does not have the same positive response on bone as an increase in loading from an earlier lower level up to a higher level. Previous studies concerning habitual leisure-time physical activity and bone loss have been cross-sectional or training trials. The difference or change in BMD or BMC in those trials has been at least partly due to the short-term effect of increased physical activity. Population-based follow-up studies examining the association between continuous leisure-time physical activity and bone loss have not been performed. From public health viewpoint these would be of most interest. The link between walking or jogging and diminished lumbar bone loss found in our study is important because of the popularity and suitability of these sports for most women. Although extensive overlapping of different sports activities makes it difficult to estimate an association between other specific sports and bone loss, it is possible, even probable, that some less popular and increased-load weightbearing sports such as aerobics have a more preventive effect on bone loss, perhaps also in the femoral neck, although we were unable to detect such an effect in our study. The differences in loss of lumbar BMD between physically active and sedentary women found in our study were smaller compared with those in loss of lumbar BMC. Shortening of the lumbar spine is caused mainly by loss in height of vertebrae and less
Table 3. Mean adjusted annual changes (95% confidence intervals) in lumbar and femoral BMC and BMD according to regular (at least 1 h each week) weight-bearing leisure-time physical activity (walking or jogging only) Regular leisure-time walking or jogging
Adjusteda annual change in L2–4 BMC (mg) N Adjustedb annual change in L2–4 BMD (mg/cm2) N Adjusteda annual change in femoral neck BMC (mg) N Adjustedb annual change in femoral neck BMD (mg/cm2) N
Yes
No
p
⫺272 (⫺392, ⫺149) 101 ⫺2.96 (⫺4.79, ⫺1.14) 126 ⫺19.9 (⫺30.1, ⫺9.8) 133 ⫺5.14 (⫺6.63, ⫺3.61) 133
⫺452 (⫺554, ⫺361) 151 ⫺5.74 (⫺7.32, ⫺4.18) 179 ⫺21.8 (⫺29.9, ⫺13.8) 185 ⫺4.49 (⫺5.74, ⫺3.20) 185
0.022 0.029 0.778 0.525
Refer to Table 1 for abbreviations. Adjusted for age, baseline body weight, time from menopause to second densitometry, chronic diseases at baseline, baseline smoking, duration of HRT during follow-up, and baseline BMC. b Adjusted for age, baseline body weight, time from menopause to second densitometry, chronic diseases at baseline, baseline smoking, duration of HRT during follow-up, and baseline BMD. a
Bone Vol. 29, No. 5 November 2001:442– 446
E. Puntila et al. Leisure activity and bone loss in women
by degeneration and shallowing of intervertebral disks.9,23 Due to the measuring protocol of lumbar DXA scans (intervertebral spaces included), the individual shortening of the lumbar spine and reduction of the measured anteroposterior area include a source of error in the follow-up lumbar BMD measurements, which may explain some of the aforementioned differences. Thus, due to its greater accuracy, reporting of changes in BMC instead of BMD seems to be preferable in future prospective studies investigating lumbar osteoporosis and DXA, as has been suggested earlier.1 The use of self-administered questionnaires includes a risk of misestimation that cannot be entirely avoided, even with an assisted questionnaire. With leisure-time physical activity, direct measurement is still too complicated in population studies, so the use of self-administered and assisted questionnaires in measuring leisuretime physical activity can be justified. In our study, physically active women were more slender and more often used HRT than sedentary women. Even with these adjustments, the differences indicate a possible inequality in overall lifestyle between our study groups. This type of inequality cannot be completely adjusted and thus may have had an impact on our results. The results of the present population-based follow-up study in peri- and postmenopausal women are important from the public health perspective. It was shown that as little as 1–3 h/week of leisure-time weight-bearing physical activity appears to be sufficient to diminish lumbar bone loss, but light-to-moderate physical activity seems to be ineffective in the prevention of femoral bone loss. The optimal physical activities appear to be brisk walking or jogging.
445
Therefore, regular weight-bearing physical activity may be recommended to prevent osteoporosis in peri- and postmenopausal women. Nevertheless, further studies of weight-bearing physical activity and femoral bone loss are clearly needed.
Acknowledgments: This study was partially supported by the Yrjo¨ Jahnsson Foundation, the Juho Vainio Foundation, and the Academy of Finland.
Appendix 1 Questions concerning leisure-time physical activity 1989 and 1994: May 1989: 11. Do you regularly take conditioning exercise? 1. No 2. Yes, average
hours per week
May 1994: 28. Did you regularly take conditioning leisure-time exercise during the past 12 months? 1. No hours and minutes per week in winter 2. Yes, average and hours and minutes per week in summer
Appendix 2 Leisure-time physical activity questions in connection with the second bone densitometry:
15. Which of the following exercises have you undertaken during the last 12 months?
How many times per month? Type of activity Walking to work Conditioning walking Jogging, running, orienteering Skiing Bicycling Bicycling to work Swimming Gymnastics, aerobic Ball games Gardening, snow-shoveling Hunting, picking berries, mushroom gathering Fishing Hobby crafts Rowing Tree-chopping Downhill skiing Skating Indoor keep-fit training Dance Householding, cleaning Other, what?
Jan.
Feb.
Mar.
Apr.
May
Jun.
Jul.
Aug.
Sep.
Oct.
Nov.
Dec.
Average duration per occasion (h, min)
Intensity (0 ⫽ light 1 ⫽ medium 2 ⫽ strenuous 3 ⫽ very strenuous)
446
E. Puntila et al. Leisure activity and bone loss in women
References 1. Brahm, H., Mallmin, H., Michaelsson, K., Strom, H., and Ljunghall, S. Relationships between bone mass measurements and lifetime physical activity in a Swedish population. Calcif Tissue Int 62:400 – 412; 1998. 2. Bassey, E., Rothwell, M., Littlewood, J., and Pye D. Pre- and postmenopausal women have different bone mineral density responses to the same high-impact exercise. J Bone Miner Res 13:1085–1813; 1998. 3. Be´ rard, A., Bravo, G., and Gauthier, P. Meta-analysis of the effectiveness of physical activity for the prevention of bone loss in postmenopausal women. Osteopor Int 7:331–337; 1997. 4. Brainsky, A., Glick, H., Lydick, E., Epstein, R., Fox, K. M., Hawkes, W., Kashner, M., Zimmerman, S. I. and Magaziner, J. The economic cost of hip Lifetime leisure exercise and osteoporosis. The Rancho Bernardo study. Am J Epidemiol 141:951–959; 1995. fractures in community-dwelling older adults: A prospective study. J Am Geriatr Soc 45:281–287; 1997. 5. Dequecker, J., Nijs, J., Verstraeten, A., Geusens, P., and Gevers, G. Genetic determinants of bone mineral content at the spine and radius: A twin study. Bone 8:207–209; 1987. 6. Du¨ ppe, H., Ga¨ rdsell, P., Nilsson, B., and Johnell, O. A. single bone density measurement can predict fractures over 25 years. Calcif Tissue Int 60:171–174; 1997. 7. Eastell, R. Treatment of postmenopausal osteoporosis. N Engl J Med 338:736 – 746; 1998. 8. Greendale, G., Barrett-Connor, E., Edelstein, S., Ingles, S., and Haile, R. 9. Holm, S. Pathophysiology of disc degeneration. Acta Orthop Scand 251(Suppl.):13–15; 1993. 10. Kannus, P., Haapasalo, H., Sankelo, M., Sieva¨ nen, H., Pasanen, M., Heinonen, A., Oja, P., and Vuori, I. Effect of starting age of physical activity on bone mass in the dominant arm of tennis and squash players. Ann Intern Med 123:27–31; 1995. 11. Krall, E. and Dawson-Hughes, B. Walking is related to bone density and rates of bone loss. Am J Med 96:20 –26; 1994. 12. Kro¨ ger, H., Tuppurainen, M., Honkanen, R., Alhava, E., and Saarikoski, S. Bone mineral density and risk factors for osteoporosis—A population-based study of 1600 perimenopausal women. Calcif Tissue Int 55:1–7; 1994. 13. Lakka, T. A. and Salonen, J. T. Intra-person variability of various physical activity assessments in the Kuopio Ischaemic Heart Disease Risk Factor Study. Int J Epidemiol 21:467– 472; 1992. 14. McClung, M. R. Therapy for fracture prevention. JAMA 282:687– 689; 1999. 15. Michaelsson, K., Bergstro¨ m, R., Holmberg, L., Mallmin, H., Wolk, A., and Ljunghall, S. A high dietary calcium intake is needed for a positive effect on bone density in Swedish postmenopausal women. Osteopor Int 7:155–161; 1997.
Bone Vol. 29, No. 5 November 2001:442– 446 16. Lane, J. M. Osteoporosis. Medical prevention and treatment. Spine 22(Suppl.): 32S–37S; 1997. 17. Netelenbos, C. Osteoporosis: Intervention options. Maturitas 30:235–239; 1998. 18. Prince, R. L., Smith, M., Dick, I. M., et al. Prevention of postmenopausal osteoporosis. N Engl J Med 325:1189 –1195; 1991. 19. Reginster, J. Y., Gillet, P., Ben Sedrine, W., Brands, G., Ethgen, O., De Froidmont, C., and Gosset, C. Direct costs of hip fractures inpatients over 60 years of age in Belgium. Pharmacoeconomics 15:507–514; 1999. 20. Rubin, L., Hawker, G., Peltekova, V., Fielding, L., Ridout, R., and Cole, D. Determinants of peak bone mass: Clinical and genetic analyses in a female Canadian cohort. J Bone Miner Res 14:633– 643; 1999. 21. Slemenda, W., et al. Genetic determinants of bone mass in adult women: A reevaluation of the twin model and the potential importance of gene interaction on heritability estimates. J Bone Miner Res 6:561–567; 1991. 22. Snow-Harter, C. and Marcus, R. Exercise, bone mineral density, and osteoporosis. Exerc Sport Sci Rev 19:351–388; 1991. 23. Twomey, L. T. and Taylor, J. R. Age changes in lumbar vertebrae and intervertebral discs. Clin Orthop 224:97–104; 1987. 24. Ulrich, C. M., Georgiou, C. C., Snow-Harter, C. M., and Gillis, D. E. Bone mineral density in mother– daughter pairs: Relations to lifetime exercise, lifetime milk consumption, and calcium supplements. Am J Clin Nutr 63:72– 79; 1996. 25. Welten, D. C., Kemper, H. C. G., Post, G. B., Van Mechelen, W., Twisk, J., Lips, P., and Teule, G. J. Weight-bearing activity during youth is a more important factor for peak bone mass than calcium intake. J Bone Miner Res 9:1089 –1095; 1994. 26. Wolff, I., van Croonenborg, J. J., Kemper, H. C. G., Kostense, P. J., and Twisk, J. W. R. The effect of exercise training programs on bone mass: A metaanalysis of published controlled trials in pre- and postmenopausal women. Osteopor Int 9:1–12; 1999. 27. Youm, T., Koyal, K. J., and Zuckerman, J. D. The economic impact of geriatric hip fractures. Am J Orthop 28:423– 428; 1999. ¨ hlen, G. The 28. Zethraeus, N., Stro¨ mberg, L., Jo¨ nsson, B., Svensson, O., and O cost of a hip fracture. Estimates for 1,709 patients in Sweden. Acta Orthop Scand 68:13–17; 1997.
Date Received: June 23, 2000 Date Revised: May 10, 2001 Date Accepted: May 15, 2001