Effect of physical activity on femoral bone density in men

BMJ 1995; 310 doi: (Published 10 June 1995) Cite this as: BMJ 1995;310:1501
  1. A G Need, directora,
  2. J M Wishart, research assistantb,
  3. F Scopacasa, research assistantb,
  4. M Horowitz, associate professorb,
  5. H A Morris, principal hospital scientista,
  6. B E C Nordin, visiting professorc
  1. a Division of Clinical Biochemistry, Institute of Medical and Veterinary Science, Adelaide, South Australia 5000
  2. b Department of Medicine, Royal Adelaide Hospital, Adelaide, South Australia
  3. c Department of Pathology, University of Adelaide
  1. Correspondence to: Dr Need.
  • Accepted 21 March 1995

Although most patients with osteoporosis are women, up to one third of hip fractures occur in men. There is little information about which factors influence bone density in men.1 Vigorous activity may lead to bone gain, while immobilisation causes bone loss. A sedentary lifestyle could, therefore, increase the risk of fractures.2 We therefore examined the relation between physical activity and bone density in normal men.

Subjects, methods, and results

One hundred and thirty seven healthy white men, comprising husbands of women attending our osteoporosis clinic, laboratory staff, and hospital workers who were enrolled in a normal bone study, listed their regular physical and sporting activities. Subjects taking drugs or with diseases likely to affect calcium metabolism were excluded. Forty eight were smokers (mean 15 cigarettes/day), and 120 drank alcohol (mean 8.6 g of alcohol/day). The time per week spent on each activity was multiplied by the energy expenditure for that activity and expressed in METs per week (a MET is an arbitrary unit of metabolic activity equivalent to average sitting oxygen consumption); 56 METs per week were subtracted for sleeping time.3

Bone density was measured in the forearm by single photon absorptiometry (Molsgaard Bone Mineral Analyser) and in the spine (L2-L4), femoral neck, Ward's triangle, and trochanter by dual energy x ray absorptiometry (Lunar DPX-L). Results were analysed by Student's t test and linear regression; age correction was done by multiple linear regression with age expressed as a quadratic function. Since bone loss accelerates in men over 50,1 measurements were evaluated separately in men aged 50 and over and in those aged under 50. A P value < 0.05 was considered significant.

The subjects' mean age was 47 (range 20-83), weight 77 kg (range 60-98), height 177 cm (range 157-200), and body mass index 24.7 kg/m2 (range 18.8-32.1). Forearm bone density was lower in those aged 50 and over than in those aged under 50 (472 (SD 77.6) v 522 (56.5) 95% confidence interval for difference: 25.6 to 74.4; P<0.001), but energy expenditure was similar in the two groups (129 (14.7) METs per week in the former and 131 (17.1) in the latter), equivalent to about 2.5 hours' jogging or 5.5 hours' walking. In the whole set activity was related to age-corrected bone density in the femoral neck (r=0.26, P<0.01) but not in the forearm or spine (r=-0.056 and 0.140 respectively). Activity and bone density were not related in men aged 50 and over. In those aged under 50 they were significantly related in the spine, femoral neck, Ward's triangle, and trochanter. These relations were unaffected by adjustment for age (table). Femoral neck bone densities in all subjects who jogged (22), walked regularly (53), or were sedentary (21) were 1.098 (0.121), 0.984 (0.152), and 0.962 (0.120) g/cm2 respectively; the difference between the joggers and sedentary men was significant (0.136; 95% confidence interval: 0.097 to 0.225; P<0.01).

Partial coefficients of correlation (age corrected) of bone density variables and activity score

View this table:


The major new finding in this study is of a continuous positive relation between physical activity and bone density in normal white men. This relation may reflect selection bias, inherent in a cross sectional study, but this is unlikely. Men who undertake regular outdoor activity have been shown to have a lower prevalence of hip fracture than those who do not,3 and active men have a higher bone density than sedentary men.4 Our data suggest that any exercise is beneficial to bone and that, within the normal range, the more the better.

Slemenda et al5 reported that the high leg and pelvic bone densities in young female figure skaters aged 10-25 were not evident until their mid-teens, suggesting that selection of skaters from a biased population was unlikely. Furthermore, the effect of exercise on bone appears to be related to site. Most of the activities we recorded were weight bearing and likely to stress the upper femur. The mean difference in femoral bone density across the range of activities recorded was 12%, equivalent to 18 years of bone loss.1 The observation that bone density was related to physical activity only in those aged up to 50 suggests that exercise may have its major effect on peak bone density.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
View Abstract