Regular physical activity is routinely recommended by clinicians for patients at risk for osteoporosis. There are; however, no specific guidelines for clinicians regarding the type, duration or intensity of physical activity that is most appropriate for these patients. Furthermore, patient compliance regarding exercise programs is generally low in the majority of clinical settings. As such, physical activity in any form becomes important in the prevention of bone loss. The current study sought to determine whether a relationship exists between the amount of regular physical activity performed on a weekly basis and bone mineral density in Canadian women aged 75 and over.
Physical activity in women aged 75 and over
The vast majority of participants reported some level of involvement in moderate physical activity, i.e. that which could be considered activity over and above the general activity of day-to-day life, such as brisk walking, golfing, housecleaning, etc. Close to three-quarters of participants (71.7%) reported that they are moderately active for at least 4 hours per week. This is an encouraging finding, in light of the fact that, in Canada, up to 64% of female seniors are considered inactive  while in the United States, over 60% of senior women were reportedly not meeting the minimum recommendations for regular physical activity (approximately 15–20 minutes daily) .
Effects on bone mineral density
The findings of this study indicate that regular physical activity at a moderate level can help to improve bone density in post-menopausal women, although these improvements were limited largely to the hip region. These findings echo those of similar studies that have shown that the benefits from exercise or physical activity are generally noted in the hip but not in the lumbar spine. Bolton et al.  demonstrated in a recent randomized, controlled trial of post-menopausal women that an increase in regular physical activity can have a positive impact on bone mineral density. In their study, over the course of one year, participants took part in a general exercise program that included 60-minute exercise training three times each week, where control participants continued in their normal daily routine. The exercise training group performed tasks including resistance training, moderately intense exercise and training. The authors found that there was a positive (although not statistically significant) effect on bone density in the hip region but a negative (although also not statistically significant) decrease in bone density in the lumbar spine. The measured difference in BMD in the current study closely approximates that of the Bolton study, especially regarding the location of improvements. These findings are likely not unexpected, as the benefit gained from resistance or impact exercise relates largely to the effect of loading on the skeleton [21–23]. The hip joint will absorb the majority of the forces applied during land-based exercise, while the lumbar spine will absorb very little physical force. As such, the majority of exercises are designed to address the hip, an important fact due to the simple fact that the hip, being the structure that absorbs more force during these type of tasks, is also the structure more likely to be damaged (i.e. to suffer a fracture).
The results from the current study contradict the findings of Gerdhem et al. , who noted no correlation between previous and current physical activity level and bone mineral density in women up to age 75. The questionnaire used in their study contained 10 questions relating to past and current physical activity level, half of which related to physical activity associated with employment. Only one question related to the current level of physical activity. The more detailed information gained from the CaMos questionnaire regarding current activity level may help explain the contradictory findings. The lack of consistency between physical activity/training programs and duration are discussed by Gerdhem as possible explanations for the contradictory findings in previous studies. Such variations could explain the inconsistencies between the results of these two studies.
The results from this study indicate that there was a statistically significant improvement in bone density associated with a step increase in the amount of moderate physical activity performed on a regular basis. The essential question, then, is: is this improvement clinically important? The most common treatment for osteoporosis are the bisphosphonates. These medications have been shown to induce an average increase of approximately 0.019 g/cm2 following a one-year course of treatment . The findings from the current study indicate that the improvements in bone density range from 0.006 g/cm2 (for femoral neck, Ward’s triangle and the trochanter) to 0.008 g/cm2 (for the total hip). These improvements represent between 30-50% of the improvement expected from bisphosphonate treatment. Warming et al.,  performed a prospective study to evaluate the normal changes in BMD in the forearm, hip, spine and total body, in otherwise healthy men and women. They used DXA measurements at 2 year intervals in over 500 participants and found that, in women, the only pre-menopausal bone loss was noted at the hip (<0.003 g/cm2/year). In women after menopause, though, bone loss ranging from 0.002 g/cm2/year to 0.006 g/cm2/year was noted in all sites. The greatest post-menopausal bone loss was found in forearm, where 1.2% (0.006 g/cm2/year) was lost following menopause, a change that remained constant throughout life. While the changes noted in this study do not meet the level of bisphosphonate treatment, it appears that an increase in the amount of MPA on a daily basis may be enough to offset the normal bone loss that occurs following menopause. If this is indeed the case, the importance of encouraging elderly patients to remain active on a daily basis is underscored.
Several factors were considered possible confounding factors in this study, based on their ability to affect bone mineral density. The results regarding medication use indicate that, perhaps expectedly, the use of anti-resorptive therapy reversed the negative effect on BMD in the lumbar spine and increased the protective effect in each of the other BMD sites, although only the improvements in the lumbar spine and femoral neck were statistically significant. It is not surprising that anti-resorptive therapy counteracted the observed decrease in BMD noted in the lumbar spine and result instead in a positive regression coefficient and a relative increase in BMD.
Other factors considered in this study included race, body mass index (BMI) and participant age. Because race and/or ethnicity are known to impact on bone loss and the incidence of osteoporosis, race was initially intended to be considered as a secondary factor. Analysis of the database; however, indicated that the large majority of participants (97.9%) identified themselves as “white”, which essentially made an examination of the effect of race on bone loss impossible. The relationship between BMI and BMD indicated that increased BMI resulted in a relative protective effect on bone density. These findings support those of several authors [27–29], who have also observed that increased BMI is associated with a lower risk of osteoporosis.
This study has several limitations which prevent the direct application of its findings to clinical settings.
The homogeneity of the cohort with respect to racial and/or ethnicity make-up makes application of the results difficult. With 97.9% of participants identifying themselves as “white”, the ability to determine racial differences is impossible. The CaMos cohort, while sampling from a large proportion of the Canadian population as a whole, does not fully reflect Canadian society as a whole. Indeed, taking the entire CaMos cohort into account, 94.9% of the 9423 participants identified themselves as white. While this may a valuable factor when considering that Caucasian women are at a higher risk of osteoporosis as compared to other racial groups such as blacks or hispanics, the ability to apply the findings to an increasingly racially diverse Canada is limited by these demographics.
The initial plan for this study was to compare physical activity considered part of normal day-to-day activity with more strenuous activity, to determine the relative effects on bone density and fracture rate. The observation that over 96% of the study cohort took part in no vigorous or strenuous activity whatsoever made that analysis impossible. It is unfortunate that more participants were not active to these greater degrees, as it would have better reflected the potential role of exercise in the protection against fracture. However, this finding is mitigated by the fact that beneficial effects were noted simply by increasing the amount of MPA performed each day, which is likely easier in it implementation than incorporating a vigorous exercise program into the routines of elderly patients.
An important factor in this study was the use of anti-resorptive medication by some participants. These medications certainly have a positive effect on bone density; however, their use, in combination with exercise and activity, is an important clinical consideration, especially when clinicians are faced with the choice of prescribing medication for their patients. Of the 1169 participants in this study, only 150 reported currently using anti-resorptive medication. Of those, only 50 were using bisphosphonates, the most common anti-resorptive medication, and a mere 5 were using SERMs. This represents less than 0.5% of the entire study cohort, an amount insufficient to determine a possible confounding effect of anti-resorptive use.