Exercise, Bone Density, and Fracture Risk

Elaine Mansfield

Many of my female clients between the ages of 35-70 have exercised aerobically and kept a healthy weight for years only to be diagnosed with osteoporosis or low bone mineral density (BMD).  Like them, I assumed that my exercise and nutrition program would result in excellent bone density, but a year ago, I requested a dexascan (bone density test).  It didn’t seem fair to advise women about their dexascan results without knowing my own.  I was shocked to find that I have mild osteoporosis in my hips and osteopenia (more moderate bone loss) in my lumbar spine.  I spent a day feeling weak and defeated, but at 2 am that sleepless night, I pulled out my health library and began re-reading information about bone density.  Where did I go wrong?  How frail am I?  How could this happen to a woman who lifts weights and eats a bone-friendly diet?

A diagnosis of low bone density is increasingly common with wide use of dexascan machines.  The emotional reaction is also common.  Even women who are in great aerobic condition suddenly feel frail and vulnerable.  In the past, women with low BMD were advised to use hormone replacement therapy (HRT) for bone strength, but this is no longer recommended because of HRT related health risks.  Now, healthcare providers give women literature about osteoporosis that focuses on the ever-expanding drug treatments available—treatments that are quite new and considered safe just as HRT was recently considered safe.  Along with scant nutritional advice, women are told to do weight-bearing exercise.  The exercises recommended are for unfit or even frail beginners.

After my night of reading and reviewing, I realized that there are tests that could help me learn if I am losing bone density now or if my low BMD is a result of past habits.  Although I’ve focused on good health for many years, I didn’t begin resistance training until my mid 50s and between the ages of 15 and 25, the years when bone-building is most efficient, I assaulted my bones with cigarettes and severe weight-loss diets without taking adequate calcium or other minerals.  I requested a urinary bone resorption test to measure the level of breakdown products of bone.  This test and a second one 6 months later show that I am not losing bone now.  I also requested a serum vitamin D (25-hydroxyvitamin D) test to make sure vitamin D levels are adequate for good calcium absorption and utilization.  They are.

I was relieved to find that my bones are stable or possibly growing denser, but since bone density changes only a few percent a year in women my age (58), it will be a few years before a dexascan shows statistically relevant changes.  It is likely that I’ve had low bone density for years—probably lower before beginning strength training, although I’ll never know.  Genetics could also be a factor, although no female family members have had obvious osteoporosis.

This brings me to some essential points.  Low bone density as measured by dexascan is only one of many factors related to fracture risk.  There are many equally important risk factors, such as inability to get out of a chair without pushing with the arms, lack of exercise, use of prednisone, cortisone, or aluminum-containing antacids, high caffeine intake, and low vitamin D intake.  Also, research shows that exercise lowers fracture risk even if bone density doesn’t show improvement on a dexascan.  Still, my clients and I want to do all we can to both lower our fracture risk and increase BMD.  This paper will discuss appropriate exercise and leave nutritional concerns for another time.

Basic Information About Bone Mineral Density

Osteoporosis is a disease that affects older individuals but begins when we’re young.  Young people need appropriate exercise and adequate bone-building nutrients to build healthy BMD for life, but bone health is the last thing on a young person’s agenda.  Most of the research is done on aging women, since osteoporosis and fracture prevention has become big business, but men have problems with bone density, too, and the research results are just as important for them.  Men have denser bones in their youth because of testosterone and don’t have the precipitous drop in hormone levels that women experience at menopause.  Still, with increased longevity, hip and spinal compression fractures are becoming common for men over 80.  For women, peak bone mass comes between 2 years post menarche (13-15 years for most women) and 30 years of age, followed by an average of 1% annual loss and a loss between 2-5% a year during menopause.[i]  After menopause, it’s possible to gain bone density, but without high levels of estrogen, changes are slow. 

Low BMD is most closely associated with low lean body mass (LBM), low calcium and vitamin D intake, and past and present inactivity.  On average, strong people have strong bones.  For overweight individuals, high body fat correlates with high BMD since the bones and muscles are working hard to support excess weight.  Individuals with small bone structure are more likely to have low BMD.

In the general population, yard work and strength training are the types of exercise most associated with high BMD and low fracture risk.  Women who have a lifetime habit of walking have about 30% lower fracture rate and slightly higher bone density than sedentary women, but walking and other low impact exercise like rowing, biking, swimming, or even running have little short-term impact on bone density in older individuals.[ii]  Exercise programs designed to increase bone density have the most impact on sedentary populations.

What Does Research Tell Us About Osteoporosis And Exercise?

Problems with Research

Many exercise studies resulted in little measurable change in bone density, and the researchers questioned whether the subjects were exercising with adequate intensity to change bone.  Looking at the meager strength gains for untrained subjects in some studies, I suspect they were not.  Many studies were short duration, while bone change in older individuals is slow and different sites in the body respond at different rates to exercise stimulation.  The femoral neck (FN) is the narrowed section of leg or femur bone located just below the insertion point of the femur into the hip joint.  The FN is the most common area of hip fracture in the elderly and responds more slowly than the rest of leg and hip to exercise stimulation.  All hip sites respond more slowly than the spine.  Research results are further complicated because bone density may initially decrease during high intensity strength training, caused by an initial phase of bone breakdown that initiates the process of bone remodeling and strengthening.[iii]  Often the exercises used seem inadequate.  For example, no studies used deadlifts or back-hip machines that provide excellent stimulus for back and hip bone, since they build strength in those area.  Progression techniques seemed inadequate in many studies.  Load was sometimes raised every 6 or 8 weeks rather than weekly or at each session.  Finally, research results were complicated by use of hormone replacement therapy (HRT) and calcium and vitamin D supplementation.

Despite these limitations, bone density research gives important clues about designing an exercise program to improve bone density and lessen fracture risk.

1. Site specificity

Bone becomes strong when muscles attached to it become strong.  The one-year BEST (Bone Estrogen Strength Training Study) included resistance training 3 times a week.  The most statistically significant positive effect for hipbone strength was from weighted squats.  Weighted marching (raising legs to form 90 degree angle of hip and knee joints) showed the most significantly correlation between the total amount of weight used and changes in total body bone density (11% increase). [iv]  It’s widely known that individuals who play racquet sports have higher BMD in the racquet arm.[v]

2. Different sites change at different rates

The trabicular bone of the spine remodels more rapidly than the cortical bone of hip and wrist.  A yearlong study found a 1.3% increase in lumbar spine (LS) BMD compared to –1.2% in the control group.[vi]  At the same time, there was a loss of bone at the hip in both groups (statistically significant only in the non-exercising group).  In a different 9-month study, LS density increased 1.6 % compared with a loss of 3.6% in controls, but there was no change in FN or wrist BMD.[vii]  In a study of women 28-39, training increased BMD at the LS by 2.8% at 5 months.  The FN had no change at 5 months, but increased by about 2% at 18 months.[viii] 

3. Intensity: an adequate dose yields a good response

In a meta-analysis, Wolff found that exercise-training programs prevented or reversed an average of 1% bone loss per year across randomized control studies.  Since bone loss is about 1% a year before menopause and returns to about 1% a year at approximately 5 years after menopause, strength training at least can halt bone loss.  The effect was nearly twice as large when women chose to be in the exercise vs. the control group, so the assumption is that women who were more enthusiastic about their training worked more intensely and had better results.  Wolff concluded that “strength training programs (that failed to show changes in BMD) might not all have imposed the required unusual strain distributions, high strains, and high strain rates.”[ix]

In a 1-year study of women over 60, subjects showed increases in strength in the first 6 months and a decrease by 12 months in many exercises, back to baseline levels in some exercises.  According to the study, “the smaller gains in lower body strength may partially explain the lack of change in BMD at the hip and spine.  Those sites may not have been adequately loaded to stimulate bone mineralization.”[x] Obviously, if healthy subjects aren’t progressing in the first year and strength is lessening after 6 months, the exercise program isn’t working and subjects are likely under or over-training.

A one-year study by Pruitt found no change in two groups of exercising women aged 65-79.  Pruitt reported that “most women in the present study, many of whom were in their seventh decade of life, were not comfortable training to muscular fatigue on a regular basis…”[xi]  Most women in this study began with good BMD, another reason not much change was found.

4. Frequency, load, and progression

In animal studies, Turner found that bone rapidly becomes desensitized to prolonged exercise.  “The osteogenic response to exercise can be enhanced by regimens that incorporate periods of rest between short vigorous skeletal-loading sessions.”[xii]  Animals who jumped 100 times a day didn’t improve bone mass significantly over animals that jumped 40 times.  Rats who were given 8 hours rest as opposed to 0-4 hours rest between exercise sessions had superior bone formation.  After 24 hours of rest, the osteogenic response of bone to exercise was greatest.

In humans, Cussler found that maximal load is most relevant in BMD changes, not load frequency.  In a one-year study, women exercised with 2 sets of 6-8 reps with loads at 70-80% 1RM.  Cussler reported that a small number of loading cycles worked best and there were diminishing returns with more.  She suggests 3 days a week maximum on nonconsecutive days.[xiii]

In a 6-month study, adults aged 60-83 in two groups each used 1 set.  The first group did 13 reps at 50% 1 RM max while the second group did 8 reps at 80%.  Increase in total strength was similar in the two groups, but the high intensity group increased hip BMD by 2% and increased biochemical indices of bone turnover in areas where no change in BMD were found, indicating that bone was remodeling.  Low intensity and control group had no change in BMD.[xiv]

In a 1-year study, Kerr found no change in BMD in endurance exercisers (3 x 20).  In the strength group (3 x 8), bone mass increased an average of 2% in most hip sites (except at the femur neck) and about 1% in forearm, compared to the non-exercised side which was used as a control.  The researchers suggested that osteogenesis occurs in response to local muscle tendon pull on bone and that the poor response in the FN site may be because the FN doesn’t have direct muscle attachments.  This study also reported that in the strength group, but not the endurance group, “the greater the increase in muscle strength the greater the increase in bone mass.”[xv]

Using just 5 resistance exercises, hip extension, knee extension, lateral pull-down, back extension, and abdominal flexion (3 x 8 at 80% 1RM) twice a week for 1 year, Nelson reported FN BMD increases of +1% vs. controls –2.5%.  LS increases were similar with +1% in exercise group vs. –1.8% in controls.[xvi]  This study increased the load each training session as tolerated for each muscle group, whereas most studies waited 4-8 weeks to increase resistance.  Perhaps this constant progression made the exercise more effective.  It’s also interesting that exercisers trained twice a week with only a few exercises, suggesting that lack of over-training improves BMD response.

5. High impact

High impact builds strong bones.  Intercollegiate gymnasts (despite a high prevalence of food disorders that might lead to low BMD) showed increases of over 3.5% in the spine and 2% in the hip during two 8-month competitive seasons.[xvii]  Gymnasts consistently have higher bone density than other athletes probably due to high impact and intense muscle pull.  In non-athletes, high impact exercise increased hip BMD by 2-5% in women under 30 who did 50 vertical jumps daily compared to controls who did similar overall exercise programs that excluded jumping.[xviii]

High impact exercise brings excellent bone-building results for young people, but can’t be recommended for older adults with low BMD or fussy joints.  I proved that high impact work irritates my healthy joints by practicing a vigorous jumping routine 3 days a week.  Besides irritating my knees, jumping fatigued my quads and halted all progress in squatting motions.  I wrote that off as another noble but misguided experiment. 

6. BMD vs. Fracture Risk

Exercise lowers fracture risk even if BMD doesn’t change.  In the Nurse’s Health Study, women who walked 8 hours at a moderate pace (approximately 24 mets exercise/week) had 55% less hip fracture risk than sedentary women and those who walked just 4 hours a week had 41% lower risk.  On average, the risk for fracture dropped 6% per 3 mets of exercise per week.[xix]

In animal studies on rats, BMD changed only 5.4% but resulted in a 64% increase in the ultimate force (the maximum amount of force the bone could support before breaking).  New bone formation was localized at area where mechanical load and stress was greatest.  “In rats, even small changes in bone mass, which are marginally detectable by DXA (dexascan), can significantly improve bone strength by favorably altering bone geometry.”[xx]

In both animal and human studies, exercise is more osteogenic during skeletal growth than during maturity.  “In elderly adults with low bone mass, exercise constitutes only a moderately effective bone-building therapy….  However, exercise can effectively reduce fracture risk even without dramatic effects on bone mass.”[xxi]  It’s difficult to get this point across to frightened clients who think that low BMD guarantees a fracture.  Exercise reshapes bone structure and geometry, increasing strength and decreasing fracture risk in ways not measurable by bone scan. 

7. Delayed Response

It can take 4-6 months or more for bone to remodel under the best conditions and measurable effects of exercise may only be apparent years later.  The following study is my personal favorite.  In a 2-year study of women aged 58-75, 27 exercisers (23 controls) practiced 10 back lifts 5 days a week, progressing to a maximal weight of 50 lbs in a specially constructed backpack.  After 2 years, there was no change in BMD for controls or the exercise group despite large changes in back erector muscle strength in the exercise group.  Exercise was discontinued.  In a follow-up 10 years later, the exercise group had a significantly higher BMD in the spine, and the relative fracture risk was 2.7 times greater in the control group than the exercise group.[xxii]

Conclusions

What does the research tell us about exercise, bone density, and lowering fracture risk?  Our routines need to be high load, low rep, and site specific.  A twice-weekly routine of about 8-10 exercises (1 set of 7-9 reps) that includes a squat variation, deadlift, back erector work, abdominal work, and upper body exercises that stimulate the muscle groups along the spine as well as the upper arm and forearm, satisfies the major criterion.  We want our exercise sessions brief and intense with excellent form, paying special attention to back safety.  If the trainee has spinal osteoporosis, avoid resistance exercises that use a forward trunk curl since this position makes a weakened spine vulnerable to fracture.  We need to provide adequate time between exercise sessions and supply the body with bone-building and other nutrients for optimal recovery.  Finally, research results teach us to be patient, a virtue for any lifter.  Bone changes are slow, much slower than strength changes.

Using exercise to improve BMD brings us back to the basic advice Master Trainer has been giving for years.  If we use high load and low rep routines of compound exercises that stimulate muscle development around the hips, spine, and arms, we’re building bone strength in those vulnerable areas and throughout the body.  Even if BMD is not improved as measured by dexascan, resistance training with adequate intensity will dramatically lower our lifetime fracture risk.  When it comes to bones, that’s the result we’re after.

Elaine Mansfield is a nutritionist and personal trainer practicing in upstate New York.  Check her web site for practical exercise and nutrition ideas:  www.elainemansfield.com

© 2004 Elaine Mansfield