Over 30 years ago I gave a paper at an osteoporosis meeting in Jerusalem titled “The Paradox of Irreversibility of Age-Related Bone Loss.” By ”irreversibility” I meant that once the bone was lost there was not much that could be done to restore it.
Perhaps “puzzle” would have been a better word than “paradox.” From our experience with other similar situations, we would have expected that the lost bone would be restored. The underlying facts are that during the postmenopausal period bone loss occurs rapidly as estrogen levels drop to low values. Estrogen replacement therapy started at menopause prevents that loss, showing clearly that it is the estrogen deficiency that is responsible. Similarly, severe calcium deficiency also leads to bone loss, and maintaining a high calcium intake does slow that loss, and perhaps even prevent it. And, as is generally recognized, low calcium intake and low estrogen status are common in contemporary women during the post-menopausal years. These factors are the principal reasons for age-related bone loss in women.
But neither estrogen, nor calcium, nor the combination of the two, will restore the lost bone after it is gone. This seemed puzzling because with many other nutritional and hormonal deficiencies, restoring the lost hormone or nutrient does generally return the body to its pre-deficiency state. Examples would be hypothyroidism, which responds to thyroid hormone replacement, and iron-deficiency anemia, which responds fully to replacement of the lost iron.
The explanations for this seeming irreversibility which I offered at the time were twofold. 1) bone building (or rebuilding) requires weight-bearing or impact exercise, and physical activity generally declines after midlife; so a condition necessary for rebuilding was missing. 2) much of the lost bone is trabecular in character, i.e., the spongy latticework in the center of bones such as the vertebral bodies of the spine; once that lattice is lost, there is no longer a scaffolding or framework on which to rebuild.
I believe that both reasons are at least partially correct, but today they seem to me far from satisfactory explanations for this puzzling irreversibility. There is, I think, a better, more complete explanation, one that can be tested (and thus proved or disproved), and one that, if correct, could revolutionize the treatment and prevention of osteoporosis.
Bone is not just calcium. It is made up, first of all, of a protein matrix within which the calcium salts are embedded. Soak a bone in acid and you remove the calcium. But what’s left still looks like the bone you started with, except now it’s rubbery rather than hard. It’s now all protein and no mineral. The key point is that, while bone is the body’s reservoir of calcium, that calcium is tied up as part of a structure, the largest component of which is protein. When the body needs calcium and has to make withdrawals from the skeletal reserves, it does so not by leaching the calcium from this protein-mineral complex, but by physically tearing down microscopic units of bone and scavenging the calcium that is released in the process. Inevitably, therefore, the protein matrix – the structure – goes as well.
In order to profit fully from a high calcium intake, a patient who has lost bone needs to consume enough protein to allow the body to rebuild the lost structure. Otherwise all that a high calcium intake can do is to prevent the body’s further tearing down of bone to meet the calcium needs of other body systems and tissues. That’s a good thing to do, but it is not enough. Nevertheless, it is precisely to prevent that draining of the body’s calcium reserves that a high calcium intake (whether from food or supplements) is today a vital part of the standard of care for patients with osteoporosis. Even so, the failure of nutritional replacement to rebuild lost bone is what originally set the stage for the entry of pharmaceutical agents, some of which can produce substantial bone rebuilding.
That landscape began to change a few years ago when an insightful investigator at the Tufts Nutrition Research Center on Aging in Boston noticed that a high calcium intake did, in fact, lead to increased bone gain if the patient’s intake of protein was high. Bess Dawson-Hughes had previously published the results of a calcium and vitamin D supplementation trial, producing a better than 50 percent reduction in fracture risk in healthy elderly Bostonians with those two nutrients alone. But, like others before her, she noted that, while high calcium intakes reduced or stopped bone loss in her treated subjects, the two nutrients didn’t lead to bone gain. They didn’t, that is, in individuals consuming usual protein intakes. However, in a subset of her treated patients, who, it turns out, had protein intakes above 1.5 times the RDA (0.8 g/kg body weight), bone gain was dramatic (while it was zero in those with more usual – and usually thought “adequate” – protein intakes). The figure below shows the 3-year change in bone mineral density (BMD) at the hip in the calcium- and vitamin-supplemented participants in the Tufts study. Only with the highest protein intakes was there appreciable bone gain.
For me, it was an “Aha!” moment. Why hadn’t we thought of that? It was known that bone is 50 percent protein by volume (but only about 20 percent calcium by weight). And it was known that when bone is torn down (as with estrogen or calcium deficiency), its protein is degraded in the process. So it made sense that, to rebuild the lost bone, you would need not just calcium but fresh protein as well.
When I first heard of this result, I immediately went to our own Creighton database on calcium metabolism in midlife women (the “Omaha Nuns Project) and looked to see whether protein intake (which we had recorded and measured) made a difference in the bone metabolism of our nuns. There it was, just as the Tufts investigator had shown. Our nuns with protein intakes below the median for the group could not retain calcium, no matter what the intake (i. e., they couldn’t build bone). By contrast, those with protein intakes above the median for the group retained extra calcium reasonably well.
So, here were two distinct data sets, two quite different investigations, exhibiting the same interdependence of calcium and protein. What we, and probably most clinical nutritionists, had failed to recognize, was that the adult RDA for protein is just barely enough to prevent muscle loss, and is not enough to support tissue building or rebuilding. But, as already noted, when calcium deficiency leads to bone loss, the bone protein is lost as well, and that has to be rebuilt to restore the lost bone.
This mutual dependence of calcium and protein provides a good illustration of two key (and often underappreciated) aspects of nutrition. The first is that nutrients almost always act together with other nutrients. The second feature is what Bruce Ames of the University of California, Berkeley, has called a “triage” system within nutrition. The body operates a triage mechanism, ensuring that the most vital functions receive the nutrients first and leaving the other tissues and systems of the body to get by on what is left over. It seems that this triage mechanism is at work with respect to adult bone rebuilding. With limited protein intake, the body ensures that its most vital functions are served first. Bone, in effect, gets the leftovers. We need a high protein intake precisely to ensure that there will be something left for bone.
Two unplanned observations such as those of Dawson-Hughes and our own Creighton group, even if they make perfect sense, would not generally be considered enough to change public policy, particularly when it comes to nutrient intake recommendations. So, if we are to be certain that supplementing both protein and calcium will permit rebuilding of lost bone, it will be necessary to mount one or more clinical trials testing that hypothesis.
Such a trial would likely be designed to start with a group of probably several hundred postmenopausal women who had already lost bone and whose protein intakes were in the range of the current RDA, that is about 0.8 g/kg/day. All would be supplemented with sufficient calcium to permit maximum bone building if the individuals concerned could, in fact, use the calcium efficiently. They would all also receive sufficient vitamin D to ensure a serum concentration of 25(OH)D of 40 ng/mL or higher. Then half would be given a diet, probably involving a protein supplement, which would raise their protein intakes to above 1.2 g/kg/day. [Some might argue that there should be a third group, one with protein intakes at the RDA, but without the substantial calcium and vitamin D supplementation envisioned above.] In either case, trial duration would be about three to four years, and the endpoint would be the observed change in BMD over that treatment period. The predicted outcome would be that the lower protein group receiving calcium and vitamin D would have no appreciable change in BMD, while the higher protein group, also receiving extra calcium and vitamin D, would exhibit clinically significant bone gain. As outlined here, such a trial could not be blinded, mainly because the diets would be perceptibly different.
Even if such a trial were to start today, it would probably be at least five years before the results would be clear and actions could be taken to change official recommendations and influence individual dietary behaviors. What should one do in the meanwhile?
This is a matter for individual decision, but it is helpful to know that high protein intakes are safe. Their principal negative impact is on the wallet, not the body (as rich protein foods tend to cost somewhat more than foods high in added sugars, for example, or other types of empty calories). For me the decision is easy; I’d opt for the high protein intake without a second thought (with, of course, adequate calcium and vitamin D, as well). As a dividend, I should note that there is one food group that is both a very rich source of protein and at the same time the principal source of calcium in the diets of first world populations – dairy. Moreover, if consumed as milk, its cost is less than the average cost of the other foods in your grocery cart.
Even if the trial were to be successful, It would be naïve to think that would be the end of the story. Recall that the pharmaceutical industry stepped into this field 25 years ago when it appeared that nutritional therapy was not up to the task (at least as it was conceived at the time). If this protein hypothesis is correct, then better nutrition could be a much better form of prevention than pharmacotherapy. However, I suspect that the pharmaceutical industry will not back out of the field as readily as it got into it.
To be fair, resistance from big Pharma should not be surprising. After all, they’ve invested billions of dollars in helping us solve a critical health problem for an aging population. Naturally, they (and our pension funds who are their stockholders) want to protect that investment. Still, if diet can do the job for us, few would choose a lifetime of pill taking or injections over better eating.
This blog would be incomplete if I did not call attention to the fact that bone structure and density are designed by natural selection to resist mechanical loads, in other words, to permit a person to do physical work. In the absence of continuous mechanical loading, there is no diet, by itself, that will allow an older adult to regain the bone he/she had as a child. So, yes, calcium is important. And protein is important. But physical work is important, too. How much Ca? – probably 1500–1800 mg/day. How much protein? – probably at least 1.2 g/kg body weight/day. How much exercise? – probably about what the cardiovascular exercise people recommend, with special emphasis in this instance on impact exercise, such a jumping rope. Look at toddlers. Look at the impact forces to which they subject their skeletons. That’s how they grow bone.