Robert P. Heaney, M.D., Osteoporosis Research Center, Creighton University Medical Center, Omaha, USA.
“Most of the studies investigating the connection between calcium intake and body weight found that increased calcium intake augmented the weight loss of energy-restricted diets, protected lean body mass, and reduced age-related weight gain, or that it had null effect. A recent meta-analysis reported results also favoring calcium (1). However, the authors seem concerned to characterize the effect as “small” and of “uncertain clinical relevance.” Many other authors writing on this topic have also waffled, characterizing the issue of calcium and weight as “controversial.” There is actually very little that is controversial about the facts. Where there is conflict, it is between the very different expectations that investigators, the general public, and nutritional policy bodies bring to the data. This most recent meta-analysis affords an opportunity to clarify some of the confusion surrounding this issue.
The meta-analysis reports a mean weight-loss difference of 0.74 kg in trials, most of which were of
6 months’ duration, for an annualized rate of loss of about 1.5 kg. To a person weighing 140 kg and hoping to lose 70 kg, that rate of loss would, indeed, have very limited interest; however, for a population confronting secular weight gain, this same weight effect is huge. Secular weight gain typically follows from an energy balance of as little as +50 kcal/day; clearly, this is a very small difference, but one that, when extended over many years, nevertheless produces outspoken obesity. Thus, the magnitude of the weight loss effect, when viewed from the perspective of a population gaining weight over time, must be recognized as of high clinical relevance and interest. However, an energy balance that small is extremely difficult to detect in clinical studies. It is to be expected, therefore, that counteracting that small positive energy balance (as with high calcium intakes) would be equally hard to detect. Thus, it is hardly surprising that there are many null trial reports for calcium and weight.
Yet another instance of this same apparent discrepancy between effect size and importance is found in the matter of calcium intake and blood pressure control. Multiple meta-analyses have confirmed a blood pressure lowering effect of adequate calcium intake. As with weight, though, the average decrease is small (1–5 mmHg), and in one of the meta-analyses the authors went so far as to write “... the effect is too small to support the use of calcium ... for preventing or treating hypertension ” (2). Such a statement is at best misguided, as population data show that each decrement of 1 mmHg can reduce the prevalence of hypertension by as much as 5% (or more). Yet, in an individual patient with high blood pressure, a drop of just a few mmHg is indeed a small effect. The problem, as noted earlier, lies entirely in the expectations we bring to the issue.
So, what can responsibly be said about calcium and body weight? Simply this: Other things being equal, a generous dietary calcium intake 1) will reduce or eliminate the gain in weight that commonly occurs in mid-life; 2) should be a component of any weight-loss regimen, as it augments the weight loss of a caloric deficit while protecting lean body mass; 3) is not a substitute for control of an energy intake/output imbalance; and 4) is not a drug and is certainly not a magic bullet.
Apart from the issue of weight, we must ask ourselves the following question: Is it realistic, or even appropriate, to expect a large effect size for nutrients – that is, any nutrient and any effect? For the most part, nutrient effects are, as just noted for calcium, small in size, even if large in cumulative impact. Moreover, they typically involve multiple body systems, each with a distinct, if small, effect. They thereby exert a broad range of effects, many of which can be clinically important. Nutrients, however, are not drugs, and they should not be thought of as drugs. Nor can they be appropriately tested or meta-analyzed as if they were drugs (3).”