To the Editor:
A meta-analysis of calcium supplementation in children in a recent
issue of the Journal reported a positive effect, but deemed it too small
to be useful (1). Service, in a Perspective on meta-analysis, noted that
many meta-analytic teams lack members with content expertise in the area
being summarized (2). As a consequence, while the meta-analyses may be
rigorous with respect to methodologic issues, they can overlook important
biological flaws in the studies concerned or pool studies that are
incommensurable. The publication by Winzenberg et al. (1) constitutes a
good case in point.
Among several biologic errors in this meta-analysis, two in particular
stand out: the use of bone mineral density (BMD) as the outcome variable
and the failure to ensure that all studies included a low calcium control
group. Both issues have been explored extensively elsewhere and are
treated in depth in at least one standard reference book (3–6).
Briefly, during growth, bone expands in three dimensions. BMD, as measured
(and by design), eliminates the increase in mass associated with two of
those dimensions. Thus BMD change misses most of the actual change in bone
mass during growth. (And when true density can be measured, it misses mass
change entirely.) In this instance, the meta-analysis found a significant
aggregate positive effect, but the authors concluded that it was too small
to be useful. They were apparently unaware 1) that changes in bone mass
during growth are typically several times larger than changes in BMD; and
2) that bone strength is directly related to mass and size, and only
indirectly to BMD.
Furthermore, use of BMD explicitly precludes finding an effect of the
intervention on bone size. The latter point is not simply a theoretical
objection, as several investigations have suggested that high calcium
intakes, in addition to improving bone mineral acquisition, can increase
bone size as well (7,8). Thus BMD is precisely the wrong outcome measure
for nutritional effects during growth.
Calcium, like iron, vitamin D, and many other nutrients, exhibits
threshold behavior, i.e., effects accrue up to only a certain threshold
intake value, above which further increases in intake produce no
additional effect (4,5). Studies without a low calcium intake contrast
group are not capable of testing whether calcium produces skeletal (or
other) benefits. Several of the studies included by Winzenberg et al.
exhibited precisely that problem.
The fact that the investigators in the studies assembled in this meta-
analysis had themselves made these same mistakes does not justify their
inclusion. One of the points of a meta-analysis is to weed out flawed
studies – flawed not just because of poor randomization or blinding, but
because of inapposite methods and poorly posed biological questions.
In both their title and their conclusions the authors are careful to use
the term “healthy children”, a designator which would logically include
the notion that their diets were adequate in all essential nutrients. In
that sense, therefore, one cannot disagree with the conclusion that more
calcium does not produce much benefit. But since that point has been shown
many times over and is implicit in the notion of a threshold nutrient, one
wonders why this analysis was done or published in the first place. It is
likely that the message for the average reader would be that calcium
intake in children is not important. That would be not only wrong, but
potentially dangerous, as well.
Robert P. Heaney, M.D.
Creighton University, Omaha, Nebraska, USA
Connie M. Weaver, Ph.D.
Purdue University, West Lafayette, Indiana, USA
1. Winzenberg T, Shaw K, Fryer J, Jones G. Effects of calcium
supplementation on bone density in healthy children: meta-analysis of
randomized controlled trials. BMJ 2006;
2. Service FJ. Idle thoughts from an addled mind. Endocr Prac 2002;8:135-
3. Prentice A, Parsons TJ, Cole TJ. Uncritical use of bone mineral density
in absorptiometry may lead to size-related artifacts in the identification
of bone mineral determinants. Am J Clin Nutr 1994;60: 837-842.
4. Heaney RP. Design considerations for clinical investigations of
osteoporosis. In: Osteoporosis, 3rd Ed. Marcus R, Feldman D, Nelson D,
Rosen C, eds. Elsevier Inc., San Diego, CA (in press) 2007.
5. Heaney RP, Bachmann GA. Interpreting studies of nutritional prevention.
A perspective using calcium as a model. J Women’s Health 2005;14:990-897,
6. Heaney RP. BMD: The problem. Osteoporos Int 2005;16:1013-1015.
7. Prentice A, Ginty F, Stear SJ, Jones SC, Laskey MA, Cole TJ. Calcium
supplementation increases stature and bone mineral mass of 16-18 year old
boys. J Clin Endocrinol Metab 2005;90:3153-3161.
8. Bonjour J-P, Carrie AL, Ferrari S, Clavien H, Slosman D, Theintz G,
Rizzoli R. Calcium-enriched foods and bone mass growth in prepubertal
girls: a randomized, double-blind, placebo-controlled trial. J Clin Invest
Competing interests: No competing interests