Calcium Fortificants: Overview and Strategies for Improving Calcium Nutriture of the U.S. Population


  • K. Rafferty,

    1. Authors Rafferty and Heaney are with Creighton Univ., Osteoporosis Research Center, 601 N. 30th St., Suite 4820, Omaha, NE 68131, U.S.A. Author Walters is with PURAC America, Food and Nutrition, 111 Barclay Boulevard, Lincolnshire, IL 60069, U.S.A. Direct inquiries to author Rafferty (E-mail:
    Search for more papers by this author
  • G. Walters,

    1. Authors Rafferty and Heaney are with Creighton Univ., Osteoporosis Research Center, 601 N. 30th St., Suite 4820, Omaha, NE 68131, U.S.A. Author Walters is with PURAC America, Food and Nutrition, 111 Barclay Boulevard, Lincolnshire, IL 60069, U.S.A. Direct inquiries to author Rafferty (E-mail:
    Search for more papers by this author
  • R.P. Heaney

    1. Authors Rafferty and Heaney are with Creighton Univ., Osteoporosis Research Center, 601 N. 30th St., Suite 4820, Omaha, NE 68131, U.S.A. Author Walters is with PURAC America, Food and Nutrition, 111 Barclay Boulevard, Lincolnshire, IL 60069, U.S.A. Direct inquiries to author Rafferty (E-mail:
    Search for more papers by this author


ABSTRACT:  Despite more than 20 y of awareness of the importance of calcium to health, U.S. calcium intakes remain suboptimal. Fortification of foods with shortfall nutrients is probably the optimal strategy for dealing with widespread nutrient deficiencies, as it has the best chance of reaching the population segments most at risk, as contrasted with attempts at changing individuals' food choices or relying on voluntary supplement taking. Given the wide array of potential calcium fortificants and fortification levels, there is not much to guide manufacturers interested in improving the nutritional value of their products. In this review, we assemble the calcium salts/complexes that have been used or proposed for use as fortificants and describe certain of their measured characteristics that relate to incorporation into foods, particularly what is known of their absorbability. The calcium salts most commonly used as supplements or fortificants exhibit similar absorbability when tested in pure chemical form. Choice of salt will depend mainly upon cost, compatability with the manufacturing process, and consumer acceptability. However, interaction with food, tablet, or beverage matrices can degrade intrinsic absorbability substantially. As a consequence, each product must be explicitly tested to establish the degree to which its calcium is available to consumers.


In 1982, the American Society for Bone and Mineral Research held a press conference on osteoporosis, calling the public's attention to the protective role of a high calcium intake. Media coverage of that conference led in the mid-to-late 1980s to what Newsweek Magazine called “The Calcium Craze.”“Craze” apart, the underlying science was sound, and no less than 3 subsequent NIH Consensus Development Conferences recommended increased calcium intakes for most age groups in the population, precisely to lower the risk of osteoporosis.

Yet despite more than 20 y of awareness of the importance of calcium to health, U.S. calcium intakes remain suboptimal. As recently as 2004 the U.S. Surgeon General, in his report on bone health (USDHHS 2004), stated that “calcium has been singled out as a major public health concern today because it is critically important to bone health and the average American consumes levels of calcium that are far below the amount recommended.”

As America's chief health educator, the Surgeon General gives Americans the best scientific information available on how to improve their health and reduce the risk of illness and injury. Earlier, the 1988 Surgeon General's Report on Nutrition and Health (USDHHS 1988) had suggested that calcium be considered for food fortification. Today calcium intake is recognized to be important not only for bone health but also for lowering the risk of a host of other disorders as well, ranging from hypertension to colon cancer to obesity.

Among the earliest studies of calcium metabolism and bone health (and one of the longest running, continuously supported projects in NIH history) is the Omaha Nuns Study conducted by the Creighton Univ. Osteoporosis Research Center. Begun in 1967 and continuing to the present, 40 y later, the project studies calcium metabolism in the population at risk for osteoporosis before and after manifestation of the bone disorder occurs.

The Omaha Nuns study literally “wrote the book” on the operation of the calcium economy in mid-life women. Its findings have provided the principal scientific basis for the NIH recommendations for adult calcium intake. The project has also established the gold-standard measurement of calcium absorption using a radioactive tracer method. Absorption of calcium from over 50 foods, beverages, supplements, and fortificants has subsequently been measured in human subjects at the Creighton Univ. Osteoporosis Research Center.

The Need for Calcium Fortification of Foods and Beverages

The principal reason for fortifying modern diets is the decline in per capita calorie expenditure in the industrialized nations and the corresponding need to reduce energy intake. This fact, coupled with an increase in energy-dense, nutrient-poor foods, results in substantial shortfalls of many key nutrients in contemporary diets, in contrast with the diets of just 2 generations ago.

Fortification of foods with nutrients is probably the optimal strategy for dealing with wide-spread nutrient deficiencies, as it has the best chance of reaching the population segments most at risk, as contrasted with attempts at changing individuals' food choices or relying on voluntary supplement taking. Successful instances of fortification-induced eradication of disease include the conquest of pellagra in the United States by niacin fortification of white flour, the virtual elimination of goiter by iodination of salt, and the reduction of neural tube defects by folate fortification of cereal grain products. Two of these involved mandatory fortification, while iodination of salt remains voluntary in the United States (though mandated in Canada and many European countries).

While there is a pending citizen petition before the FDA for mandatory calcium fortification of most cereal-based products, mandatory fortification seems unlikely at the present time. It took the FDA 24 y to require folate fortification after the 1st request to do so from the Natl. Academy of Sciences. Hence the emphasis for calcium will likely remain on voluntary fortification, with food manufacturers confronting an absence of standards and a multiplicity of options, both for calcium source and for level of fortification.

Early efforts of food manufacturers in the 1980s included such notable examples as Procter & Gamble's “Citrus Hill Plus Calcium” and Continental Baking's “Wonder Calcium” (both subsequently withdrawn, though for quite different reasons). Similar examples appeared on supermarket shelves through most of the 1990s, but few lasted long. Then, beginning in 1999, calcium-fortified brands appeared in large numbers and have, for the most part, been apparent commercial successes. Over 1100 calcium-fortified products (Clark 2006) were introduced in a recent 5-y period (more than two-thirds of these in the beverage and snack categories). Whatever their commercial success, several of these products are what must be judged fortification failures (Heaney and others 2005a, Heaney and Rafferty 2006) and need reformulation.

Moreover, fortification has only partially penetrated most major food categories, taking all brands into consideration. Thus, in the category of ready-to-eat cereals, many of General Mills™ breakfast cereals are now fortified with calcium, whereas those produced by Kellogg's™ and Post™, for example, are not. Continental Baking has reintroduced a calcium-fortified Wonder Bread™ (though at a lower level than its earlier attempt), but many breads remain unfortified. Thus, population penetration remains spotty.

The importance of calcium to the consumer is reflected in consumer attitudes and beliefs, if not in food choices. A survey of consumer preferences for minerals in beverages (Rokosh 2005) reveals:

  • • The general public rates the claim “Good Source of Calcium” as most important when choosing fortified beverages
  • • 70% have consumed fortified beverages
  • • 23% once per day or more
  • • 24% 1 to 6 times per week
  • • 30% believe such beverages are vital to health
  • • 67% see themselves as deficient in calcium, magnesium, iron, zinc, and/or potassium
  • • 69% consider calcium one of the most important ingredients in fortified beverages, necessary for maintaining a healthy lifestyle

Given the wide array of potential fortificants and fortification levels, there is not much to guide manufacturers interested in improving the nutritional value of their products. In this review, we assemble the calcium salts/complexes that have been used or proposed for use as fortificants and describe certain of their measured characteristics that relate to incorporation into foods, particularly what is known of their absorbability. We also call attention specifically to instances of fortification failures that should, if possible, be avoided in future attempts at fortification.

Absorbability: Meaning and Importance

With respect to absorbability, it should be noted at the outset that, while for disorders such as osteoporosis, absorption is of paramount importance, unabsorbed calcium in the gut exerts useful functionality in its own right (blocking absorption of potentially harmful by-products of digestion) (Heaney 2003a). Hence emphasis on absorbability as the primary criterion of a desirable fortificant may not always be pertinent. Absorbed and unabsorbed calcium are both important. A calcium-rich diet supports both.

Absorption actually has 2 meanings that need to be distinguished. One is the unidirectional movement of calcium from the gut lumen through the intestinal mucosa into the blood. The other is the algebraic sum of this 1st movement and a contrary flux of calcium from the blood into the lumen in the form of digestive secretions and sloughed mucosal cells. The first, gross absorption, is a feature of the calcium source and is the most relevant measure to use in characterizing fortificants and fortified foods. The second, net absorption, is heavily dependent upon host factors and is the measure of greatest nutritional relevance to the individual.

Absorption of calcium, as for many divalent cations, is relatively inefficient. In healthy adults, at typical calcium intakes, gross absorption averages in the range of 25% to 35% of typical ingested calcium loads, while net absorption averages about 10% to 12%. Additionally, absorption is a function of load size, varying inversely as the logarithm of the test load (Heaney and others 1990a). For loads as small as 10 to 20 mg, gross absorption will exceed 80% of the ingested load, while for the same source, at a load size of 1000 mg, gross absorption will typically average about 20%. These numbers illustrate an often misunderstood feature of absorption. Eighty percent (80%) would seem to be clearly better than 20%. However, 80% of a 20 mg load means that only 16 mg is transported from the gut lumen into the blood, while 20% of a 1000 mg load results in 200 mg calcium actually entering the blood stream. It is the amount absorbed, not the fraction (or percent) absorbed that has nutritional significance.

It is beyond the scope of this review to describe methods of measuring absorption, but it is necessary here to point out that the isotopic tracer methods, when correctly implemented, directly measure gross absorption fraction, while the pharmacokinetic and intestinal balance methods reflect, instead, the net absorbed quantity. The distinction is important because different sources may require different methods, and to compare sources it is necessary not only that they be tested at the same load size but by the same method as well. When the emphasis is on the calcium source (that is, the fortificant or the fortified food) as in this review, gross absorption fraction will be the preferred measure.


Literature search

Medline was searched for all studies of calcium absorption in humans, with primary emphasis on the absorbability of specific sources. A total of 45 articles were found. After deciding whether they met the inclusion criteria (see below), these were combined with 31 studies performed in the authors' laboratory and in that of TNO Nutrition and Food Research Inst., Zeist, the Netherlands. Several of these provided data for multiple salts. Some few of these had, themselves, been previously published (but most of which had previously appeared only in project reports or internal communications to sponsors).

Inclusion criteria

We included in this study primarily studies in which gross absorption was measured from a chemically defined calcium load. This allowed both quantitative estimation and comparison of the absorbability of various calcium salts, and estimation of the effect of incorporating the salt into various food/supplement matrices. In studies in which a reference material such as milk or precipitated calcium carbonate was used, the absorbability of a test calcium source was also expressed relative to that of the referent in the same test participants (that is, a quotient was calculated for the absorbability of the test calcium source divided by that of the referent calcium source).

All told, we were able to identify 55 discrete studies (or arms of studies) of specific calcium salts, measured either in pure form or in a food or supplement matrix. Most of those studies evaluated at least 2 sources, and several up to 4 sources. We excluded 1 study employing the gastrointestinal lavage method, 4 using recovery of tracer solely from fecal collections, 3 in which the only measurement was of urine calcium, 3 biodynamic studies measuring only change in serum parathyroid hormone, 1 study measuring whole body calcium retention, 2 studies in children, and 1 study involving patients with liver disease. This makes a total of 14 studies not yielding usable data. Additionally, there were 6 studies using pharmacokinetic methods that while useful for comparing products, produce results that cannot be straightforwardly converted to fractional absorption values. These studies were used only to the extent that they contained a referent allowing expression of absorption relative to the referent. Finally, 2 papers consisted of meta-analyses that, of course, contained no original data.

Studies involving fecal recovery of calcium tracers were eliminated because such studies measure net rather than gross absorption, and because, when properly executed, such studies require prolonged fecal collections for accuracy, and most such reports did not meet this criterion. Although a method exists (Heaney 2003b) for converting data from pharmacokinetic studies into fractional absorption values, this method has not been validated across large ranges of load size, and hence studies involving pharmacokinetic or intestinal methods were included only when comparisons were made between sources, permitting calculation of relative values when a suitable referent material had been included in the study. Two pharmacokinetic studies were excluded because serum calcium measurements were obtained for fewer than 5 h. In all, 18 published reports were included in this analysis (Recker 1985; Smith and others 1987; Miller and others 1988; Heaney and others 1989, 1990a, 1999, 2000, 2001, 2003, 2005b; Andon and others 1996; Nickel and others 1996; Weaver and others 1996, 2002; Martin and others 2002; Brink and others 2003; Rosado and others 2005; Heaney 2006).

In this respect, the word “test” is used to designate an individual measurement in a single individual, and the word “study” to designate a group of individual tests on the same calcium source as part of a single investigative protocol. If more than 1 calcium source was evaluated in a single protocol, we counted them as separate studies.

Analytical methods

For the 30 studies performed in the authors' laboratories and not previously published in the scientific literature, the test procedure followed published methods (Heaney and Recker 1985, 1988), which, briefly, were as follows. All tests were performed in healthy adults; all studies had been approved by the Institutional Review Boards of Creighton Univ. or TNO Nutrition and Food Research Inst.; and all subjects gave written consent. All tests were carried out in the morning after an overnight fast, usually with the test calcium source being ingested in the middle of a light breakfast meal. For those studies in which the protocol required that the substance be ingested without food, on an empty stomach, the results were segregated and separately analyzed because of the previously established fact that calcium absorption for many sources tends to be enhanced in the presence of food (Heaney and others 1989). (This latter effect probably relates to the influence of food on gastric emptying, rather than to any feature of the calcium source itself.) When the test method was pharmacokinetic in character, multiple blood samples were taken over a 9- to 24-h period following ingestion of the test source. The tracer-labeled sources were tested either by the double-tracer, gold-standard method (de Grazia and others 1965), or by a single tracer method requiring a single blood sample obtained 5 h after ingestion of the test source. The latter procedure had been explicitly validated against the double-tracer method (Heaney and Recker 1985, 1988).

For tracer-labeled experiments, the source was generally intrinsically labeled by addition of a suitable calcium isotope to a solution of the chemical salt concerned, prior to its precipitation and formulation into a specific dosage form (for example, capsule, tablet, food fortificant). This method ensured uniform distribution of the tracer through the carrier atoms of the calcium source. In beverages in which the calcium salt was in solution, extrinsic labeling was implemented by direct addition of a suitable quantity of the tracer to the source, allowing equilibration for 17 h at 4 °C before dosing. As the protocols for several studies called for different ingested calcium load sizes, fractional absorption values were normalized to a 300 mg load by using the following equation


derived from a study of the effect of load size on absorption fraction (Heaney and others 1990a). In this equation, AbsFx= gross absorption fraction, and load = the calcium load of the test source in milligrams.

Statistical analysis

As the purpose of this analysis was primarily descriptive, simple summary statistics were computed using the routines in Microsoft Excel (Microsoft Corp., Redmond, Wash., U.S.A.). For studies of the same calcium salt, means and standard deviations for each report and/or study arm were pooled to develop an aggregate estimate of the results across all of the studies concerned, with weighting for sample size. Where quotients were available for contrasting absorbability from a particular source and that of a referent, a single-sample t-test was used to evaluate whether the ratio departed significantly from a value of 1.0 (= equivalent absorbability).

Comparison of Calcium Salts

Our comparative analysis examines 11 calcium salts tested in various configurations as single salts, compound/combination salts, pure salts, salts prepared as pharmaceutical formulations, and/or salts within a food/beverage matrix. Some tests were conducted without food on an empty stomach, and others were co-ingested with a standardized meal. Roughly half of the tests include a milk comparator and 20%, a precipitated calcium carbonate comparator.

Table 1 sets forth the various calcium salts studied under either the tracer or pharmacokinetic methods of measurement, together with test conditions, number of subjects, and subject characteristics. Roughly 70% of the over 2100 individual tests were carried out in premenopausal women. Calcium carbonate accounts for the majority mainly because CaCO3 was often chosen as the referent in a crossover study design. In cases where both men and women have been tested on the same salt, both tend to exhibit similar absorption values. Similarly, the data show that there is little difference in the absorption efficiency of pre- and postmenopausal women. Hence, in presenting the results that follow (Table 2 to 5), we aggregate studies of the same salt from men and both pre- and postmenopausal women.

Table 1—.  Numbers of individual tests by salt and by test conditions.
SaltSubjects testedTest conditions
nPre-menoPost-menoMenPure saltbNo mealw/meal
  1. aCaLactate Gluconate; CaLactate MCP/DCP/TCP; Ca/P/K/Mg/Na; Ca/P/K/Citrate.

  2. b“Pure salt” means the salt ingested in a gelatin capsule, not incorporated into a food-like matrix, and irrespective of co-ingested food.

Calcium carbonate927683234 10112137790
Tricalcium phosphate137116 10 11 10 137
Dicalcium phosphate 36 36  36
Calcium citrate116 17 89 10 44  7109
Calcium sulfate 34 34   9  9 25
Calcium lactate 30  30  30 10 20
Calcium citrate malate558434124  10 558
Calcium hydroxide 42 18 24  42
Bisglycinocalcium 13 13 13 13 
Calcium glycerophosphate 24 24  24
Combination saltsa184 94 46 44 10 184
Total2101 1469 557 752381761925 
Table 2—.  Fractional absorption of various salts co-ingested with a low-Ca test meal.
SaltaN (studies)N (subjects)Weighted adj absFxbWithin-study SDBetween-study SD
  1. a“Pure salt” means the salt ingested in a gelatin capsule, not incorporated into a food-like matrix, and irrespective of co-ingested food.

  2. “In matrix” refers to incorporation of the salt into a food during its preparation or to admixture of the salt with the components commonly encountered in nutritional supplement tablets or capsules, in both cases prior to fabrication of the final ingested form; weighting is by sample size.

  3. bWhere load size in tests concerned differed from the standard load of 300 mg, resulting values adjusted to 300 mg (see text).

Calcium carbonate—pure salt 5 710.36820.09810.0565
Calcium carbonate—in matrix225570.25800.07520.1094
Calcium citrate—in matrix 3 370.38170.08480.0029
Tricalcium phosphate—in matrix 71370.24290.09750.0212
Dicalcium phosphate—in matrix 2 360.26100.07740.0178
Calcium sulfate—in matrix 2 250.40020.10090.0577
Calcium lactate—in matrix 3 300.33000.09060.0708
Calcium glycerophosphate—in matrix 1 240.27140.0720
Calcium hydroxide—in matrix 2 180.29160.03440.0236
Calcium citrate malate—in matrix175080.33800.09090.0355
Table 3—.  Comparative salt absorbability—salt-to-milk quotient.
 N (subjects)All subjects
  1. aSignificantly different from 1.0.

Pure salts, fasting without meal 
 Calcium carbonate41 0.999
 Calcium citrate 7 0.935
 Calcium lactate10 1.466a
 Calcium sulfate 9 0.985
Pure salts, co-ingested meal 
 Calcium carbonate34 1.116a
 Tricalcium Phosphate10 0.789a
 Calcium lactate20 1.033
 Calcium citrate malate10 1.007
 Calcium lactate gluconate10 1.092
Salts in food/supplement matrix, co-ingested meal 
 Calcium carbonate168  0.987
 Tricalcium Phosphate137  0.834a
 Dicalcium Phosphate36 0.965
 Calcium sulfate25 1.094
 Calcium citrate malate228  1.114a
 Comb. (Ca Lactate MCP/DCP/TCP)84 0.864
 Comb. (Ca, P, K, citrate)33 0.817
Table 4—.  Comparative absorbability—salt-to-calcium carbonate (pure CaCO3) quotient.
Salt/productQuotient—all subjectsN
Research CaCO3 product A0.61126
Research CaCO3 product B0.78716
Commercial CaCO3 product A0.88026
Commercial CaCO3 product B0.94124
Commercial calcium citrate product0.92524
Table 5—.  Comparative absorbability—calcium-citrate-malate (CCM) in various beverage matrixes.
Beverage matrixMean absFxN
CCM in water0.278410
CCM in orange juice0.342457
CCM in apple juice0.402457

Table 2 sets forth the mean absorbability values of 9 salts that have been tested either in pure form (that is, without excipients and without prior incorporation into a food or beverage matrix) or studied after fabrication as supplements or fortified foods and beverages. Mean fractional absorption values of the 9 salts tested span a range from 0.24 to 0.40. The mean values of the table, however, obscure a high degree of variance for some of the calcium salts. The table contains 2 columns labeled “SD.” The first is the within study standard deviations around the respective means, and the second is the standard deviation for the aggregate of all the component study means. (With 1 salt, calcium glycerophosphate, for which there was only a single study, only the within-study SD is shown.) Had all the studies been random samples of a single population of absorption values, the SD of the means (that is, the between-study SD) would be predicted to be smaller than the within-study SD, by a factor of 1/?N. In the case of CaCO3 (in matrix), however, the between-study SD is not smaller but is substantially larger than the within-study DS, indicating the presence of substantial heterogeneity among the sample means.

This effect is most apparent for a single salt (for example, calcium carbonate) and suggests either (1) differences in the absorptive performance of the various test participant panels or (2) differences in test conditions (for example, with and without a test meal) or (3) matrix effects on inherent absorbability. Figure 1 captures the mean values for 25 studies of calcium carbonate and shows 2 points very clearly: (1) absorption from an empty stomach is less efficient than that from the same salt fed with a light breakfast; and (2) there is a 2-fold spread of fractional absorption values for calcium carbonate once the salt is incorporated into various food, supplement, or beverage matrices, with mean absorbability values ranging from 0.21 to 0.42. The highest value was for a product fully as well absorbed as the pure salt, while the lowest was for a product less than half as well absorbed. Much of this variability must reflect interaction of food, supplement, or beverage constituents with the added salt, altering its inherently high absorbability.

Figure 1—.

Mean values for absorption fraction from 25 studies of calcium carbonate absorption. Point plot of mean (± 1 SEM) absorbability values for the pure salt co-ingested with a meal (left column), for the pure salt without any co-ingested food (middle column), and for the salt fabricated into various food, supplement, or beverage matrices (right column). (Copyright, Robert P. Heaney, 2007. Used with permission.)

A greater degree of uniformity is shown for the majority of the other calcium salts co-ingested with a meal. Note that the between-study SD for many of these salts is smaller than for calcium carbonate, indicating that the different matrices did not appreciably alter the absorbability of the noncarbonate salt concerned (Table 2).

These discordances in calcium bioavailability highlight the reality that bioavailability cannot be predicted based on the current chemical knowledge of the source, but must be directly tested. Nor can bioavailability of a salt in 1 matrix be extrapolated to other untested matrices.

The ideal model for testing calcium bioavailability is a crossover design in which the source calcium in question is tested against a referent calcium source such as dairy milk. Within a subject, calcium absorption efficiency is highly consistent (Heaney and others 1990b) and the milk (or other) comparator effectively eliminates the interindividual variation. Testing calcium absorption against a consistent comparator such as milk or precipitated calcium carbonate standardizes the test and allows for credible conclusions (that is, comparing “apples” to “apples”). Table 3 sets forth calcium absorbability not as absolute fractional absorption (AbsFx) but as a quotient relative to milk, and Table 4 does the same for a precipitated calcium carbonate referent. Absorption equal to that of the referent calcium has a value of 1.0. The following conclusions may be drawn:

  • • Most calcium salts ingested under fasting conditions show reduced absorbability relative to the same salt ingested with food, with the exception of calcium lactate, which, in a single study, demonstrated better absorption from a fasting than from a fed state.
  • • Most pure calcium salts co-ingested with a meal show improved absorbability, relative to milk, with the exception of tricalcium phosphate. While this observation does not yield any actionable information for consumers, as consumers are unlikely to obtain supplemental calcium in the form of pure salts, the food and beverage industry may engage the challenge of identifying food or beverage vehicles that preserve the pure salts' absorbability characteristics.
  • • As single salts, calcium lactate and calcium lactate gluconate tested high relative to milk. In calcium lactate–calcium phosphate combination salts, however, absorbability was low relative to milk. Unfortunately, the use of a combination precludes evaluation of the separate contributions of its constituents.
  • • Among the 12 calcium salts co-ingested with a meal, none of the 6 exhibiting improved absorption (that is, quotients > 1.0) contained a calcium phosphate salt, while 5 of the 6 exhibiting reduced absorption (that is, quotients < 1.0) had phosphate as a component of the salt. Although it is beyond the scope of this analysis to explore the totality of the calcium economy, this less efficient absorption of the phosphate salts is not necessarily a negative consideration. Calcium absorption is just 1 component in the calcium economy. Calcium is retained in bone as a calcium phosphate salt and the total nutrient intake of phosphorus is critical for the accumulation of bone mineral. Though phosphorus is less likely to be as severely underconsumed as calcium in adults, certain segments of the population, such as elderly women taking bone-building medications, may benefit from the phosphorus contained in the calcium phosphate fortificants.
  • • Overall, while a few salts seem to stand out (for example, CCM, calcium lactate, and CaSO4 on the high side, and TCP on the low side), most salts exhibit roughly similar bioavailability and tend to fall within ± 10% of the absorbability of milk calcium.

Clearly, what is true for pure calcium salts cannot be extrapolated or generalized to the same salts tested as calcium supplements or food fortificants. The pharmaceutical formulation, which includes a broad range of factors such as excipients (flavorings, coatings, sugars, starches, lubricants), encapsulation, and granulation, can result in great differences in bioavailability. The effect of various tableting formulations on calcium bioavailability both from test products and from marketed calcium supplements is extreme, resulting in a reduction in absorbability of from 7.5% to 39% in a tableted CaCO3 preparation, relative to the pure CaCO3 salt.

Another source factor that may influence bioavailability is the presence of enhancers or inhibitors in the source itself. This is particularly significant in the instance of calcium added as a fortificant to foods natively absent any calcium. For example, calcium-citrate-malate as a fortificant in fruit juice shows superior absorbability in an apple juice matrix compared to the same complex salt in an orange juice matrix when tested in the same individuals, and poorer absorbability when tested as a pure salt in water (Table 5). In an animal model (not included in the studies examined here), Mehansho and others (1989) found good absorbability for CCM calcium in orange and grapefruit juices, but poor absorbability for the same salt in lemon juice.

Barriers to Effective Calcium Fortification

In the Surgeon General's Report on Bone Health and Osteoporosis, previously cited (USDHHS 2004), a substantial shortfall was noted between prevailing and recommended calcium intakes. The report stressed the need to do a better job of applying what we know toward the goal of improving the calcium nutriture of the U.S. population. Food fortification is a way to improve intake of a nutrient with generally recognized effectiveness. Hence better and more widespread utilization of this option for calcium by food manufacturers would seem to be indicated. Some barriers to effective fortification can be identified and need to be addressed. None would seem to be insuperable.

First there is the evident variation in bioavailability between products, highlighted in Table 2 and Figure 1. This can be solved in several ways, but preferably by developing a better understanding of the matrix factors that alter bioavailability. Additionally, bioavailability needs to be tested for each product, as it is difficult to predict, especially for calcium carbonate which, in most other respects, would appear to be a highly desirable fortificant.

The choice of a salt will inevitably depend to a great extent on compatibility with the manufacturing process, on taste effects, and on product stability characteristics. If these factors should dictate choice of a less bioavailable salt, confirmed by testing, the relatively lower absorbability can generally be offset by the simple device of fortifying to a higher level. The focus in all such fortification strategies should be on the quantity of the fortificant absorbed, not the quantity added. Initially, this may create some regulatory problems, but it seems to us that no other approach makes sense.

Related to this matter of bioavailability is another quality-related issue. In beverages, such as soy “milk” and some fortified orange juices, it has been shown that the fortificant tends to settle to the bottom of the carton (Heaney and others 2005a, Heaney and Rafferty 2006). Even vigorous shaking was not sufficient to resuspend the calcium salt in some products. Here bioavailability is moot, as it is ingestion itself that is compromised. In all such instances, the labeled content, while technically accurate, is irrelevant and will likely be misleading to consumers.

Second, and of some contemporary interest, is the issue of whether or not the added calcium will confer a net benefit. The report from the Women's Health Initiative (WHI) of a small increase in kidney stones in women receiving large calcium supplement doses (Jackson and others 2006) has resulted in some cooling of consumer interest in calcium. That reaction is almost surely inappropriate, but needs addressing nonetheless. It is important to understand that the character of the stones reported in WHI has not been identified and that stones were not a planned outcome variable in WHI. Hence, the observation may well be spurious. It is important also to recall that calcium intakes in WHI, in the supplemented arm, were nearly twice currently recommended values, while the purpose of any proper fortification strategy would be to bring individuals up from low intakes to recommended levels, not to exceed them. Finally, it is important also to view the WHI report in the context of the totality of the evidence. In the only published randomized controlled trial of calcium intake in stone-formers, a diet providing calcium at recommended levels resulted in 50% fewer stones than did a calcium-restricted diet (Borghi and others 2002). This finding is concordant with other data showing reduced stone risk as calcium intake rises from low to currently recommended levels (Curhan and others 1993). (The reason is simply that unabsorbed calcium in the digestive residue complexes with, and blocks the absorption of, dietary oxalate which, if absorbed, is a potent stone risk factor.) Hence the available evidence, taken together, indicates that raising calcium intake to recommended levels does not increase stone risk, but reduces it instead.

Finally, it should be noted that a significant nonsource factor affecting calcium absorption is the vitamin D status of the individual. It has long been recognized that vitamin D is necessary for the active transport of calcium across the intestinal mucosa. Only recently, however, has the vitamin D status that fully optimizes calcium absorption been identified and quantified. Serum 25-hydroxyvitamin D concentration (25OHD) is the accepted functional indicator of vitamin D status. Heaney has shown that intestinal calcium absorption improved by 68% when serum 25OHD was raised from 50 to 80 nmol/L in postmenopausal women (Heaney and others 2003). While both values fall within the current laboratory reference range for “normal” serum 25OHD status, these results indicate that improving vitamin D status up to repletion is important for fully optimal calcium absorption efficiency.

Recently, it has been realized that vitamin D status of the U.S. population is at least as insufficient as that of its calcium intake, and hence there is beginning to be interest in vitamin D fortification of suitable foods, as well. While fortification with vitamin D seems a sound strategy for many reasons, the added vitamin D is biologically inactive and does not produce an acute calcium absorptive enhancement effect. Dietary vitamin D must first be absorbed, and then undergo a series of hydroxylations, first in the liver and second in the kidney, to produce the active form of the vitamin that is the regulator of calcium absorption. Vitamin D, as an additional fortificant to a calcium-fortified food or beverage, while doubtless contributing to the improvement or maintenance of the recipient's serum 25OHD status, does not itself directly influence the absorption of the calcium contained in the product. Vitamin D does not, therefore, have to be present in the calcium source in order to insure calcium absorption.

Additional Notes about Effective Calcium Fortification

As already stated, the decision about which calcium source to use in a food or beverage formulation is related to ease of use in that application, taste effects, and stability in the finished food or beverage. But also the decision is based on overall cost and desired health claims. In regard to health claims, in general, calcium is best absorbed by the body when consumed in smaller amounts, taken several times throughout the day. Therefore, added calcium is probably most effective at fortification levels up to 30% DV (daily value), that is, 300 mg per serving. The associated health claims are:

  • • Foods containing 10% of the calcium DV, when compared to a standard serving size of a similar food, can say they are “Calcium Enriched,”“Calcium Fortified,” or have “More Calcium.”
  • • Foods containing 10% to 19% of the calcium DV can say “Contains Calcium,”“Provides Calcium,” or is a “Good Source of Calcium.”
  • • Foods containing 20% or more of the calcium DV can be listed as “High in Calcium,”“Rich in Calcium”, or an “Excellent Source of Calcium.”

In regard to ease of use in an application, calcium carbonate is often used in items where clarity or dry and chalky mouthfeel would likely not be noticeable, such as food bars and cereal. Because calcium carbonate is the least costly calcium source, opaque meal replacement beverages and soy beverages also become a vehicle for this salt, especially in its micronized forms. This approach, however, works best when stabilizers and ingredients to reduce interaction with protein (when present) are employed to keep this insoluble calcium suspended and to minimize protein–calcium complexation and sedimentation. Calcium phosphate is another popular, low solubility calcium source used in foods and nonclear beverages because it can also act as a buffer for control of acidity. Once a formulator moves into clear beverages, the emphasis shifts to a search for a calcium that either can be soluble at a certain pH, such as calcium citrate, or be soluble in a broad pH and concentration range. The latter functionality moves the calcium choices primarily into lactate, lactate gluconate, and gluconate salts of calcium because other highly soluble forms of calcium, such as calcium chloride, have undesirable taste acceptance at usages over 10% DV. Calcium fumarate would appear to be useful for clear beverages for taste and solubility, but dissolution is significantly slower than the lactate or lactate-gluconate salts, and solubility in cold water is lower as well. Finally, concentrated slurries or syrups are difficult or impossible with calcium fumarate while achievable with lactates and lactate gluconates.


Despite more than 25 y of awareness of the importance of calcium to health, and 2 decades of U.S. manufacturers' launching calcium fortified foods, the mean U.S. calcium intakes remain suboptimal. More widespread food fortification with calcium, well executed, offers the promise of effectively dealing with that problem at virtually no cost and essentially no risk.


Partial support for this review provided by PURAC America, Food and Nutrition, is gratefully acknowledged.