Although calcium and vitamin D have long been known to be important for long-term fracture risk, more recent research has shown that bone mass is, in fact, sensitive to a wide variety of nutritional exposures. Dietary intake is a centrally important modifiable factor in the development of peak bone mass and in the protection against bone loss with aging. Because bone undergoes continuous remodeling, an adequate supply of nutrients is required to support
bone formation and retention. The bone matrix is composed of calcium, phosphorus, protein, and trace minerals, including magnesium, and these are of primary importance. However, additional dietary components affect the remodeling balance through effects on calcium regulation, inflammation, DNA methylation, and other regulatory processes that stimulate bone resorption or formation. Understanding these relationships is important, because it points to dietary quality as a critical factor in bone status, as opposed to an earlier focus mainly on calcium supplementation for prevention of bone loss (
21,
22).
Calcium
Calcium is the major mineral component of bone mass, and nearly 99% of the calcium in the adult human body is contained in bones in the form of hydroxyapatite. Children need relatively large amounts of calcium to lay down new bone with rapid growth. The 1997 Food and Nutrition Board set adequate intakes at 1300 mg (32.5 mmol) per day for children 9 to 18 years old, for maximizing peak bone mass to protect against osteoporosis later in life (
23). However, supplementation studies in children have shown mixed results. A 2000 review of calcium supplementation and bone concluded that calcium contributed to higher BMD primarily at cortical bone sites, was most effective in populations with low baseline calcium intake, and was more effective in pubertal than prepubertal children (
24). A more recent review of 19 calcium supplementation trials in 2859 children found that calcium supplementation had a small effect on upper limb BMD but no effect on femoral neck or lumbar spine (
25). No evidence was found for effect modification by sex, baseline calcium intake, pubertal stage, ethnicity, or level of physical activity; and it was concluded that calcium supplementation in children is unlikely to reduce the risk of fracture either in childhood or later adulthood.
A broader review of calcium trials, including adults, found that all but 2 of 52 trials showed better bone balance with calcium intervention, greater bone gain during growth, reduced bone loss with aging, or reduced fracture (
26). In contrast, a more recent metaanalysis of 4 clinical trials, with 6504 subjects and 139 hip fractures, calculated a pooled RR between calcium and placebo of 1.64 (95% confidence interval, 1.02, 2.64), indicating higher, not lower, risk (
27). A follow-up after completion of a large, 3-year, placebo-controlled trial of calcium and vitamin D supplementation in older men and women showed that most of the BMD benefits accrued during the trial were lost 2 years after supplementation ended (
28). A lack of efficacy of calcium supplements was also seen in the Women’s Health Initiative (WHI), in which 36,282 postmenopausal women 50 to 79 years of age were randomly assigned to 1000 mg calcium and 400 IU (10 μg) vitamin D3 daily, versus placebo, and followed for 7 years. Although the calcium-vitamin D supplementation resulted in small improvements in hip BMD, it did not reduce hip fracture risk in all healthy postmenopausal women. However, subgroup analyses showed that supplemented women more than 60 years old, but not younger women, did have a lower risk of hip fractures (
29). Together, these studies question the conventional wisdom that calcium supplementation has major protective effects against fracture risk. However, few studies have considered baseline status. It seems likely that those with inadequate calcium intake would benefit more from supplements than those with adequate intakes.
It is also likely that dietary sources of calcium may be more effective than calcium supplements. An NHANES III follow-up analysis found that low recalled milk intake during childhood and adolescence was associated with significantly lower hip BMD and a doubling of fracture risk among women 50 years of age and older (
30). A large 5-year study of British adults found that fracture risk was 75% higher among women with baseline calcium intakes less than 525 versus 1200 mg or more per day, and the association was stronger for women less than 50 years of age than for older women (
31). However, in the Nurses’ Health Study (NHS), women who reported drinking two or more glasses of milk per day versus up to one per week did not differ significantly in hip fracture incidence (
32).
Intervention studies with calcium-rich foods have shown beneficial effects on bone. In one, spinal bone loss was significantly lower in premenopausal women who used dairy foods to raise calcium intake from 900 to 1500 mg (22.5 to 37.5 mmol) per day, relative to controls (
33). In another, three additional servings of yogurt per day led to significant reduction in urinary excretion of bone turnover markers in older women (
34). Calcium in foods like milk and yogurt may be used more effectively than supplements because it comes packaged with other important nutrients that work together, including vitamin D, protein, potassium, and magnesium.
Phosphorus
Phosphorus is essential for bone, but too much phosphorus in combination with low calcium intake can lead to reduced calcium bioavailability and potential bone loss. Although uncommon, phosphorus deficiency can lead to reduced mineralization and bone resorption. Deficiency has been seen in older adults with malnutrition, intestinal malabsorption, or long-term use of medications that bind phosphorus, including antacids (
35). In the general population, excess phosphorus is more of a concern than deficiency. The US diet tends to be high in phosphorus relative to calcium. Mean intakes of phosphorus in the NHANES 2007 to 2008 study were 1123 and 1550 mg/day, for women and men, relative to a recommended dietary allowance (RDA) of 700 mg; whereas mean intakes of calcium were 833 and 1038 mg for women and men relative to an RDA of 1000 to 1200 mg (
36,
37).
Excess phosphate form complexes with calcium that interfere with calcium absorption, which may in turn lower serum calcium and lead to secretion of parathyroid hormone (PTH), lower 1,25(OH)2D production, lower
intestinal calcium absorption and, consequently, bone resorption to release calcium from bone (
38). Short-term metabolic studies have documented some of these mechanisms (
39,
40).
One major source of excess phosphorus in the US diet is phosphoric acid from cola drinks. Two studies in teenage girls found that cola consumption significantly increased the odds of fracture (
41). In the FOS, women consuming cola daily had significantly lower hip BMD than those who consumed cola less than once per week (
42). In contrast, a short-term metabolic study, showed negligible effects of phosphoric acid-containing beverages on urinary calcium excretion, and concluded that the effects seen in observational studies may be caused by milk displacement (
43). However, milk displacement did not explain the significant negative effects seen for cola in the FOS, and there was no effect of other soft drinks. It is likely that regular exposure to phosphoric acid may cause small BMD losses, which accumulate to measurable losses over time.
Magnesium
Magnesium is important to the formation of pure hydroxyapatite and may enhance bone strength through its role in crystallization (
44). It also is known to regulate active intestinal calcium transport. In animal studies, experimental magnesium deficiency decreased bone volume and trabecular thickness, bone mass, PTH, 1,25(OH)
2 vitamin D concentration, and osteoprotegerin (OPG), whereas it increased receptor activator of nuclear factor B ligand (RANKL) and osteoclastogenesis (
45,
46,
47,
48).
Magnesium concentrations were significantly lower in women with osteoporosis than those with normal bone mass (
49). In observational studies, magnesium intake was significantly positively associated with BMD, and protective against bone loss (
50,
51,
52). This is important because magnesium intakes tend to be consistently low; daily median intakes from NHANES data ranged from 177 mg among African-American women to 326 mg among non-Hispanic white men, relative to recommendations of 320 and 420 mg for women and men, respectively (
53). Some intervention studies with magnesium have shown benefit on bone mass in adolescent girls (
54), in suppressing bone turnover markers in young men (
55), and protecting against bone loss in osteoporotic women (
56), findings suggesting that this important mineral may be underappreciated for its important role in bone health. However, too few randomized controlled trials with magnesium exist to support widespread use of magnesium supplementation to prevent osteoporosis.
Potassium
Potassium promotes renal calcium retention and is also important in neutralizing the acid load of most diets, which may protect against calcium loss from the bones. Potassium administration increased serum osteocalcin concentration and decreased urinary hydroxyproline excretion (
57). Several population-based studies have demonstrated protective associations between potassium intake and bone status. In premenopausal women, a difference of 8% in femoral neck BMD between the highest and lowest quartiles of potassium intake was seen (
58). In perimenopausal and early postmenopausal women, dietary potassium was associated with lower bone resorption and greater BMD (
57). In older adults in the FOS, potassium showed protective associations with BMD in men and women at baseline, and with lower BMD loss over time in men (
50). In another study of elderly women, higher urinary potassium excretion at baseline was associated with 4% greater total body BMD and 11% greater trabecular BMD at 5 years (
59). One author noted that relative to preagricultural humans, the modern human diet is deficient in potassium (2500 mg versus 7000 mg/day) and contains excess sodium (˜4000 mg versus 600 mg/day) (
60). This combination may have particularly negative effects on bone.
Sodium
Sodium intake in the United States is considerably higher than recommended. NHANES 2007 to 2008 data showed that in contrast to recommendations of approximately 1500 mg, mean daily sodium intakes in US adults were 4043 mg for men and 2884 mg for women (
36,
61). This likely contributes to renal calcium excretion. Studies have shown that each 1000 mg of additional sodium was associated with a 20-mg increase in urinary calcium loss—the amount likely to be absorbed from 80 mg of dietary calcium (
62), and consequently with lower BMD. The optimal intake balance for protecting bone was approximately 1000 mg calcium and less than 2000 mg sodium per day.
The effect of sodium also may depend on potassium exposure. A metabolic study found that postmenopausal women given 5175 mg of sodium per day had increases in urinary calcium and N-telopeptide, whereas those given sodium plus potassium citrate had decreases in urinary calcium and no increase in N-telopeptide (
63). In the Dietary Approaches to Stop Hypertension (DASH) sodium trial, a diet high in fruit, vegetables, low-fat dairy, and thus high in potassium, was randomly assigned, versus a control diet, for 30 days. The DASH diet significantly reduced serum markers of bone turnover and additionally reducing sodium led to further reduction in serum osteocalcin decreased PTH in the control group, and lowered urinary calcium in both (
64). In another study of postmenopausal women who reduced sodium intake to less than 2000 mg/day for 6 months, urinary calcium excretion and bone turnover markers were reduced (
65). However, other studies have been less clear about the effects of sodium on bone (
66,
67); and one study showed no adverse effect on BMD of 3000 mg of sodium, compared with 1500 mg/day, when participants were supplemented to ensure adequate calcium and vitamin D intakes (
68).
Fluoride
Fluoride has long been known to prevent tooth decay and has been added to most water supplies in the United States. Fluoride substitutes for the hydroxyl group in hydroxyapatite, forming fluorapatite. Fluoride has been shown to result in bone with larger crystals and higher BMD, but lower elasticity (
69). The effect of fluoride on fracture has been controversial, with both protective effects (
70) and increased risk (
71,
72) being reported. In the largest randomized, placebo-controlled trial of sodium fluoride in postmenopausal women with osteoporosis, spine BMD increased, but so did vertebral fracture risk (
71). A metaanalysis of 25 studies showed that fluoride treatment increased spine and hip BMD, but with no effect on fracture risk. However, the protective effect was seen with low doses (≤20 mg/day of fluoride equivalents) (
73). A more recent comparison of bone tissue from individuals in municipalities with and without fluoridated water, showed no differences in the physical characteristics of bone (
74). Fluoride supplementation, either in short- or long-acting forms, is not approved by the US Food and Drug Administration for the prevention or treatment of osteoporosis.
Iron
Iron is an important cofactor for hydroxylases in collagen formation. Both low iron intake and iron overload have been negatively associated with bone. Iron overload has been associated with low BMD in patients with genetic hemochromatosis and with African hemosiderosis (
75,
76). However, low iron is more of a concern in the general population. Rats fed iron-deficient diets showed compromised bone morphology, strength and density, and decreased serum osteocalcin (
77,
78). Studies in postmenopausal women have reported that higher iron intake was associated with greater baseline BMD (
79) and, prospectively, with lower loss of BMD in a subset of women using hormone replacement therapy and taking 800 mg calcium per day (
80). In contrast, however, another study showed no association between iron status and BMD in women (
81).
Silicon
Silicon is important for collagen and glycosaminoglycan formation in bone and cartilage, influencing the formation of the organic matrix. Silicon is also a major ion of osteogenic cells. Orthosilicic acid, the form of silicon absorbed in the diet, appears to be associated with bone formation through increased synthesis of collagen type I and stimulation of osteoblasts (
82,
83). Chicks fed a silicon-deficient diet developed abnormally shaped bones (
84), whereas the addition of silicon to the diet of depleted rats resulted in fewer osteoclasts, increased bone formation, decreased bone turnover, and increased BMD (
85,
86).
Few studies have been conducted in humans, but those studies have shown protective effects. In the FOS, dietary silicon was positively associated with BMD at hip sites for men and premenopausal women but not postmenopausal women (
87). French patients with osteoporosis showed significant improvements in trabecular bone volume with silicon treatment (
88), and femoral BMD increased in female osteoporosis patients given intramuscular silicon, compared with others given fluoride, oral magnesium, or controls (
89). These results suggest that higher silicon intake may be protective of BMD, but more studies are needed to confirm this.
Other Minerals
Copper is a cofactor for lysyl oxidase, which catalyses cross-linking of lysine and hydroxyproline in collagen. Animals with induced copper deficiency have reduced bone strength (
90) and greater bone loss with aging (
91). In women, plasma copper concentration was correlated with lumbar spine BMD (
92), and lower copper status has been noted in elderly patients with fracture, relative to matched controls (
93). A controlled feeding study in men showed increased activity of bone resorption markers when moved from a high (6 mg/day) to low (0.7 mg) copper diet, and this was reversed by returning to the high copper diet (
94). However, another study did not replicate this effect (
95).
Zinc may affect bone through its role in nucleic acid and protein metabolism (
96). Lower serum and bone zinc and higher urinary zinc have been noted in patients with osteoporosis (
97). In animals, zinc increased alkaline phosphatase and DNA synthesis, which may stimulate bone formation (
98). Supplementation with zinc gluconate has been shown to increase alkaline phosphatase activity (
99). In one study, postmenopausal women were randomized to treatment with calcium plus copper and zinc versus calcium plus corn starch. After 2 years, women with usual daily zinc intakes less than 8.0 mg benefited from the copper and zinc supplements, but women consuming adequate amounts of dietary zinc actually lost more BMD than control women (
100).
Boron intake may protect bone by decreasing urinary calcium, phosphorus, and magnesium losses and increasing serum estradiol (
101). In rats, boron deprivation altered trabecular bone and reduced the force needed to break the femur, confirming the importance of boron to cortical bone strength and trabecular bone microarchitecture (
102). The systemic administration of boric acid reduced alveolar bone loss in periodontal disease in rats (
103). However, no randomized trials of boron supplementation to prevent bone loss or prevent fractures exist.
Strontium has similarities to calcium, and it has received increasing interest as a treatment for osteoporosis. Doses of 1 to 2 g/day of strontium ranelate for 2 years or longer increased BMD in postmenopausal women by 2% to 3%, relative to placebo (
104), and reduced both vertebral and nonvertebral fracture risk (
105,
106). The increase in BMD is predictable and occurs in all treated individuals because of the ability of strontium to incorporate within the hydroxyapatite crystal. However, fracture risk cannot
be predicted based on this increase in BMD. A metaanalysis of two phase III clinical trials showed that strontium ranelate was associated with 31% fewer osteoporotic clinical fractures and 40% fewer morphometric vertebral fractures (
107). Bone biopsies showed that strontium was incorporated preferentially in newly formed bone pockets, and collagen cross-link ratio and bone quality were preserved (
108). Strontium ranelate is approved for the prevention and treatment of postmenopausal osteoporosis in Europe but not in the United States.
Manganese also may contribute to bone status, although it has rarely been examined independently from other trace minerals. In rats, supplementation with manganese led to significantly increased BMD of the lumbar vertebrae and increased serum osteocalcin, suggesting that manganese contributes to bone formation (
109). One study randomized older postmenopausal women to daily supplementation with trace minerals (15 mg zinc, 5 mg manganese, 2.5 mg copper), placebo, 1000 mg elemental calcium, or calcium plus trace minerals for 2 years. Relative to other groups, women receiving calcium plus trace minerals had less bone loss and had increases in spinal BMD (
110).