Should calcium and vitamin D still be a part of our effort to prevent osteoporosis in post-menopausal women?

The publication of a large-scale clinical trial in the New England Journal of Medicine in February, and the ensuing media fracas, has certainly led to a lot of confusion in the clinical community and among our patients. I believe the answer is still a strong, “Yes,” and closer analysis of the data supports this.

The study, led by investigators at the Department of Endocrinology, Ohio State University, surprised women and practitioners throughout the country because overall, there was no significant reduction in hip fractures, the primary outcome measured, but a slight increase in risk of kidney stones among women taking the supplements (Jackson R, LaCroix A, Gass M, et al. N Engl J Med. 2006; 354(7): 669–683). The study, part of the NIH-sponsored Women’s Health Initiative, involved more than 36,000 women, aged 50 to 79, randomized to placebo or 1,000 mg calcium carbonate plus 400 IU of vitamin D.

Although headlines announced negative results, a closer look at the study reveals that there were 12% fewer hip fractures in the group assigned to take the supplements than in the placebo group. This difference was not statistically significant which is why researchers and reporters announced “No benefit.” When women who were noncompliant with the supplement program were excluded from the analytical results, there were 29% fewer fractures in the supplement group when compared to the placebo group

Further, in the postmenopausal women, calcium supplementation decreased bone loss by as much as 50% at non-vertebral sites. The effects were greatest in women whose baseline calcium intake was low, in older women, and in women with osteoporosis. This is in accord with an earlier study showing a significant decrease in vertebral bone loss following supplementation with 1,000–2,000 mg of calcium (Elders PJ, et al. J Clin Endocrinol Metab. 1991; 73: 533–540).

Another interesting twist on the WHI study has to do with how much calcium women were taking at baseline. The mean amount from diet and supplement sources was 1,150 mg per day. Women in both the intervention and placebo groups were allowed to take their own calcium supplements, meaning those in the placebo group were getting an average of over 1,000 mg per day; those in the treatment group were getting an additional 1,000 mg per day plus vitamin D. The point is, calcium intake was already pretty high, even in the placebo group.

We know that women who are already consuming adequate amounts of calcium are less likely to benefit from calcium supplement compared to women with low calcium intake. The data bear this out: in the subjects whose baseline calcium intake was low or moderate, calcium plus vitamin D reduced hip fracture by 22% but by only 12% in those whose baseline calcium intake was 1,200 mg or more.

An equally important finding was a 17% increase in the risk of kidney stones in the treatment group. So while calcium and vitamin D in postmenopausal women with a mean age of 62 years old does support bone health, excessive calcium intake gives no additional benefit, and may induce renal stones in some individuals.

Calcium and vitamin D do still have a place in our practices, but the WHI data do suggest that our approach needs to be more refined and selective than we might have thought. While dietary calcium is essential throughout a woman’s life, it is not effective in preventing accelerated bone loss during menopause. But bear in mind that 10 years post-menopause, calcium supplementation again becomes effective in reducing age-related bone loss (Licata AA. Cleve Clin J Med. 1994; 61: 451–460). We, and our patients, may have been overly optimistic about calcium and vitamin D in women in their 50’s and 60’s. The crucial times for increasing calcium and vitamin D are in prepubertal and teenage years, and then in women over 75, when risk of fractures is significantly increased.

Also consider the fact that calcium salts differ in their content. Calcium carbonate contains 40% calcium in elemental forms; calcium citrate, 24%; tribasic calcium phosphate 39%; calcium lactate, 13%; and calcium gluconate, 9%. Many have cited data that calcium citrate is more bioavailable than calcium carbonate and may be more beneficial for bone. But the comparison studies were done in fasting individuals. Calcium citrate is better absorbed with food, due to the HCl production by the stomach. When taken with meals, there appears to be no difference between carbonate and citrate absorption. This is worth remembering in older women. As we age, HCl production tends to diminish, so calcium citrate may be more appropriate for older patients. It remains speculative whether citrate is better than carbonate for those on antacids and proton pump inhibitors.

In general, calcium intake alone may have only a slight protective effect for bone mass and fracture risk. The best outcomes have come from a combination of calcium and vitamin D. One study in people over age 65 showed the combination could significantly reduce nonvertebral fractures (Dawson-Hughes B, et al. N Engl J Med. 1997; 337(10): 670–676). Another study of postmenopausal women undergoing hip replacement surgery showed that those with hip fractures were more likely to have a vitamin D deficiency than those undergoing elective joint replacement (LeBoff M, et al. JAMA. 1999; 281(16): 1505–1511). More recently, the Nottingham Neck of Femur (NoNOF) study of hip fracture patients showed that vitamin D, either orally or by injection, suppresses parathyroid hormone, increases bone mineral density and reduces falls. The effects were more pronounced when the vitamin was co-supplemented with 1,000 mg of calcium per day (Harwood RH, et al. Age Ageing. 2004 Jan; 33(1): 45–51).

We have typically attributed the moderate protective benefit of vitamin D to its effect on bone mineral density. However, there is a good body of evidence that it may also directly improve muscle strength, and as a result, reduce fracture risk by preventing falls. Several randomized controlled trials have found that vitamin D reduced fractures within two to three months and have benefits in improving muscle strength. Trials evaluating the effect of vitamin D on fall rates have given mixed results. A meta-analysis in 2004 attempted to determine the overall efficacy of vitamin D in preventing falls in the elderly, especially women (Bischoff-Ferrari H, et al. JAMA. 2004; 291(16): 1999–2006). Based on 5 randomized clinical trials involving 1,237 individuals, the data show vitamin D reduced the risk of falling by 22% (corrected OR, 0.78; 95% CI, 0.64–0.92) compared with calcium alone or placebo.

The most recent study on this subject is from Australia, and involved 625 older residents of nursing homes and assisted-living facilities (95% women). Patients were randomized to receive either vitamin D (ergocalciferol; 10,000 IU weekly or 1,000 IU daily) or placebo. None of the participants had a vitamin D deficiency, and all also received 600 mg elemental calcium daily. During two years of follow-up, the incidence of falls was significantly lower in the vitamin D group compared to the placebo group (1.37 vs. 1.86 falls per person-year). Those who received vitamin D were also less likely to sustain a fracture (8% vs. 11%), although the difference was not statistically significant (Flicker L et al. J Am Geriatr Soc. 2005 Nov; 53: 1881–1888).

The daily requirements for vitamin D were last set in 1997 by the Food and Nutrition Board of the Institute of Medicine. The current dietary recommendations for vitamin D are:

At the time, the dietary reference intake (DRI) panel members argued that there was insufficient evidence for establishing estimated average requirements (EARs), better known as recommended dietary allowances (RDIs), for vitamin D. Instead, they defaulted to publishing “adequate intakes” (AI) as listed above.

EARs have been difficult to establish for vitamin D because it is metabolized like a hormone, and inputs from sunlight and food are hard to determine. In addition, the relationships between adequate vitamin D levels and health were not as clear in 1997 as they are now. The AI values for vitamin D were set on the basis of intakes necessary to achieve “normal” serum levels, which, at 25–137.5 nmol/L (10–55 ng/mL), represent a very broad range.

I believe it is high-time for a formal re-evaluation of the vitamin D requirements in light of all the new data published since 1997. It would seem important to start with a better determination of what levels of serum vitamin D are necessary to achieve optimal functional outcomes. Granted, this will be difficult, because serum 25-hydroxyvitamin D concentrations alone do not reflect optimal status in target tissues. What we need is to determine optimal concentrations of vitamin D metabolites in various tissues, especially bone, breast, and other tissues the health of which are directly affected by vitamin D.

Scientists may also need to determine the optimal serum 25-hydroxyvitamin D concentrations that cause serum parathyroid hormone (PTH) to reach a minimum. Other functional outcome measures for vitamin D would be related to calcium absorption and/or bone resorption. We now know that maximal suppression of serum PTH and calcium absorption efficiency increase when 25-hydroxyvitamin D concentrations reach 80–90 nmol/L.

Until the Institute of Medicine revises the vitamin D requirements, I would recommend that we test selected members of our patient population and try to achieve serum concentrations of 80 nmol/L. This is particularly important for elderly women, because they typically have a resistance to 1,25-dihydroxyvitamin D compared with young women, and may need concentrations well above the 50 nmol/L reference standard for post-menopausal women.

Other subgroups that warrant closer attention include: individuals who need to be evaluated for secondary causes of bone loss; dark-skinned women; those living in northern latitudes; women who remain indoors or who routinely protect their skin with clothing or sunscreen; and those with other risk factors for low bone density and fractures.

One approach would be to test these individuals, dose at higher levels of vitamin D if they are not at 80 nmol/L, and then recheck in 2 months, while also checking serum and urinary calcium levels to assure safety. Doses up to 2,000 IU per day will be safe in most individuals. Alternatively, one can try once-weekly dosing of 10,000 IU, or once monthly doses 50,000 IU for those with documented vitamin D deficiency. Be very careful when using doses over 2,000 IU per day.

Cod liver oil is a popular source of vitamin D, and many people take it regularly. I believe this should be done very cautiously. It is true that 1 tablespoon of cod liver oil supplies a healthy 1,400 IU dose of vitamin D, but it also contains very high levels of vitamin A, which can exacerbate calcium loss and promote low bone density. Cod liver oil is probably not the optimal source of vitamin D for most patients.

Recommended daily elemental calcium intakes for peri- and postmenopausal women: Keep in mind these figures are for total calcium. Don’t forget to estimate dietary sources. Once you estimate that number, then you supplement the difference to get to the total amount desired.