NPC Archive Item: Calcium supplements without vitamin D may increase the risk of MI

NOTE – This is an archive post from the NPC and has not been updated since first publication. Therefore, some hyperlinks may no longer be working.
MeReC Rapid Review NPC Logo

23rd August 2010

This meta-analysis found that calcium supplements without vitamin D increased the risk of myocardial infarction (MI) versus control by about 30% in relative terms, but did not find a statistically significant effect on risk of stroke or death. The analysis did not include studies of calcium plus vitamin D versus control; in the light of UK practice which is to usually prescribe calcium with vitamin D, this limits the relevance of these new data.

Level of evidence:
Level 1 (good quality patient-oriented evidence) according to the SORT criteria.

Action
We anticipate that the regulatory authorities will be investigating this safety signal. Until any advice is published, health professionals may wish to consider their prescribing of calcium alone taking into account:

  • the safety signal from this study regarding calcium alone possibly increasing cardiovascular (CV) risk
  • the lack of a safety signal from this study about an increased CV risk from calcium plus vitamin D combinations, and the limited evidence that suggests that addition of vitamin D may negate or at least mitigate any harmful effects of calcium supplementation alone
  • the evidence that the combination of calcium 1000mg/day or greater plus vitamin D 800 units/day or greater has been shown to be more effective than calcium alone in reducing the risk of falls and fracture.

In accordance with NICE guidance on primary and secondary prevention of osteoporosis, unless prescribers are confident that women who receive treatment (e.g. with bisphosphonates) have an adequate calcium intake and are vitamin D replete, they should consider calcium and/or vitamin D supplementation.

What is the background to this?
An earlier secondary analysis of a randomised controlled trial (RCT) of calcium monotherapy for primary prevention of fracture in postmenopausal women had suggested an increased risk of MI. However, this study had certain limitations (see MeReC Rapid Review 57). This meta-analysis (which has many of the same authors) examined this question further. It included only RCTs of at least calcium 500mg/day versus placebo for more than one year and excluded studies of calcium plus vitamin D versus placebo. Trials were eligible if vitamin D was given to both intervention and control groups, and one such study was included. Patient-level data were available from five studies (n=8,151) and summary-level data were available from these plus 6 further studies (total n=11,921).

What does this study claim?
In the patient-level analysis, the hazard ratio (HR) for incident MI in those allocated to calcium compared to placebo was 1.31 (95% confidence interval [CI] 1.02 to 1.67, P=0.035). The risks of stroke or death (separately) or the composite of MI, stroke or sudden death were not statistically significantly different from placebo. The authors calculated that the number needed to harm (NNH) with calcium for five years to cause one incident MI was 69 (although in practice this would depend on the person’s baseline risk of MI). Adjusting for prespecified covariates related to cardiovascular (CV) outcomes (age, sex, smoking status, and diabetes) did not change the results of the primary analyses.

In the summary level analysis of data from eight RCTs, allocation to calcium supplements was associated with a statistically significant increased risk of MI (relative risk [RR] 1.27, 95%CI 1.01 to 1.59, P=0.038) but not stroke, death or the composite end point. Adding data from the remaining three trials (where data were available only for subgroups of participants) did not significantly change the results for any end point.

How does this fit with the evidence on the effects of calcium, and vitamin D, on fracture rates?
The evidence for calcium and vitamin D supplementation, separately and together, to reduce the risk of osteoporotic fracture and falls is not straightforward.

To examine this in more detail, a large MA found that calcium alone and also when used in combination with vitamin D produced a 12% relative reduction in risk of all fractures overall (RR 0.88, 95%CI 0.83 to 0.95, P=0.004). Subgroup analysis found that calcium supplementation had a modest effect on fracture risk versus placebo, at the limits of conventional levels of statistical significance (RR 0.90, 95%CI 0.80 to 1.00, P not stated). The risk reduction in studies of calcium and vitamin D was slightly greater and showed a greater degree of statistical significance: RR 0.87 (95%CI 0.77 to 0.97, P not stated). However, the RR from calcium plus vitamin D was not significantly different from the RR from calcium alone (P=0.63); this finding could be due to limitations in the data. Calcium with daily doses of vitamin D of 800 units/day or greater had a significant effect on fracture rates (RR 0.84, 95%CI 0.74 to 0.94) but doses of vitamin D less than this did not (RR 0.87, 95%CI 0.71 to 1.05, P for difference in RRs=0.03). Most of the patients in trials in the MA who received calcium and vitamin D received lower doses of vitamin D, and this may have diminished the observed difference between the calcium and calcium plus vitamin D groups. The dose of calcium was also important: overall, doses of 1200mg/day or greater had a significantly greater effect on fractures than lower doses (P for difference 0.006), although around two thirds of such patients were also taking vitamin D 800 units/day.

In addition, the treatment effect from calcium and calcium plus vitamin D supplements was found to be significantly greater than the overall figure in some subgroups. These included people living in institutions compared to those living in the community, those whose daily calcium intake was less than 700mg per day and in trials with a compliance rate 80% or greater. Treatment was not significantly effective in people aged 50 to 69 years but was effective in older people, and the effect increase with age.

As discussed in MeReC Rapid Review 1052, a more recent MA, looking specifically at vitamin D, also found that daily calcium and vitamin D supplementation significantly reduced the overall risk of fracture across a wide age range of people, irrespective of sex and fracture history. In contrast, vitamin D alone in doses equivalent to 400 to 800 units/day did not prevent fractures. Finally, another MA found that found that high-dose vitamin D supplements (700 to 1000 units/day) reduced the relative risk of falling by 19% in individuals aged 65 years or more, but lower daily doses did not.

Overall, the evidence suggests that the greatest benefits are likely to come from a dose of calcium 1000 to 1200mg/day plus vitamin D 800 units/day. NICE guidance on primary and secondary prevention of osteoporosis specifically states that unless prescribers are confident that women who receive treatment (e.g. with bisphosphonates) have an adequate calcium intake and are vitamin D replete, then calcium and/or vitamin D supplementation should be considered.

So what?
This study raises questions over the use of calcium supplements alone to reduce the risk of osteoporotic fractures. However, it has some limitations. Most importantly, it did not consider the risk of calcium plus vitamin D versus placebo. It did include data  from one trial on CV outcomes in participants randomised to calcium 1000mg/day plus vitamin D 800 units/day compared to those randomised to vitamin D 800 units/day plus placebo. There was an observed increase risk of MI in the calcium plus vitamin D group compared to the vitamin D plus placebo group in that study, but it was not statistically significant.  However, a meta-analysis of vitamin D supplementation versus control in 18 RCTs found a reduction in risk of all-cause mortality (RR 0.93, 95%CI 0.87 to 0.99). Moreover, the summary RRs were similar in the trials which did and did not include calcium supplements in the intervention group. It is therefore possible that addition of vitamin D negates or at least mitigates any harmful effects of calcium supplementation alone.

In addition, CV outcomes were not pre-specified primary outcomes in any of the studies included in the MA, in none of them as single studies was an increased risk of MI (or other outcomes examined) statistically significant, and confidence intervals were wide in both individual studies and summary results.

Study details
Bolland M et al. Effect of calcium supplements on risk of myocardial infarction and cardiovascular events: meta-analysis. BMJ 2010;341:c3691

Design: Meta-analysis of RCTs.

Patients and trials: Trials were included only if elemental calcium was administered at a dose of 500mg/day or greater, the participants’ mean age at baseline was more than 40 years; 100 or more participants were randomised; participants of either sex were studied; and the trial duration was more than one year.Trials were excluded if calcium and vitamin D were given together with a placebo comparator (trials were eligible only if vitamin D was given to both intervention and control groups), if calcium was administered in the form of dietary modification or a complex nutritional supplement; and if most participants had a major systemic disease other than osteoporosis. Patient-level data were available from five studies (n=8,151) and summary-level data were available from these plus 6 further studies (total n=11,921)

Intervention and comparison Elemental calcium at a dose of 500mg/day or greater or placebo, for average of 4 years

Outcomes and results In the patient-level analysis, the HR for incident MI in those allocated to calcium was 1.31 (95%CI 1.02 to 1.67, P=0.035). The HR for stroke was 1.20 (95%CI 0.96 to 1.50, P=0.11), the HR for death was 1.09 (95%CI 0.96 to 1.23, P=0.18) and the HR for a composite of MI, stroke or sudden death was 1.18 (95%CI 1.00 to 1.39, P=0.057)

In the summary level analysis of data from eight RCTs, allocation to calcium supplements was associated with a statistically significant increased risk of MI (RR 1.27, 95%CI 1.01 to 1.59, P=0.038) but not stroke (RR 1.12, 95%CI 0.92 to 1.36, P=0.25), death (RR 1.07, 95%CI 0.95 to 1.19, P=0.26) or the composite end point (RR 1.12, 95%CI 0.97 to 1.30, P=0.13). Adding data from the remaining three trials (where data were available only for subgroups of participants) did not significantly change the results for any end point

Sponsorship: Health Research Council of New Zealand and the University of Auckland School of Medicine Foundation

More information on osteoporosis can be found on the relevant section of NPC

Please comment on this blog in the NPC discussion rooms, or using our feedback form.

Make sure you are signed up to NPC Email updates — the free email alerting system that keeps you up to date with the NPC news and outputs relevant to you