Re: Long term calcium intake and rates of all cause and cardiovascular mortality: community based prospective longitudinal cohort study
We thank Dr Grant for his interest in our study. He suspects that the results are not sustainable and we agree that observational studies should be interpreted with caution. The title wording “long term calcium intake” refers to the available repeatedly measured information on dietary habits enabling estimation dietary intake over a longer period, as is common in nutritional epidemiology. However, it is noteworthy that Dr. Grant presents an article by Heaney and colleagues (1) as an argument that a high calcium intake does not increase the risk of cardiovascular disease. Readers should be aware that this overview was initiated by the Council for Responsible Nutrition (CRN), the leading trade association representing the dietary supplement industry. This fact has been previously presented in BMJ by Reid and colleagues. (2) After publication of our study, the last author of the Heaney study (1), Dr Wallace (senior director, scientific and regulatory affairs of CRN), sent a press release from CRN (http://www.crnusa.org/CRNPR13-BMJCVD021213.html) criticizing our study on non-objective grounds, including claims that the results were confounded by estrogen therapy use, even though we clearly described in the article that this was not the case. In an e-mail correspondence with Dr. Wallace (February 15 2013), we have presented how our estimates change after adjustment for estrogen therapy, but despite this presentation to him, no retraction of his untrue statements has been made by CRN. The multivariable hazard ratios for total mortality of highest dietary calcium intake (>1400 mg/day) was, after addition of estrogen therapy use in the model, 1.41 (95% CI 1.18 to 1.68). The corresponding HR for IHD mortality was 2.15 (1.49 to 3.10). Both these values are marginally different from those presented without estrogen therapy adjustment.
Moreover, Dr. Grant also refers to a previous study from our group which demonstrated a U-shaped relationship between serum 25-hydroxyvitamin D levels and death from cancer and with cancer incidence.3 For cardiovascular death, there was an increased risk only at low levels of 25-OH-D. The link between high serum vitamin D and all-cause mortality was driven thus by an increased risk of cancer death. Dr Grant does not explain this detail even though he is well aware of the study's content. Grant now also claims that our research group is the only one who found the U-shaped relationship but this is not correct.4-9 There are, however, a number of studies that have not found similar correlations. Several explanations for these discrepancies may exist such as genetic constitution, differences in exposure range of serum 25-hydroxyvitamin D, differences in precision and accuracy of assays, outcome information bias, statistical approach and mere statistical power, as also discussed in the article.3 Dr. Grant's one-sided view that "more is better" when it comes to vitamin D and calcium should be supported by proper evidence. As long as there is a remaining uncertainty of benefits to supplement those who already have a good nutritional status, a precautionary principle is prudent to prevail. I would also remind Dr. Grant, that the number of hip fractures increase with supplementation with calcium10 - a fact not widely recognized. Results that do not suit one’s purposes should not be swept under the carpet but should be viewed with interest - no one benefits in the long run of a one-sided reporting of research results, not even Dr. Grant's Sunlight, Nutrition and Health Research Center (SUNARC).
Dr Gupta discusses the seemingly contradictory results of the cross-sectional study INTERSALT displaying that a low calcium intake in Japan was associated with increased risk of high blood pressure and our study results which showed a U-shaped relationship between calcium intake and death, particularly cardiovascular death. We do not see that there is a contradiction. In our study, we also noted an increased risk of death at a low calcium intake, although this risk was largely biased according to results from our marginal structural model.
Dr Gada correctly notes that the calcium in combination with vitamin D may have a different effect than calcium alone. This fact made us to consider vitamin D intake in an interaction analysis, as presented in the article. No such effect modification was found. Unfortunately, we had not the possibility to measure serum 25-hydroxyvitamin D levels in our cohort, considered to best reflect vitamin D status.
Dr Pena Pradel makes two observations. First, we fully agree that with an observational study you can definitely not be sure of a causal relation. However, the prospects for managing comorbidity are relatively good with our Swedish national patient registers. (11) We were able to take diseases both before baseline and until the second questionnaire into account, rather than excluding women with cardiovascular disease. Nevertheless, we cannot be certain that our estimates are correct and our study may be viewed as a puzzle piece among others. Second, we could not separate prescribed supplements from over-the-counter supplements in our data and it is therefore unclear to us what the actual indication the women in our cohort had with their supplement use. General screening for osteoporosis do not occur in Sweden. It is possible that a previous fracture caused individuals starting out with supplementation. When we took previous fractures into account in our multivariable model - as we described in the article - the estimates remained unchanged. Since Dr. Pena Pradel seems to be of the opinion that supplementation with calcium is of importance to prevent osteoporotic fractures, we want to make clear that the evidence for this claim is thin and that supplementation with calcium leads to an increased risk of hip fractures,10 the fracture in the elderly with most devastating consequences and also associated with a significantly increased mortality.12 The recent US Preventive Services Task Force (USPSTF) opinion on combined vitamin D and calcium supplementation to prevent fractures in adults is that there is at present insufficient evidence to recommend daily supplementation to non-institutionalized postmenopausal women. (13)
We agree with Dr Naneria that calcium is probably over-prescribed to individuals with an already good nutritional status. At the same time, we want to stress that if you suffer a hip fracture it may have serious implications. Premature mortality following hip fracture is well recognized and despite improved hospital care, there is no trend of decreasing mortality after hip fracture. (14)
Prof Katan argues that our relationship between calcium intake and death can be explained by a high intake of saturated fat. In the article, we reported only that our estimates did not change after adjustment for total fat intake. Anyhow, he calls for estimates after adjustment for saturated fat. In our Table 2 it can be noted that a calcium intake greater than 1400 mg per day was associated with a multivariable-adjusted HR of 1.40 (95% CI 1.17 to 1.67) for all-cause mortality and an HR of 2.14 (95% CI 1:48 to 3:09) for IHD mortality. After adjustment for saturated fat intake, these estimates were changed to 1.45 (95% CI 1.24 to 1.47) and 2.32 (95% CI 1.60 to 3.36), respectively. Thus, the associations do not appear to be explained by the saturated fat intake.
We agree that it may look odd that the estimate for the highest category of calcium intake change from an estimate below (1) in an age-adjusted model to a heightened risk after multivariable adjustment for supplement use. One fact we have to consider is that the supplement users in our setting, on average, are healthier than the general population. Again, this is due to the fact that we cannot separate prescription supplement use from over-the-counter supplement use. The latter users are healthier and more health conscious compared to the general population. (15) This bias has to be considered in our observational study. Ideally, however, the effect of supplementation itself is best investigated in randomized trials while large-scale dietary intervention studies are difficult to implement. We want to stress that even among those who did not report use of supplements we noted an increased mortality at a high calcium intake through diet (See figure).
Mr Farrar has misunderstood the conditions for epidemiological research in Sweden. Although some cohort members no longer lived within the catchment area, we can still track them until they die by using the individual's personal identity number provided to all Swedish citizens, combined with our national mortality registries. (11) Thus, the information lost pertains to updated information on nutrients and other questionnaire data, not follow-up on outcomes. Implausible energy intake was defined as an energy intake lower or higher than 3 SDs from the mean value for loge-transformed energy intake, which is a standardized criterion for also many other large cohort studies. This is described in the article’s references 17 and 18. The majority of women which were excluded based on this criterion in our cohort had reported low dietary intakes, not high.
References
1. Heaney RP, Kopecky S, Maki KC, Hathcock J, Mackay D, Wallace TC. A review of calcium supplements and cardiovascular disease risk. Advances in nutrition 2012;3(6):763-71.
2. Bolland MJ, Grey A, Reid IR. An inappropriate response? BMJ 2013;346:f942.
3. Michaelsson K, Baron JA, Snellman G, Gedeborg R, Byberg L, Sundstrom J, et al. Plasma vitamin D and mortality in older men: a community-based prospective cohort study. Am J Clin Nutr;92(4):841-8.
4. Dror Y, Giveon S, Hoshen M, Feldhamer I, Balicer R, Feldman B. Vitamin D Levels for Preventing Acute Coronary Syndrome and Mortality: Evidence of a Non-Linear Association. J Clin Endocrinol Metab 2013.
5. Durup D, Jorgensen HL, Christensen J, Schwarz P, Heegaard AM, Lind B. A reverse J-shaped association of all-cause mortality with serum 25-hydroxyvitamin D in general practice: the CopD study. J Clin Endocrinol Metab 2012;97(8):2644-52.
6. Melamed ML, Michos ED, Post W, Astor B. 25-hydroxyvitamin D levels and the risk of mortality in the general population. Arch Intern Med 2008;168(15):1629-37.
7. Meyer HE, Robsahm TE, Bjorge T, Brustad M, Blomhoff R. Vitamin D, season, and risk of prostate cancer: a nested case-control study within Norwegian health studies. Am J Clin Nutr 2013;97(1):147-54.
8. Albanes D, Mondul AM, Yu K, Parisi D, Horst RL, Virtamo J, et al. Serum 25-hydroxy vitamin D and prostate cancer risk in a large nested case-control study. Cancer Epidemiol Biomarkers Prev 2011;20(9):1850-60.
9. Tuohimaa P, Lou YR. Optimal serum calcidiol concentration for cancer prevention. Anticancer Res 2012;32(1):373-81.
10. Bischoff-Ferrari HA, Dawson-Hughes B, Baron JA, Burckhardt P, Li R, Spiegelman D, et al. Calcium intake and hip fracture risk in men and women: a meta-analysis of prospective cohort studies and randomized controlled trials. Am J Clin Nutr 2007;86(6):1780-90.
11. Calltorp J, Adami HO, Astrom H, Fryklund L, Rossner S, Trolle Y, et al. Country profile: Sweden. Lancet 1996;347(9001):587-94.
12. Haentjens P, Magaziner J, Colon-Emeric CS, Vanderschueren D, Milisen K, Velkeniers B, et al. Meta-analysis: excess mortality after hip fracture among older women and men. Ann Intern Med 2010;152(6):380-90.
13. Moyer VA, on behalf of the USPSTF. Vitamin D and Calcium Supplementation to Prevent Fractures in Adults: U.S. Preventive Services Task Force Recommendation Statement. Ann Intern Med 2013.
14. Brauer CA, Coca-Perraillon M, Cutler DM, Rosen AB. Incidence and mortality of hip fractures in the United States. JAMA 2009;302(14):1573-9.
15. Rautiainen S, Akesson A, Levitan EB, Morgenstern R, Mittleman MA, Wolk A. Multivitamin use and the risk of myocardial infarction: a population-based cohort of Swedish women. Am J Clin Nutr 2010;92(5):1251-6.
Competing interests:
No competing interests
19 April 2013
Karl Michaëlsson
Professor in medical epidemiology and senior consultant in orthopaedic surgery
Liisa Byberg
Department of surgical sciences, Uppsala university, Uppsala, Sweden
Department of surgical sciences, Uppsala university, SE-751 85 Uppsala, Sweden
Rapid Response:
Re: Long term calcium intake and rates of all cause and cardiovascular mortality: community based prospective longitudinal cohort study
We thank Dr Grant for his interest in our study. He suspects that the results are not sustainable and we agree that observational studies should be interpreted with caution. The title wording “long term calcium intake” refers to the available repeatedly measured information on dietary habits enabling estimation dietary intake over a longer period, as is common in nutritional epidemiology. However, it is noteworthy that Dr. Grant presents an article by Heaney and colleagues (1) as an argument that a high calcium intake does not increase the risk of cardiovascular disease. Readers should be aware that this overview was initiated by the Council for Responsible Nutrition (CRN), the leading trade association representing the dietary supplement industry. This fact has been previously presented in BMJ by Reid and colleagues. (2) After publication of our study, the last author of the Heaney study (1), Dr Wallace (senior director, scientific and regulatory affairs of CRN), sent a press release from CRN (http://www.crnusa.org/CRNPR13-BMJCVD021213.html) criticizing our study on non-objective grounds, including claims that the results were confounded by estrogen therapy use, even though we clearly described in the article that this was not the case. In an e-mail correspondence with Dr. Wallace (February 15 2013), we have presented how our estimates change after adjustment for estrogen therapy, but despite this presentation to him, no retraction of his untrue statements has been made by CRN. The multivariable hazard ratios for total mortality of highest dietary calcium intake (>1400 mg/day) was, after addition of estrogen therapy use in the model, 1.41 (95% CI 1.18 to 1.68). The corresponding HR for IHD mortality was 2.15 (1.49 to 3.10). Both these values are marginally different from those presented without estrogen therapy adjustment.
Moreover, Dr. Grant also refers to a previous study from our group which demonstrated a U-shaped relationship between serum 25-hydroxyvitamin D levels and death from cancer and with cancer incidence.3 For cardiovascular death, there was an increased risk only at low levels of 25-OH-D. The link between high serum vitamin D and all-cause mortality was driven thus by an increased risk of cancer death. Dr Grant does not explain this detail even though he is well aware of the study's content. Grant now also claims that our research group is the only one who found the U-shaped relationship but this is not correct.4-9 There are, however, a number of studies that have not found similar correlations. Several explanations for these discrepancies may exist such as genetic constitution, differences in exposure range of serum 25-hydroxyvitamin D, differences in precision and accuracy of assays, outcome information bias, statistical approach and mere statistical power, as also discussed in the article.3 Dr. Grant's one-sided view that "more is better" when it comes to vitamin D and calcium should be supported by proper evidence. As long as there is a remaining uncertainty of benefits to supplement those who already have a good nutritional status, a precautionary principle is prudent to prevail. I would also remind Dr. Grant, that the number of hip fractures increase with supplementation with calcium10 - a fact not widely recognized. Results that do not suit one’s purposes should not be swept under the carpet but should be viewed with interest - no one benefits in the long run of a one-sided reporting of research results, not even Dr. Grant's Sunlight, Nutrition and Health Research Center (SUNARC).
Dr Gupta discusses the seemingly contradictory results of the cross-sectional study INTERSALT displaying that a low calcium intake in Japan was associated with increased risk of high blood pressure and our study results which showed a U-shaped relationship between calcium intake and death, particularly cardiovascular death. We do not see that there is a contradiction. In our study, we also noted an increased risk of death at a low calcium intake, although this risk was largely biased according to results from our marginal structural model.
Dr Gada correctly notes that the calcium in combination with vitamin D may have a different effect than calcium alone. This fact made us to consider vitamin D intake in an interaction analysis, as presented in the article. No such effect modification was found. Unfortunately, we had not the possibility to measure serum 25-hydroxyvitamin D levels in our cohort, considered to best reflect vitamin D status.
Dr Pena Pradel makes two observations. First, we fully agree that with an observational study you can definitely not be sure of a causal relation. However, the prospects for managing comorbidity are relatively good with our Swedish national patient registers. (11) We were able to take diseases both before baseline and until the second questionnaire into account, rather than excluding women with cardiovascular disease. Nevertheless, we cannot be certain that our estimates are correct and our study may be viewed as a puzzle piece among others. Second, we could not separate prescribed supplements from over-the-counter supplements in our data and it is therefore unclear to us what the actual indication the women in our cohort had with their supplement use. General screening for osteoporosis do not occur in Sweden. It is possible that a previous fracture caused individuals starting out with supplementation. When we took previous fractures into account in our multivariable model - as we described in the article - the estimates remained unchanged. Since Dr. Pena Pradel seems to be of the opinion that supplementation with calcium is of importance to prevent osteoporotic fractures, we want to make clear that the evidence for this claim is thin and that supplementation with calcium leads to an increased risk of hip fractures,10 the fracture in the elderly with most devastating consequences and also associated with a significantly increased mortality.12 The recent US Preventive Services Task Force (USPSTF) opinion on combined vitamin D and calcium supplementation to prevent fractures in adults is that there is at present insufficient evidence to recommend daily supplementation to non-institutionalized postmenopausal women. (13)
We agree with Dr Naneria that calcium is probably over-prescribed to individuals with an already good nutritional status. At the same time, we want to stress that if you suffer a hip fracture it may have serious implications. Premature mortality following hip fracture is well recognized and despite improved hospital care, there is no trend of decreasing mortality after hip fracture. (14)
Prof Katan argues that our relationship between calcium intake and death can be explained by a high intake of saturated fat. In the article, we reported only that our estimates did not change after adjustment for total fat intake. Anyhow, he calls for estimates after adjustment for saturated fat. In our Table 2 it can be noted that a calcium intake greater than 1400 mg per day was associated with a multivariable-adjusted HR of 1.40 (95% CI 1.17 to 1.67) for all-cause mortality and an HR of 2.14 (95% CI 1:48 to 3:09) for IHD mortality. After adjustment for saturated fat intake, these estimates were changed to 1.45 (95% CI 1.24 to 1.47) and 2.32 (95% CI 1.60 to 3.36), respectively. Thus, the associations do not appear to be explained by the saturated fat intake.
We agree that it may look odd that the estimate for the highest category of calcium intake change from an estimate below (1) in an age-adjusted model to a heightened risk after multivariable adjustment for supplement use. One fact we have to consider is that the supplement users in our setting, on average, are healthier than the general population. Again, this is due to the fact that we cannot separate prescription supplement use from over-the-counter supplement use. The latter users are healthier and more health conscious compared to the general population. (15) This bias has to be considered in our observational study. Ideally, however, the effect of supplementation itself is best investigated in randomized trials while large-scale dietary intervention studies are difficult to implement. We want to stress that even among those who did not report use of supplements we noted an increased mortality at a high calcium intake through diet (See figure).
Mr Farrar has misunderstood the conditions for epidemiological research in Sweden. Although some cohort members no longer lived within the catchment area, we can still track them until they die by using the individual's personal identity number provided to all Swedish citizens, combined with our national mortality registries. (11) Thus, the information lost pertains to updated information on nutrients and other questionnaire data, not follow-up on outcomes. Implausible energy intake was defined as an energy intake lower or higher than 3 SDs from the mean value for loge-transformed energy intake, which is a standardized criterion for also many other large cohort studies. This is described in the article’s references 17 and 18. The majority of women which were excluded based on this criterion in our cohort had reported low dietary intakes, not high.
References
1. Heaney RP, Kopecky S, Maki KC, Hathcock J, Mackay D, Wallace TC. A review of calcium supplements and cardiovascular disease risk. Advances in nutrition 2012;3(6):763-71.
2. Bolland MJ, Grey A, Reid IR. An inappropriate response? BMJ 2013;346:f942.
3. Michaelsson K, Baron JA, Snellman G, Gedeborg R, Byberg L, Sundstrom J, et al. Plasma vitamin D and mortality in older men: a community-based prospective cohort study. Am J Clin Nutr;92(4):841-8.
4. Dror Y, Giveon S, Hoshen M, Feldhamer I, Balicer R, Feldman B. Vitamin D Levels for Preventing Acute Coronary Syndrome and Mortality: Evidence of a Non-Linear Association. J Clin Endocrinol Metab 2013.
5. Durup D, Jorgensen HL, Christensen J, Schwarz P, Heegaard AM, Lind B. A reverse J-shaped association of all-cause mortality with serum 25-hydroxyvitamin D in general practice: the CopD study. J Clin Endocrinol Metab 2012;97(8):2644-52.
6. Melamed ML, Michos ED, Post W, Astor B. 25-hydroxyvitamin D levels and the risk of mortality in the general population. Arch Intern Med 2008;168(15):1629-37.
7. Meyer HE, Robsahm TE, Bjorge T, Brustad M, Blomhoff R. Vitamin D, season, and risk of prostate cancer: a nested case-control study within Norwegian health studies. Am J Clin Nutr 2013;97(1):147-54.
8. Albanes D, Mondul AM, Yu K, Parisi D, Horst RL, Virtamo J, et al. Serum 25-hydroxy vitamin D and prostate cancer risk in a large nested case-control study. Cancer Epidemiol Biomarkers Prev 2011;20(9):1850-60.
9. Tuohimaa P, Lou YR. Optimal serum calcidiol concentration for cancer prevention. Anticancer Res 2012;32(1):373-81.
10. Bischoff-Ferrari HA, Dawson-Hughes B, Baron JA, Burckhardt P, Li R, Spiegelman D, et al. Calcium intake and hip fracture risk in men and women: a meta-analysis of prospective cohort studies and randomized controlled trials. Am J Clin Nutr 2007;86(6):1780-90.
11. Calltorp J, Adami HO, Astrom H, Fryklund L, Rossner S, Trolle Y, et al. Country profile: Sweden. Lancet 1996;347(9001):587-94.
12. Haentjens P, Magaziner J, Colon-Emeric CS, Vanderschueren D, Milisen K, Velkeniers B, et al. Meta-analysis: excess mortality after hip fracture among older women and men. Ann Intern Med 2010;152(6):380-90.
13. Moyer VA, on behalf of the USPSTF. Vitamin D and Calcium Supplementation to Prevent Fractures in Adults: U.S. Preventive Services Task Force Recommendation Statement. Ann Intern Med 2013.
14. Brauer CA, Coca-Perraillon M, Cutler DM, Rosen AB. Incidence and mortality of hip fractures in the United States. JAMA 2009;302(14):1573-9.
15. Rautiainen S, Akesson A, Levitan EB, Morgenstern R, Mittleman MA, Wolk A. Multivitamin use and the risk of myocardial infarction: a population-based cohort of Swedish women. Am J Clin Nutr 2010;92(5):1251-6.
Competing interests: No competing interests