Vitamin D: Cardiovascular Disease - Major New Frontiers for Prevention

I'm David Sain from Wake Forest University, and I'm a cardiologist, so I'm going to speak about cardiovascular diseases as a potential major new frontier for prevention, using vitamin D, of course. I want to thank Carol for the invitation to speak here and for organizing this incredible event. Both the speakers and the audience are very distinguished, knowledgeable, and enthusiastic on this topic, so I feel very honored to be here. So I'm going to talk about the heart, vascular biology to some extent, effects on blood pressure, and some effects on biomarkers that relate to prognostication of cardiovascular events. I'm not going to speak about the kidneys, except maybe in the context of my disclosure, and I'm the only member who has something to disclose, and there it is. I have spoken at symposia that were sponsored by Abbott, who makes pericalcitol, and as it turns out, there is some industrial interest in vitamin D analogs for patients with chronic kidney disease, which is not an uncommon, or there's a considerable prevalence of chronic kidney disease in the United States. In fact, in the most recent U.S. RDS analysis, 13 million subjects in the United States have chronic kidney disease, and chronic kidney disease in and of itself is a major risk factor for cardiovascular events, so there is an interest in that, and thereby I have this disclosure for you. I have two big objectives. First, to talk about mechanisms by which vitamin D status might contribute to vascular disease and adverse cardiovascular outcomes, and then secondly, to review a whole host of studies, cross-sectional and some prospective studies, linking low vitamin D status with adverse cardiovascular disease outcomes and subclinical cardiovascular disease. So first, I'd just like to go back a number of years, say 10 years ago, and look at what sort of epidemiology might have propelled one to study vitamin D in the context of cardiovascular events, cardiovascular disease, and so these slides are similar to the ones shown by Drs. Garland, except that there's no smiley face, and that's because this is plotted as absolute distance from the equator, either north or south, and here we're looking at the associations between latitude, distance from the equator, and mean blood pressure, either diastolic blood pressure, this is systolic blood pressure, this is diastolic blood pressure, absolute blood pressure, up here, and then down here, looking at the prevalence of hypertension, this is from the InterSalt study that was published by Stephen Rosten back in 1997, you can see that the farther you are from the equator, the higher your blood pressure is, and the more likely you are to have hypertension. This is France, and there was an interesting study published by Boullée back in 1999, looking at circannual heart failure rates, or yearly heart failure rates, in France, and you can see this interesting periodicity where mortality from heart failure, in-hospital deaths, and hospitalizations all show a peak in the winter months, so as Dr. Frank Garland just mentioned, this is one of the clues that there might be a vitamin D deficiency going on. Going south of the equator, this is New Zealand, which has two big islands, the North Island and the South Island, and looking at CHD rates, we also see a periodicity, although in this case, the phase has shifted, and the peak occurrence, the top graph here is in males, CHD, so these would be coronary events, MI, sudden cardiac death, you can see that there is a periodicity for males and females, males of course have higher rates than women, but the peak occurrence, instead of being in our winter months, is in June, July, and August, again, when the vitamin D, when the UV radiation would be at a nadir. And furthermore, if we compare death rates, just between the two islands, the North Island and the South Island, which differ only by a couple of degrees of latitude, we find that the event rates per million are higher in the South Island, which is farther, again, from the equator, than they are in the North Island. Again, interesting epidemiology, suggesting that vitamin D is playing some role here. Now, if you went back ten years ago, that's about all you would have. In fact, if you went back five years, you wouldn't have too much more than that, but really, in the past two to three years, there's been just an explosion of data in the cardiovascular frontier on vitamin D. And, you know, as cardiologists, we tend to be fairly skeptical of simple, non-invasive treatments. Okay. I shouldn't have said that. Take that back. But we, you know, if we can put a stent in, or a pacemaker, or an ICD, or, you know, but to treat with vitamin D, which can be incredibly cheap, and prevent diseases, that's an interesting concept for cardiologists, but we're catching on. And there's some nice studies, and also nice reviews. Dr. Hollick co-authored a review that just came out in one of our major journals, Journal of American College of Cardiology. I saw it just this morning, and read it at 4 a.m. this morning. I'm still in Eastern Standard Time. So, the first, so how do you explain the potential for vitamin D to have an effect on the cardiovascular system? Well, you know, the receptor has to be there. And it is. And you can show that it's present in the heart and blood vessels. This was a study done by Robert Simpson's lab at University of Michigan, where he did, they did immunofluorescent confocal microscopy of adult rat cardiomyocytes. And they found the vitamin D receptor in these isolated cardiomyocytes. And they also found evidence that it tended to localize with a dihydropyridine receptor, which is a calcium channel that could mediate the flux of calcium ions. And they found a lesser tight association with a channel called the circa. And so they concluded that the vitamin D receptor is present along the T-tubules in the cardiomyocytes. And for those of you who don't study cardiac anatomy every day, this is, these T-tubules are important in the electrical, mechanical association of the heart. So, the coupling the electrical activity to the contractile activity. The vitamin D receptor is also present in the endothelium, and the smooth muscle cells here in a rat aorta. It was demonstrated by Paul Van Hooty's group in Singapore with Wong as the first author. So it's present in the tissues where it could be active. And has been alluded to earlier, it's present in a lot of tissues. And we're of course interested in cardiomyocytes, smooth muscle cells, and endothelial cells. And I should also point out that at least for smooth muscle cells and endothelial cells, there's evidence for a 1-alpha hydroxylase in these cells as well. Meaning that there can be conversion of 25-hydroxy to 1-alpha, 125-dihydroxy locally for a paracrine effect in the vessel wall as well. But it's not all just the cardiac and vascular tissues that we're interested in. Of course the heart and the blood vessels sit within the entire body and are influenced by other cells and other tissues. And as Dr. Garland just mentioned, pancreatic beta cells of course have the vitamin D receptor and could influence insulin secretion and glycemic control. And as he also mentioned, it's present in immune cells. And of course now as Russell Ross taught us years ago and Peter Libby and others have further taught us that atherosclerosis is largely an inflammatory disease and we find the presence of the VDR in inflammatory or immune cells. But even beyond that, when we test our patients and we evaluate exercise performance, we're also not just testing the cardiac output per se or the heart per se, we're also testing muscle strength. One of the tests we do for heart failure for example is a six-minute walk test. Well that's a test not only of cardiac output and pulmonary capillary wedge pressure and such, but it's also a test of muscle strength. So the presence of the VDR within striated muscles may impact upon our evaluation of patients with cardiac diseases. Now to further break this down, there's evidence for both genomic and rapid responses to vitamin D. This is actually from a review of Tony Norman's published a couple years ago in Endocrinology adapted from his review and their characteristics of the genomic response and the rapid response mainly differentiated by the time element wherein the rapid response occurs over seconds to minutes and the genomic response occurs over hours to days. The main mediator in both seems to be the vitamin D receptor, although for the rapid response there has been postulated a second receptor. This rapid response occurs by vitamin D interacting at the membrane level, presumably by the vitamin D receptor localized within cavioli in the membrane. In terms of the genomic response, it's a transcription response, in the rapid response it's an activation of various enzymes and second messengers. And Dr. Norman has proposed a conformational ensemble model wherein the calcitriol being a very flexible molecule can have different conformations that interact differentially with the receptors in the nucleus or cytoplasmic nucleus versus in the cell membrane. Now cardiovascular examples of these two responses are for the genomic response the suppression of renin, which I'll discuss a little bit more later. For the rapid response you can take cardiomyocytes, bathe them with 125-dihydroxyvitamin D and show an immediate effect within virtually within a few minutes. And Robert Simpson's group has shown that treating cardiomyocytes, you know, in a petri dish within minutes reduces the time to relaxation, so it directly affects the contractile and the relaxation properties of the cardiomyocytes. It also immediately affects the vascular response, reducing endothelial cell-dependent contractions in the vessel wall. But in terms of long-term health, certainly we believe at least right now that this genomic response is the more important element that we should consider. And so here we have the vitamin D receptor interacting with the retinoic X receptor binding to a vitamin D response element and then assembling this transcriptional pre-initiation complex that either turns genes on or turns genes off. In the case of renin, it shuts down the production of renin. And this is very important as renin has broad effects both on the vessel wall in terms of inflammation, oxidative stress, and the progression of atherosclerosis in the vessel wall. Of course renin is a major regulator of the blood pressure and renin, even independent of blood pressure, can affect the development of left ventricular hypertrophy. So this is certainly a major effect of vitamin D receptor activation. But it's not just renin. Actually it's been now described using microarray analysis that more than 200 genes in heart and blood vessels can be regulated, either turned up or turned on or shut down by the presence of activated vitamin D. So these may have many effects on all of these systems and more that are yet to be explored. Now in terms of renin, we should point out that renin itself has been identified as a prognosticator for future occurrence of myocardial infarction. Back in 1997, Alderman showed that renin level here when coupled with traditional risk factors added to the prognostication ability so that patients who had high renin levels and high conventional risk factors had exceptionally high rates of future occurrence of myocardial infarction. And more recently in 2005 from the PROGRESS study, this was a study of an ACE inhibitor we see that higher quartiles of renin and in this case when coupled with another marker N-terminal probe B and P were associated with substantially higher rates of future development of myocardial infarction. And we believe that renin is not just a marker but actually a mediator of these events and the ability of vitamin D to suppress renin levels certainly is likely a very significant beneficial attribute of vitamin D. Now let's turn to heart. There's a very good or actually now two good animal models of vitamin D deficiency on the VDR level. If vitamin D receptor is knocked out, shown here by this negative negative, you see high levels of renin in the juxtaglomerular apparatus of the kidney. The cardiomyocytes are large. In the cardiac tissue you see fibrosis. And so the hearts are structurally altered and structurally different presumably because of renin. You can demonstrate high renin levels in these mice, higher angiotensin 2 levels, higher blood pressure. And actually when you take the hearts and weigh them, the heart to body weight ratio in these knockout mice is substantially higher. So these mice have LVH. And this model has been also shown for a knockout of the 1-alpha hydroxylase. So the terminal step of activation of vitamin D into the fully active hormone. Turning to the vessel wall, there are many effects that have been described for activated vitamin D on the vessel wall. This is a pictorial representation of a slice of a diseased coronary artery. And here we have an atheroma. This dark blue represents calcification. We have inflammatory cells, macrophages, and T-cells. Here's a lipid core, a fibrous cap overlying it. This fibrous cap has been disrupted, allowing a platelet fibrin thrombus to partially occlude the lumen as well as extend down within the lipid core. So vitamin D has been, activated vitamin D has been shown to affect virtually all of these processes. For example, calcification, you can show an inverse correlation between vitamin D levels and arterial calcification. At first blush, one might think that there would be a positive correlation since vitamin D affects calcium and phosphate uptake. But in fact, it's the opposite. And what it turns out, activated vitamin D suppresses some of the initiators of calcification like bone morphogenetic protein 2, while it upregulates inhibitors of calcification like matrix GLA protein. So vitamin D appears to be inversely associated with arterial calcification, not only arterial calcification, but valvular calcification. Aortic sclerosis, mitral anterior calcification are also inversely correlated with vitamin D status. Vascular stiffness has been shown to be inversely correlated with vitamin D status, both in patients with end-stage renal disease by Girard-London, but also in healthy subjects given a large dose of vitamin D, you can show responses to the forearm endothelial response that correlate with vitamin D levels. Inflammation has already been mentioned, and there are data showing higher levels of CRP, IL-6 in patients who are vitamin D deficient, and vitamin D affects the distribution between T helper 1 and T helper 2 type cells. MMP2 and MMP9, which are matrix metalloproteinases, which are thought to be involved in plaque rupture and cause the acute expression of atherosclerosis and cause the rupture of this fibrous cap, MMP2 and 9 levels are higher in subjects who are vitamin D deficient. And then in terms of thrombosis, it's been shown that activated vitamin D suppresses pi 1, suppresses tissue factor, which are promoters of thrombosis, while upregulating thrombomodulin, an inhibitor of thrombosis. So there are multiple beneficial attributes of vitamin D in the vessel wall. So now I'd like to turn to studies that link low vitamin D status with subclinical cardiovascular disease and adverse outcomes, and we have a whole host of them now to cover. A lot has changed in the past few years. First of all, in terms of going back to the renin and its association with hypertension, Foreman and colleagues at Harvard published this analysis of two studies, really. For men, it was the health protection follow-up study and women and nurses' health study, and they looked at these categories of vitamin D and looked at the incident hypertension over four years of follow-up. So these were subjects who did not have hypertension at baseline, but they looked at whether they were developed hypertension over four years of follow-up. And you can see for men, men who had 25 hydroxy D levels less than 15 nanograms per mil had a six-fold higher risk of new-onset hypertension compared to those with greater than 30 nanograms per mil. And for women, the effect was not quite as strong, but still statistically significant with a 2.7-fold increase for those who were less than 15 nanograms per mil. Low vitamin D status has been associated with the characteristics of the metabolic syndrome, as shown in the cross-sectional NHANES-3 analysis. And here we're comparing the first quartile to the fourth quartile. The first quartile was less than 21, and the fourth quartile greater than 37 nanograms per mil. And if you looked at blood pressure, again, it's higher in the first than the fourth. Fasting blood glucose greater than 126, defining diabetes. BMI greater than or equal to 30. Triglyceride level greater than or equal to 150. Albumin less than 3.5. Albumin, you know, is a negative acute phase reactant, so this likely represents an inflammatory stimulus. So patients who had low vitamin D also tended to have low albumin. And albumin creatinine ratio greater than 200 for men and greater than 300 for women. All features of the metabolic syndrome all correlated with low vitamin D status. Low vitamin D in the NHANES data from 2001-2004 was associated with prevalence of peripheral arterial disease. Melamed et al. published this recently in ATVB, and they defined PAD by an ankle brachial index less than .9. And here, again, looking at quartiles of 25-hydroxy D levels, you can see that as the vitamin D levels were lower, the prevalence ratio of PAD by the ABI increased to about 2.2. So again, this is evidence for an association between atherosclerosis and low vitamin D status. Another association between 25-hydroxy D was shown in type 2 diabetics. Now, this is a study from Italy, 390 type 2 diabetics. You wouldn't think there would be much vitamin D deficiency in Italy, a Mediterranean country, but there actually was. And they defined hypovitaminosis D as a level less than 37.5 nanomoles per liter. Divide that by 2.5, and that gives you 15 nanograms per mil. So diabetics had a prevalence of 34, non-diabetics 16.4. So there's quite a bit of vitamin D deficiency even in Italy, and it's more common in diabetics as has been shown by a lot of studies. Then they did carotid IMTs, carotid intramedial thickness estimates on the diabetic subjects. And those with low vitamin D had higher carotid IMTs than those who had sufficient vitamin D. So the carotid intima was thicker as a function of vitamin D status. And in fact, in their multivariate regression analysis, low 25-hydroxy D independently predicted carotid IMT. Now the difference between the sufficient and deficient state in terms of the carotid IMT was .23 millimeters. And that doesn't sound very impressive, .2 millimeters. But actually if you look at meta-analyses of carotid IMT studies, it's been well shown that there's a 10-15% increase in both cardiovascular and cerebrovascular events for every .1 millimeter of carotid IMT. So actually .23 millimeter difference is actually very substantial in terms of predicting future cardiovascular events. Now I'd like to turn to another study, the Framingham cohort study that looked at low vitamin D status and incident cardiovascular disease. These were offspring of the original Framingham cohort. These were subjects without prevalent cardiovascular disease. They were followed for five years. And the outcome was the first cardiovascular event defined as a coronary event, a stroke, claudication or heart failure. And just of note, about a third of this all-white cohort from the Framingham area had levels less than 15 nanograms per mil. And this is the data looking at a relative risk for that event, that first occurrence of cardiovascular event by category of 25 hydroxy D levels less than 10, 10-15 or greater than 15 here shown in the 10 as a reference value. And you can see there's a stepwise increase in events with lower vitamin D status. And even after adjusting for age, sex, all cardiovascular risk factors, all cardiovascular risk factors and CRP, there was still an association, a statistically significant association between cardiovascular events and vitamin D status. This is another study by Ed Giovannucci from the Health Professionals Follow-Up Study. This is looking at MI. And this study was a large database, more than 18,000 men followed for about 10 years. And they looked at the occurrence of MI depending on categories of vitamin D status, greater than 30 being the referent value. And then they had sort of two intermediate values and then less than 15. And you can see that with lower vitamin D levels, there's an increased occurrence of MI. And overall, after adjusting for all the covariates, the relative risk remained at about 2.2. Sort of a bottom line, the way I think about this study is that for those men who had levels greater than 30 nanograms per mL, they had approximately one-half the risk of having a future MI, one-half. And yet, men who had such a level were only about, as I recall, 23% of the total population. So only about one-fourth of the men were in this category of having greater than 30 nanograms per mL of vitamin D, and those men had one-half the risk of future MI. Very impressive data. Now, the LURIC studies are a group of studies that have recently come out from Germany. It's a series of studies. This is the LURIC stands for Ludwigshaben Risk in Cardiovascular Health Study. First of all, I should point out that this was done in southwest Germany where the latitude is about 49 degrees. Here where San Diego is 32 degrees. Buffalo was 45 degrees. So this is even more northerly than Buffalo. This is the city that's comparable would be Vancouver, British Columbia. This is quite high up there. This was a prospective cohort of 3,300 men and women referred for cardiac catheterization. So these were not community dwellers. There was some reason for these patients to be referred for cardiac catheterization. They had some symptoms. They had risk factors and so forth. So this is not like a cross-sectional study of subjects without a risk or without potential disease. One of the strengths of this study was that both 25-hydroxy-D and 125-dihydroxy-D were measured at the baseline, and they were followed for 7.7 years. And first, just looking at all-cause and cardiovascular mortality by quartiles of 25-hydroxy-D, you can see there's a very potent effect on both all-cause mortality as well as cardiovascular-specific mortality. So if you go to lower vitamin D levels, the death rate is substantially higher, both all-cause mortality and cardiovascular mortality. It's interesting here to compare the quartiles at this high latitude. The lowest quartile is 7.6. In the Framingham study, the lowest quartile was 17.8, which shows that something you have to be careful comparing studies where only quartiles are reported. So these subjects had substantially lower vitamin D levels because of the significantly higher latitude at which this study was conducted. Now one of the, as I said, one of the interesting features of this study was that they had measured both 25-hydroxy-D as well as 125-dihydroxy-D, and they were able to show an independent predictive value of both of these analytes. In fact, that's the title of their paper. Independent association of 125 and 125 on all-cause and cardiovascular mortality. So if you go along to the lowest 25-hydroxy-D quartile here, you see a substantial increase in mortality rates, but then if you add to that the lowest 125-dihydroxy levels, your mortality rate is substantially increased. And this raises a really interesting question. If 25-hydroxy is the substrate for 125-dihydroxy, why would there be this independent association? And there are certainly people here who can better answer this question than I can, but it has been speculated that either the 25-hydroxy is locally converted in a paracon fashion at the tissue level by 1-alpha hydroxylase in the vessel wall, for example, and that you get local production of the dihydroxy, and that dihydroxy is not reflected in the serum measures, but that's one possibility. Another possibility is that the 25-hydroxy does in fact activate the vitamin D receptor, although that's certainly not been proved that I'm aware of. Yes? Is there any metastellular serum? These were all serum. I'm not aware. Someone out, does any... Yeah, yeah, I'm probably running behind, so if you could hold that until later. The LURG study, so this is the same study, these 3300 subjects in Germany referred for heart catheterization. A recent publication was on the correlation with heart failure. Again, they found a very strong correlation, lower levels of 25-hydroxy D predicting heart failure, and after adjusting for everything under the sun, no pun intended, they still came up with very, very, very strong correlation. One of the markers that we use in patients with heart failure is this internal pro-BNP, or just BNP itself, brain natriuretic peptide. They found that lower 25-hydroxy D were associated with higher levels of internal pro-BNP, and again, this held up after adjusting for everything conceivable, and I should point out that this association also occurred with 125-dihydroxy as well as with 25-hydroxy. They showed, similarly, sudden cardiac death strongly associated with lower vitamin D levels, and again, after adjustment, and I've shown the dotted line show the hazard ratios after adjustment, this very strong striking relationship persisted. Stroke, what about stroke? They also looked at future occurrence of stroke in these patients, and they reported these data a little bit differently. They reported it according to Z value. For those of you who are not biostatisticians, the Z score is the number of standard deviations away from the mean, and so this is risk for fatal stroke per 25-hydroxy D Z value, and the same for 125-dihydroxy D Z value, unadjusted data and adjusted data, and you can see that there's a strong predictive value of low 25-hydroxy as well as 125-dihydroxy for future occurrence of stroke. So what this would say is that for every standard deviation of higher, every standard deviation above the mean of 25-hydroxy D, there is a 33% lower risk of a stroke. What about, so the LURIC study, again, was not a general population study because it was looking at patients specifically referred for cardiac catheterization. What about this sort of analysis in a general population? Well, this was published from NHANES-3 data where there was more than 13,000 subjects, and they found a 26% higher all-cause mortality rate in the lowest quartile compared to the highest quartile. So the mortality rate ratio was 1.26, and this was statistically significant for all-cause mortality. Cardiovascular mortality, however, trended to be higher comparing the lowest quartile to the highest quartile, but this did not reach statistical significance. And, you know, I'm not sure why the difference between those. One possibility is that the NHANES study population has a substantially higher 25-hydroxy D than the LURIC study. For example, the lowest quartile in this study is about 17.8, whereas the mean value in the LURIC study was about 17. So because of the higher latitude, that study had substantially lower 25-hydroxy D. In this study, they found a stronger effect in women and in those without baseline cardiovascular disease, suggesting that in those without underlying disease, low vitamin D status may be an even more important predictor of future cardiovascular events. Now, one question that we should ask is what's the data for supplementing vitamin D on reducing mortality, including cardiovascular mortality, and we really don't have a lot of data there in terms of that. But there has been meta-analyses done by Atiyah et al. of 18 trials, a large number of subjects, 57,000, and these were subjects of, most of these trials were for osteoporosis, for preventing fractures. And in these trials as a group, the mean follow-up period was 5.7 years, and the mean dose of vitamin D was 528 international units, so not a large dose and not what we're recommending here to achieve adequate vitamin D status. Nevertheless, they did find a relative risk for mortality from any cause of 0.93, so a 7% reduction in all-cause mortality amongst those treated with vitamin D compared to placebo. And interestingly, this relative risk did not change according to whether the subjects received calcium supplements or not. It all appeared to be a vitamin D effect and not a calcium effect. So, in conclusion, I think that, you know, vitamin D may explain many long-term and very interesting epidemiological observations regarding the variation in cardiovascular events with season and with latitude, as I showed you. And it has, there's biological plausibility to this in that the vitamin D receptor is expressed in more than 30 tissues, including those in the cardiovascular domain or in tissues that impact upon cardiovascular events. There are further biological plausible explanations in that some of the mechanisms that have been elaborated, such as suppression of renin, correlate with beneficial attributes or beneficial effects on cardiovascular events. And then, the final conclusion is there's a lot of studies and now increasing wealth of data, especially over the past 2 to 3 years, that low vitamin D status conveys a substantially higher risk of having subclinical cardiovascular disease, presence of atherosclerosis, presence of peripheral arterial disease, as well as very hard outcomes, such as all-cause death, cardiovascular death, myocardial infarction, stroke, heart failure, and the like. The public health implications are enormous. Ensuring adequate vitamin D levels could potentially reduce the rates of hypertension, the metabolic syndrome, myocardial infarction, heart failure, stroke, sudden cardiac death, and all-cause mortality. And I think taking steps at the grassroots level, like Carol has initiated and we're all here to do, are very, very important. I would say, reflecting Dr. Cedric Garland's earlier comments, you know, where he was citing Osler that overconfidence should be guarded against, that all these are association studies and except for the meta-analyses, these have not been intervention studies. And there are certainly cases in the cardiovascular literature where interventions have not reaped the same fruit that the association studies would have predicted. So I think we still need studies where patients who have cardiovascular disease or are at risk for cardiovascular disease are randomized to ensure adequate vitamin D levels or to placebo. With that, I'll take questions. Thank you very much. Thank you. We know that there's several studies showing that there's an association between vitamin D receptor polymorphisms and autoimmune thyroid disease. And I'm wondering if similar associations have been found between vitamin D receptor polymorphisms and cardiovascular disease. And since we know that even within the normal range of TSH, higher TSH levels are associated with increased cardiovascular disease risk factors, if any of these studies have ever been controlled for thyroid status. Well, regarding the controlling for thyroid status, no, they haven't been. They have been controlled for parathyroid. Some of them have been controlled for parathyroid hormone levels. In particular, the LURK studies were all, one of the covariates that was adjusted for was PTH level. I think the issue of VDR polymorphisms is a very, very interesting one. But I'm not aware that any SNPs polymorphisms have been associated with cardiovascular outcomes per se. There have been, as you mentioned, some associations between VDR polymorphisms and other outcomes. Prostate cancer, I think, has been looked at. Dr. Hallock or others may have other additional information regarding that. There is a notion that all these association studies showing association of low vitamin D levels on cardiovascular disease really shows that vitamin D is a biomarker of activity, physical activity. People who tend to be couched potatoes don't go out and they have low vitamin D levels. People who are more active go out and they have high vitamin D levels. The same thing with seasonal variations. In the winter, we are less active than we are in the summer. What do you think about that? Well, that's a very important point. I think that would be in the category of reverse causation, partly. Someone might say, well, a diabetic is more likely to have low 25-hydroxy-D levels just because they're sicker, more likely to stay inside. Therefore, it's the diabetes status causing the low 25-hydroxy-D and not the reverse. But all I can say, and that's one point to emphasize the need for prospective studies where randomization can control for that. I think that these studies have done the best one can do, in my opinion, in controlling for covariates like that. For example, you mentioned physical activity. That was specifically adjusted for in the LURIC study. But I think there are potential unaccounted-for confounders that still could play some part in all of this. Joe Prendergast, Palo Alto, California. Where are you? Oh, yeah. Hi. There was an article in Hypertension the last few weeks about everything you just said and echoed everything and looked through a lot of the biochemical things that were done. One of the things they pointed out as a possible recommendation was to check the accuracy of what tissue levels of D was like by looking at the B and P. Have you read any other information about that? No, I have not. There are data that the vitamin D status could affect the B and P levels, but it could also be affected by other factors. I guess I would be a little skeptical about taking that value as a marker of tissue in isolation, as a marker of vitamin D status. Angina is a fairly well-categorized clinical syndrome. Has there already been any effort to look at whether patients with angina or reduced angina or reduced need for medications when their vitamin D levels are high? I'm sorry. It was your question to see whether patients have a reduction in angina after correcting from vitamin D deficiency? Are they related to their vitamin D levels? Not that I'm aware of in terms of angina per se, no. But there are plausible mechanisms by which that could occur. The effects on the vasodilatory properties of the artery, on the thrombogenicity, and so forth. That would be plausible, but I'm not aware of studies. Yes. Can you give us some characterization as to how well aware the cardiologists are of this and the vitamin D status? You're not the only one that knows this, right? Don't tell anybody else. I think we're learning. I think having articles like the one that Dr. Hallock co-authored in one of our major journals certainly helps. I think that it's not widely known, but I think the interest is certainly growing.