Page 7 - Summer2014

Letters

As I mentioned above, I agree with of such a graph is less complicated than out around 24 ng/mL in the general
Chris's basic message about vitamin D. the analysis of the continuous curve, population, while a great unknown is
But the simplified statement, especially because I believe it is simpler and easier associated with higher levels of 25(OH)
as it is highlighted in a side bar, is likely for most people to understand than the D because of the paucity of data. This
to send people running to the lab and continuous curve, and because using finding is very close to the level at which
panicking about their 25(OH)D levels if it does not substantially change any the risk of major cardiovascular events
they are over 40. Based on the evidence conclusions. bottoms out in cardiac surgery patients.
presented in the article, I'm not sure that The second paragraph after the There is no basis for believing that the
is warranted. We need more research on phrase you quote from my article makes risk increases thereafter in the general
this question. two important points that moderate the population like it does in cardiac sur-
Pia Chamberlain conclusion I drew from these data: one gery patients, but there is only a very
San Jose, California is that vitamin D status this high may weak basis for believing that it does not.
be helpful in some people but harmful As described in the last paragraph of
Chris Masterjohn replies: Thank you for in others. I cited the specific example the "Naked Ape" sidebar, and as stated
voicing your concerns about my article. of the background diet and its content more briefly in the second paragraph
The phrase you quote, "people with of vitamins A and K. The second is a after the phrase you quote, we do not
vitamin D status higher than 40 ng/ml point that you make, that the statistical know if correlations between 25(OH)D
have a higher risk of heart disease," is correlation may not represent a cause- and good or bad health outcomes rep-
part of a sentence explaining why many and-effect relationship. resent cause-and-effect relationships
people might find this surprising. The There could be other contextual because there are a variety of things
initial statement of this fact is found two factors besides the background diet. As besides vitamin D exposure that affect
sentences prior to this, where Figure 2 you point out, for example, these are 25(OH)D. Nevertheless, there is a gen-
is clearly provided as a reference. Fig- cardiac surgery patients, which means eral consistency between the epidemi-
ure 2, in turn, clearly cites the human they clearly have other cardiovascular ology and the animal experiments that
study justifying this statement and the risk factors. Similarly, the Israeli life- cannot be ignored. I cited evidence that
legend explains the data in more detail. guards with a mean 25(OH)D of just both dramatic deficiencies and dramatic
You are correct that the figure does not over 50 ng/mL and a 20-fold increase in excesses of vitamin D have been shown
show where the risk begins increas- the risk of kidney stones (referred to in to contribute to cardiovascular disease
ing. The paper from which these data the "Naked Ape" sidebar) also had signs in animals by inducing soft tissue cal-
are drawn provides a figure (Figure of dehydration and sun damage that cification, and I proposed a mechanism
2 in that paper, not in my article) that may have played a role in their kidney to explain this based on the regulation
attempts to estimate this, and it shows stones. One reasonable conclusion from of vitamin K-dependent proteins by
that the risk of heart disease in those this is that 25(OH)D over 40 ng/mL may vitamins A and D.
patients was lowest at approximately 28 contribute to soft tissue calcification in You are absolutely correct to sug-
ng/mL. Perusing the continuous curve people with other predisposing risk fac- gest that the animal experiments use
from the original research paper leads tors, but not in everyone. This does not more extreme changes in vitamin D
to the conclusion that although the risk make it any less of a concern, because, status than we find associated with
of heart disease begins increasing im- if this is true, we don't yet know what all cardiovascular disease in humans, but
mediately thereafter, it does not become the relevant risk factors are. animal experiments are always per-
statistically significantly greater until Even so, we must keep in mind the formed in the context of a genetically
about 40 ng/mL. I chose to provide my consistency between this study and homogeneous animal population with
own bar graph derived from one of their the meta-analysis cited in Figure 1 of homogeneous background diets meant
tables (Table 2 in that paper, not my my article. The meta-analysis showed to be nutritionally adequate. In humans,
article) because the statistical analysis that cardiovascular disease bottoms variations in vitamin D status occur in
Wise Traditions SUMMER 2014 Wise Traditions 7

137720_text.indd 7 7/1/14 11:40 AM