Editorials EDITORIAL (SEE JOERGENSEN ET AL., P. 1081)

Vitamin D Deficiency Is Not Good for You

n this issue of Diabetes Care, Joergensen tandem mass spectrometry (LC/MS-MS) and Nutrition Examination Survey Iet al. (1) demonstrate a strong associa- is a second asset of the present work (6). (NHANES) indicated a continuum on tion between severe vitamin D defi- Although levels of 25-hydroxyvitamin D mortality risk in the general population ciency and increased mortality in patients [25(OH)D] are high enough to allow de- with an increased cardiovascular risk in with type 1 diabetes. This observation tection by immunoassay techniques, patients with low vitamin D levels. In confirms previous findings in the general avoiding the use of antibodies results in diseases like type 2 diabetes and chronic population and in subgroups at high car- much more reliable measurements with renal failure, similar observations were diovascular risk such as patients with LC/MS-MS. Furthermore, novel data in- made, suggesting that theexcess mortality type 2 diabetes or renal impaired patients dicate that modified forms of 25(OH)D is related to an increase in cardiovascular (2–4). The data by Joergensen et al. com- may be circulating that are falsely over- events. This hypothesis is strengthened plete the picture while carefully avoiding detected or not detected by the antibodies by preclinical and in vitro data. Higher possible criticisms that have made the and may have physiological implications. levels of inflammation, higher blood pres- vitamin D field so hazardous to tread in Still, at present only a few laboratories sure, and increased vascular resistance (via recent years. Indeed, rather than hard sci- around the world are able to measure renin-angiotensin-aldosterone system) are ence, it is hype and media statements that 25(OH)D by LC/MS-MS, and most labora- observed in vitamin D–deficient patients crowd the scene. This article brings with tories therefore continue to use kits. Their and animals, whereas supplementing with itabreathoffreshair. continued widespread use might lead to regular vitamin D or activated forms of Many studies are conducted retro- an overestimation of both the frequency vitamin D (native 1,25-dihydroxyvitamin spectively or measure vitamin D levels and magnitude of vitamin D deficiency. D or a synthetic analog such as paricalci- when patients are already sick. Thanks to This brings us to the other crucial tol) lowers inflammation, lowers blood the excellent registries present in several question of where to put the cutoff that pressure, and decreases vascular resis- Northern European countries—and here defines vitamin D deficiency, even with tance (8). Interestingly, the current study in particular thanks to the database and LC/MS-MS. We do not know, and we do did not find an association between vita- tissue bank of the reputed Steno Diabetes not have a consensus. In addition, there is min D deficiency and the onset or progres- Center—this study is able to assess levels ongoing debate concerning the dosage of sion of nephropathy, whereas in previous of vitamin D measured within 3 years af- nutritional and vitamin D supplements studies in patients with type 2 diabetes ter diagnosis and during up to 25 years of required to achieve vitamin D sufficiency andinmodelsofchronicrenalimpair- clinical follow-up. This way of working (7). The turmoil around a statement re- ment, vitamin D deficiency is associated avoids justified criticisms on studies leased by the Institute of Medicine (IOM) with microalbuminuria and supplement- where vitamin D levels were measured is just an illustration of the problem. For ing with vitamin D (analogs) prevents pro- at the time of severe diseases. Obviously, bone health, defining the level of vitamin gression to proteinuria (9). This finding patients with congestive heart failure or Ddeficiency is relatively easy as the rise suggests that the pathogenesis of mi- cancer are less likely to go out and about in parathyroid hormone levels can be croalbuminuria in type 1 diabetes is a in the sun or eat fatty fish. Only prospec- used as a marker for vitamin D levels that completely different entity from microal- tive studies will help to understand the are too low. When 25(OH)D levels drop buminuria in type 2 diabetes or other relationship between vitamin D levels below 10–20 ng/mL (25–50 nmol/L), forms of renal disease. In type 1 diabetes, and disease. Because vitamin D deficiency parathyroid hormone levels rise, indicat- nephropathy is probably purely microan- in rodents in the early stages of life is ing the switching on of the feedback cas- giopathic and therefore closely related to predictive of future disease, it would be cade, suggesting that this is the level the development of retinopathy, both of of great interest to evaluate vitamin D below which 25(OH)D levels are really which were uninfluenced by vitamin D levels in neonates. We have previously deficient. Here, the authors avoid the cut- status here. demonstrated a twofold risk for the de- off discussion and its clinical relevance In view of the mean age and the sur- velopment of diabetes in NOD mice that by using the lower 10th percentile to de- vival curves of the patients, it is question- had been vitamin D deficient in the first fine deficiency. As a consequence, this able that cardiovascular events contribute 100 days of their lives (5). As vitamin D study’s conclusion is based on data from significantly to mortality in this study. As levels can be measured on dried spots of very severely vitamin D–deficient individ- we have no insight into the direct causes blood of Guthrie cards that are routinely uals. Literature on the effects of vitamin D of death, other events such as accidents or obtained from all newborns in many at levels above 20 ng/mL (50 nmol/L) be- cancer may underlie the increased mor- countries, such a study is both feasible comes very difficult to interpret because tality in vitamin D–deficient patients and could potentially further elucidate theonlyavailabledataarefromasso- with type 1 diabetes. Accidents are hard the link between vitamin D levels and ciation or preclinical studies in animal to account for by vitamin D deficiency, the development of a wide range of dis- models. but an increased cancer risk does come eases. So, what should we conclude with out of other epidemiological studies The choice of measuring vitamin D regard to vitamin D deficiency (and here (10). Almost all cancers are correlated levels using the gold standard method we agree to talk about the true deficiency, with vitamin D deficiency (with the cau- of liquid chromatography followed by thesevereone)?TheNationalHealth tion that in most studies, vitamin D is

care.diabetesjournals.org DIABETES CARE, VOLUME 34, MAY 2011 1245 Vitamin D in patients with type 1 diabetes dosed when cancer is already present), need extreme doses to achieve sufficiency 4. Doorenbos CR, van den Born J, Navis G, except for pancreatic cancer where several (16). de Borst MH. Possible renoprotection by studies point to a lower risk in individuals In conclusion, vitamin D deficiency vitamin D in chronic renal disease: beyond mineral metabolism. Nat Rev Nephrol with lower vitamin D levels (11,12). is associated with increased mortality in – What should we do now? Evidence type 1 diabetic patients as well. Being 2009;5:691 700 fi , 5. Giulietti A, Gysemans C, Stoffels K, et al. linking true vitamin D de ciency ( 10 aware of populations at risk and screening fi fi Vitamin D de ciency in early life acceler- ng/mL [25 nmol/L]) to adverse outcomes for vitamin D de ciency with appropriate ates type 1 diabetes in non-obese diabetic seems solid, and the impact of the vitamin methods is therefore essential. Cutoff – fi fi fi mice. Diabetologia 2004;47:451 462 Dde ciency happens early in life. Avoid- levels for de ciency and suf ciency are 6. Wallace AM, Gibson S, de la Hunty A, ing vitamin D deficiency is thus the under discussion, but levels below 10 Lamberg-Allardt C, Ashwell M. Mea- message. The IOM has confirmed the nu- ng/mL (25 nmol/L) are considered se- surement of 25-hydroxyvitamin D in the tritional advice that existed: 600 IU of vi- verely deficient and should certainly be clinical laboratory: current procedures, tamin D per day as supplements or in food avoided and supplemented. Guidelines performance characteristics and limita- (fortified products, or fatty fish like suggest supplements of vitamin D of tions. Steroids 2010;75:477–488 mackerel...:) for all and 800 IU for peo- 600 IU per day in all and 800 IU in 7. Ross AC, Manson JE, Abrams SA, et al. pleover70yearsofage(7).Onemayof elderly (age .70 years). Many voices The 2011 report on dietary reference in- course also make his or her own by sitting shout for higher doses, but only demon- takes for calcium and vitamin D from the Institute of Medicine: what clinicians need in the sun. Half an hour of exposure of strating these claims by hard clinical data fi fi to know. J Clin Endocrinol Metab 2011; face and hands daily should suf ce to will move the eld forward from hype to 96:53–58 maintain adequate vitamin D levels, but science. 8. Alborzi P, Patel NA, Peterson C, et al. some caveats apply (13). The wavelength Paricalcitol reduces albuminuria and in- of UV light that is necessary to make vita- 1 flammation in chronic kidney disease: a CHANTAL MATHIEU, MD, PHD min D in skin is exactly the same as the 2 randomized double-blind pilot trial. Hy- BART J. VAN DER SCHUEREN, MD, PHD one that ages skin and causes skin cancer. pertension 2008;52:249–255 Moreover, in winter, the sun does not rise 1 9. de Zeeuw D, Agarwal R, Amdahl M, et al. From the Department of Endocrinology, UZ Selective vitamin D receptor activation high enough in the sky in large parts of the Gasthuisberg, Leuven, Belgium; and the 2New York U.S., Europe, and certainly Canada, to Obesity Nutrition Research Center, St. Luke’s with paricalcitol for reduction of albumin- allow us to make any vitamin D. And if Roosevelt Hospital Center, Columbia University uria in patients with type 2 diabetes (VITAL one is dark skinned, exposing skin to sun College of Physicians and Surgeons, New York, study): a randomised controlled trial. Lan- New York. cet 2010;376:1543–1551 has almost no effect on vitamin D levels. “ Corresponding author: Chantal Mathieu, chantal. 10. Holick MF. Vitamin D: its role in cancer Exciting data point toward an individ- [email protected]. prevention and treatment. Prog Biophys ual” set point for vitamin D sufficiency DOI: 10.2337/dc11-0307 Mol Biol 2006;92:49–59 determined by the presence of polymor- © 2011 by the American Diabetes Association. 11. Stolzenberg-Solomon RZ, Jacobs EJ, Arslan Readers may use this article as long as the work is phisms in the carrier protein (DBP), but properly cited, the use is educational and not for AA, et al. Circulating 25-hydroxyvitamin D also polymorphisms in the vitamin D profit, and the work is not altered. See http:// and risk of pancreatic cancer: Cohort Con- receptor and the enzyme responsible for creativecommons.org/licenses/by-nc-nd/3.0/ for sortium Vitamin D Pooling Project of Rarer final activation of vitamin D into 1,25- details. Cancers. Am J Epidemiol 2010;172:81–93 dihydroxyvitamin D (CYP27B1). This 12. Giovannucci E. Vitamin D and cancer in- cidence in the Harvard cohorts. Ann Epi- implies one individual may have to take – more vitamin D in order to achieve vita- Acknowledgments—B.J.V.d.S. is a fellow of demiol 2009;19:84 88 13. Reichrath J. Skin cancer prevention and min D sufficiency than another (14). the Belgian American Educational Foundation. No potential conflicts of interest relevant to UV-protection: how to avoid vitamin D- In this study, no relationship between fi this article were reported. de ciency? Br J Dermatol 2009;161(Suppl. BMI and vitamin D levels was observed 3):54–60 despite overwhelming evidence that a 14. Janssens W, Bouillon R, Claes B, et al. Vi- negative correlation exists (15). The rea- cccccccccccccccccccccccc tamin D deficiency is highly prevalent in son is that these young Danes were all of References COPD and correlates with variants in the normal weight (mean BMI 20 kg/m2). In 1. Joergensen C, Hovind P, Schmedes A, et al. vitamin D-binding gene. Thorax 2010;65: overweight patients, vitamin D accu- Vitamin D levels, microvascular complica- 215–220 mulates in excess fat, leading to low cir- tions, and mortality in type 1 diabetes. Di- 15. Rodríguez-Rodríguez E, Navia B, López- culating levels. Therefore, doctors should abetes Care 2011;34:1081–1085 Sobaler AM, Ortega RM. Vitamin D in certainly screen for vitamin D deficiency 2. Melamed ML, Michos ED, Post W, Astor overweight/obese women and its relation- ship with dietetic and anthropometric var- in obese patients with diabetes and re- B. 25-hydroxyvitamin D levels and the risk of mortality in the general population. iables. Obesity (Silver Spring) 2009;17: plete if necessary. Extra attention should Arch Intern Med 2008;168:1629–1637 778–782 be paid to patients who underwent bar- 3. Joergensen C, Gall MA, Schmedes A, 16. Fish E, Beverstein G, Olson D, Reinhardt iatric surgery causing fat malabsorption. Tarnow L, Parving HH, Rossing P. Vitamin S, Garren M, Gould J. Vitamin D status of These individuals are at extreme risk of D levels and mortality in type 2 diabetes. morbidly obese bariatric surgery patients. vitamin D deficiency and sometimes Diabetes Care 2010;33:2238–2243 J Surg Res 2010;164:198–202

1246 DIABETES CARE, VOLUME 34, MAY 2011 care.diabetesjournals.org