2813.full

7/30/2019 2813.full

1/6

Serum 25-Hydroxyvitamin D, Diabetes,and Ethnicity in the Third NationalHealth and Nutrition ExaminationSurveyROBERT SCRAGG, PHD1

MARYFRAN SOWERS, PHD2

COLIN BELL, PHD3

OBJECTIVE To determine the association between serum25-hydroxyvitamin D (25OHD)and diabetes risk and whether it varies by ethnicity.

RESEARCH DESIGN AND METHODS We performed an analysis of data fromparticipants who attended the morning examination of the Third National Health and NutritionExamination Survey (1988 1994), a cross-sectional survey of a nationally representative sampleof the U.S. population. Serum levels of 25OHD, which reflect vitamin D status, were availablefrom 6,228 people (2,766 non-Hispanic whites, 1,736 non-Hispanic blacks, and 1,726 Mexi-can Americans) aged 20 years with fasting and/or 2-h plasma glucose and serum insulinmeasurements.

RESULTS Adjusting for sex, age, BMI, leisure activity, and quarter of year, ethnicity-specific odds ratios (ORs) for diabetes (fasting glucose 7.0 mmol/l) varied inversely acrossquartiles of 25OHD in a dose-dependent pattern (OR 0.25 [95% CI 0.110.60] for non-Hispanic whitesand 0.17 [0.08 0.37] for Mexican Americans) in the highest vitamin D quartile(25OHD 81.0 nmol/l) compared with the lowest 25OHD (43.9 nmol/l). This inverse asso-ciation was not observed in non-Hispanic blacks. Homeostasis model assessment of insulinresistance (log

e) was inversely associated with serum 25OHD in Mexican Americans (P

0.0024) and non-Hispanic whites (P

0.058)but notnon-Hispanic blacks(P

0.93), adjustingfor confounders.

CONCLUSIONS These results show an inverse association between vitamin D status anddiabetes, possibly involving insulin resistance, in non-Hispanic whites and Mexican Americans.The lack of an inverse association in non-Hispanic blacks may reflect decreased sensitivity tovitamin D and/or related hormones such as the parathyroid hormone.

Diabetes Care 27:28132818, 2004

There is increasing evidence that vita-min D metabolism affects the risk ofdiabetes. Initial findings from ani-

mal studies showed that insulin releasedfrom the isolated perfused pancreas of therat is lower in vitamin Ddeficient animals

than control animals (1), while pancreaticreceptors for 1,25-dihydroxyvitamin D

3in

-cells have been identified in a number

of species (2). More recently, humanstudies have shown that vitamin D sup-plementation in infancy reduces the risk

of type 1 diabetes during early adulthood(3).

Vitamin D may also have a role in thedevelopment of type 2 diabetes. TaqI vi-tamin D receptor polymorphisms havebeen associated with an insulin secretionindex among Bangladeshi Asians living inLondon, who have a high risk of type 2diabetes (4). The BsmI polymorphism wasassociated with fasting glucose in inactive

German men (5). In the Rancho Bernardostudy (6) of older U.S. Caucasians, the

ApaI polymorphism was associated withfasting plasma glucose and prevalence ofglucose intolerance and the BsmI poly-morphism with the homeostatis modelassessment (HOMA) of insulin resistance.

Given that a number of investigationshave shown that vitamin D receptor poly-morphisms are associated with variousmeasures of glucose metabolism and dia-betes risk, it seems reasonable to concludethat the latter may also be associated with

blood levels of vitamin D. However, therehave been very few epidemiological sur-veys of vitamin D status and newly de-tected type 2 diabetes. Blood levels of 25-hydroxyvitamin D (25OHD), the mainmetabolite of vitamin D, are a marker ofvitamin D status (2). A New Zealandworkforcesurvey identified that newly di-agnosed cases of type 2 diabetes and im-paired glucose tolerance had lower25OHD

3levels than matched control

subjects (7). Serum insulin and plasmaglucose levels collected from participants

in the Zutphen Study at 30-years fol-low-up decreased stepwise with increas-ing tertile of serum 25OHD (8); however,serum 25OHD levels were not related toglucose status among participants in anEnglish study (9).

There appears to be no population-based epidemiological reports of vitaminD status and type 2 diabetes from the U.S.However, the survey with the largestnumber of people with serum 25OHDmeasurements (nearly 20,000) anywhereappears to be the Third National Health

From the 1School of Population Health, University of Auckland, Auckland, New Zealand; the 2Departmentof Epidemiology, School of Public Health, University of Michigan, Ann Arbor, Michigan; and the 3School ofExercise and Nutrition Sciences, Deakin University, Geelong, Victoria, Australia.

Address correspondence and reprint requests to Dr. Robert Scragg, School of Population Health, Univer-sity of Auckland, Private Bag, Auckland, New Zealand. E-mail: [email protected].

Received for publication 27 June 2004 and accepted in revised form 7 September 2004.Abbreviations: 25OHD, 25-hydroxyvitamin D; HOMA, homeostasis model assessment; MEC, mobile

examination center; NHANES III, Third National Health and Nutrition Examination Survey.A table elsewhere in this issue shows conventional and Systeme International (SI) units and conversion

factors for many substances. 2004 by the American Diabetes Association.

E p i d e m i o l o g y / H e a l t h S e r v i c e s / P s y c h o s o c i a l R e s e a r c hO R I G I N A L A R T I C L E

DIABETES CARE, VOLUME 27, NUMBER 12, DECEMBER 2004 2813

7/30/2019 2813.full

2/6

and Nutrition Examination Survey(NHANES III), a national cross-sectionalsurvey representative of the U.S. popula-tion carried out in 19981994. The largenumber of people (6,000) with bloodmeasurements of 25OHD, fasting glu-cose, and insulin provides an ideal oppor-

tunity to confirm, as previous studiessuggest, that serum 25OHD concentra-tions are inversely related to diabetes riskand, if so, whether this association is con-sistent across the main U.S. ethnic groupssurveyed in NHANES III (non-Hispanicwhites, non-Hispanic blacks, and Mexi-can Americans). The diabetes data fromthis survey have been reported previously(10), with higher diabetes prevalencesobserved among non-Hispanic blacksand Mexican Americans compared withnon-Hispanic whites. It is possible that

these ethnic differences in diabetes riskare explained partly by ethnic variationsin vitamin D status, since non-Hispanicblacks have lower levels than non-Hispanic whites (11).

RESEARCH DESIGN AND

METHODS NHANES III is a cross-sectional survey representative of the U.S.civilian noninstitutionalized populationcarried out during 19881994 by the Na-tional Center for Health Statistics of theCenters for Disease Control and Preven-tion. Participants were recruited fromhousehold clusters using a stratified mul-tistage sampling design with over sam-pling of non-Hispa nic bla c ks a ndMexican Americans. Participants wereinitially interviewed at home, followed byphysical examinations at mobile centers.Full details of all survey methods, includ-ing sampling, interview, examination,and laboratory measurements of bloodsamples, have been published (12).

A total of 23,258 adults aged 20years were invited to take part in the sur-vey, of whom 18,825 were interviewed at

home. There were 16,573 persons whoattended the mobile examination centers(MECs), with 8,158 randomly assigned tomorning examinations.

In the home interview, informationwas collected on age, sex, ethnicity (self-assigned as either non-Hispanic white,non-Hispanic black, Mexican American,or other), combined household income inthe last 12 months, and past history ofever being diagnosed by a doctor as hav-ing diabetes (12, interviewers manual).Participants were also asked the number

of times they undertook a range of com-mon physical activities in their leisuretime during thelast month,and metabolicequivalents were assigned for each activ-ity. Participants aged 60 years wereclassified as doing moderate or vigorousactivities if the metabolic equivalents for

any activity were 3.0 or 6.0, respec-tively, and those aged 2059 years weresimilarly classified if the metabolic equiv-alents for any activity were3.5 or7.0,respectively (13).

At th e ME Cs , pa rt ic ip ants we reweighed in underpants and disposablelight clothing and slippers with electronicscales in kilograms to two decimal places,and height was measured with a fixed sta-diometer to the nearest millimeter (12,MEC interviewer manual). BMI was cal-culated as weight in kilograms divided by

the square of height in meters.During the collection of blood sam-ples, participants were asked how longthey had fasted. Participants aged 4074years were given a 75-g glucose equiva-lent oral glucose challenge(Dextol or Tru-tol) and provided an extra blood sample2 h later. Blood samples were centrifuged,aliquoted, and frozen to 70C on site.The frozen plasma and serum sampleswere shipped on dry ice to central labora-tories where they were stored at 70Cuntil analysis (12, manual for medicalt e c hnic ia ns). Pla sma gluc ose wa smeasured by a modified hexokinase enzy-matic method and separate radioimmu-noassay methods used to measure seruminsulin and 25OHD (12, laboratory pro-cedures used for NHANES III). Full meth-ods for serum 25OHD have been reportedpreviously (11). Serum 25OHD concen-trations ranged from 8.7 to 243.6 nmol/lafter excluding one person with a 25OHDvalue of 400.1 nmol/l. Diabetes was de-fined as fasting glucose 7.0 mmol/l or2-h glucose 11.1 mmol/l (10). HOMAestimates of insulin resistance and -cell

function were calculated using fastingglucose and insulin measurements (14).Data in this report are restricted to

non-Hispanic white, non-Hispanic black,and Mexican-American adults 20 yearswho attended the MECs during the morn-ing examination (after fasting overnight)and who provided a blood sample (n 6,228). Excluded participants were thosewho had fasted8 h (n 581); had pre-viously been told by a physician that theyhad diabetes (n 509), to prevent possi-ble confounding from any unknown

treatment-relatedeffects on vitamin D sta-tus; had no fasting plasma glucose or se-rum 25OHD measurement (n 353);had no BMI measurement (n 10); wereof other nationalities (n 273); or hadno sampling weight assigned for themorning examination (n 204).

Statistical analyses were carried outwith SUDAAN (version 8.0.2), using thesampling weights for the morning exam-ination to adjust for over-sampling ofnon-Hispanic blacks and Mexican Amer-icans and to correct SEs and tests of sta-tistical significance for any design effectsarising from clustered sampling. For vari-ables with skewed distributions (insulinand HOMA of insulin resistance and-cell function), the natural logarithmwas used in statistical analyses and toler-ance factors [antilog

e(1.96 SE)] calcu-

lated in place of SEs. Common cut pointsfor serum 25OHD quartiles were appliedto all ethnic groups, since recent evidenceindicates that changes in parathyroid hor-mone levels and calcium absorption oc-cur up to levels of 80100 nmol/l25OHD, which appears to be the maxi-mum threshold of effect for vitamin D(15,16) and which allowed comparison ofeffect at the same 25OHD level betweenethnic groups.

RESULTS Vitamin D concentra-tions were higherin menthan women anddeclined with increasing age (Table 1) butwere not associated with household in-come after adjusting for age, sex, and eth-nicity (results not shown). There weresubstantial ethnic variations in 25OHDconcentrations, reflected in means thatwere higher in non-Hispanic whites, in-termediate in Mexican Americans, andlower in non-Hispanic blacks. As ex-pected, serum 25OHD concentrationsshowed a typical seasonal pattern beinghighest in the summer months (July toSeptember) and lower in the winter

months (January to March) (Table 1).Serum 25OHD concentrations werealso related to lifestyle variables (Table 1).

With regard to leisure time physical activ-ity, the lowest mean concentration wasobserved in those who reported no phys-ical activity in the previous month androse in a stepwise manner with increasingfrequency of physical activity. BMI wasinversely associated with serum 25OHDconcentrations after adjusting for othercovariates.

The weighted crude prevalence of un-

Vitamin D and diabetes

2814 DIABETES CARE, VOLUME 27, NUMBER 12, DECEMBER 2004

7/30/2019 2813.full

3/6

diagnosed diabetes (fasting glucose 7.0mmol/l) was 2.8% for the total sampleand was greater in non-Hispanic blacks(3.7%) and Mexican Americans (3.3%)compared with non-Hispanic whites(2.6%). As has been previously reported,diabetes risk was positively associated(P 0.05) with age and BMI (results not

shown) and decreased in participantswho did vigorous activity (odds ratio[OR] 0.57 [95% CI 0.291.11]) com-pared with inactive participants (P 0.10). Diabetes risk was unrelated tohousehold income (results not shown).

Measures of insulin resistance and-cell function also varied with ethnicityafter adjusting for age and sex. Adjustedmean fasting insulin was significantlyhigher in Mexican Americans (62.2pmol/l [95% CI 59.8 64.6]; P 0.0001)and non-Hispanic blacks (59.7 pmol/l

[57.462.1]; P 0.0001) comparedwith non-Hispanic whites (51.4 pmol/l[49.453.5]). Mean level of HOMA of in-sulin resistance was also higher in Mexi-can Americans (2.51 [2.412.61]; P 0.0001) and non-Hispanic blacks (2.36[2.272.46]; P 0.0001) compared withnon-Hispanic whites (2.01 [1.942.09]).

Similarly, mean level of HOMA -cellfunction was highest in non-Hispanicblacks (114 [110119]; P 0.0001) fol-lowed by Mexican Americans (111 [107116]; P 0.0001) when compared withthe level in non-Hispanic whites (99 [95102]).

There was an inverse association be-tween quartiles of serum 25OHD and di-abetes, which varied between ethnicgroups (Table 2). An inverse trend be-tween odds of diabetes and 25OHD wasfound in non-Hispanic whites and Mexi-

can Americans but not non-Hispanicblacks, after adjusting for age, sex, BMI,leisure time physical activity, and season.For non-Hispanic whites, participants inthe highest vitamin D quartile had a four-fold lower odds of diabetes defined by ei-ther fasting glucose or 2-h glucose

compared with those in the lowest vita-min D quartile for fasting glucose. A sim-ilar pattern was seen among Mexican

Americans, particularly for diabetes de-fined by fasting glucose where the OR was0.17 for the highest vitamin D quartilecompared with the lowest, while for dia-betes defined by 2-h glucose there was adose-response effect of decreasing diabe-tes odds with increasing vitamin D, al-though the individual ORs were notstatistically significant (P 0.05).

The pattern for non-Hispanic blacks

was opposite that of the other two ethnicgroups (Table 2). The odds of diabetesdefined by fasting glucose did not show alinear trend across the quartiles, beingsignificantly (P 0.05) higher for the 2ndand 4th vitamin D quartiles but similar forthe 3rd quartile compared with the low-est, while for 2-h glucose, the diabetesORs for the highest 3 vitamin D quartileswere not significantly different from thelowest quartile (P 0.05).

There was no confounding betweenleisure time physical activity and serum25OHD on diabetes, with ORs in eachethnic group changing little betweenmodels that included these two variablessingularly or together; there was also noconfounding between BMI and serum25OHD on diabetes ORs (data notshown).

Multiple linear regression coefficientsfor fasting glucose, insulin, and theHOMA of insulin resistance and -cellfunction as outcome variables were calcu-lated against serum 25OHD as the inde-pendent variable, adjusting for age, sex,BMI, leisure physical activity, and season

within each ethnic group to see if therewas consistency in the variation of any ofthese measures with the ethnic patternshown in Table 2 (Table 3). Vitamin Dwas inversely associated (P 0.05) withfasting and 2-h glucose, fasting insulin,and the HOMA of insulin resistance inMexican Americans, while the negativecoefficients for fasting insulin and HOMAof insulin resistance in non-Hispanicwhites just failed to reach significance(P 0.061 and 0.058, respectively). Incontrast with the above two ethnic

Table 125OHD levels by sex, age, ethnicity, BMI, leisure time physical activity, and season,adjusted for all other variables in the table

Variable n (%)

25OHD

(nmol/l) P

Sex

Men 2,939 (47.2) 78.8 0.9 *Women 3,289 (52.8) 72.6 0.8 0.0001

Age (years)2039 2,693 (43.2) 81.0 1.1 *

4059 1,704 (27.4) 71.7 1.0 0.0001

60 1,831 (29.4) 69.5 0.9 0.0001

Race/ethnicityNon-Hispanic white 2,766 (44.4) 79.6 0.7 *

Non-Hispanic black 1,736 (27.9) 49.1 0.9 0.0001Mexican American 1,726 (27.7) 66.0 1.0 0.0001

BMI quartile (kg/m2)

23.0 1,550 (24.9) 80.6 1.2 *

23.126.1 1,549 (24.9) 77.1 1.0 0.025

26.229.9 1,577 (25.3) 73.8 1.2 0.000130.0 1,552 (24.9) 68.7 1.3 0.0001

Leisure time physical activity (times in last

month)None 1,269 (20.4) 68.4 1.5 *

Moderate12 1,741 (28.0) 72.4 0.9 0.009Moderate12 2,229 (35.8) 79.1 1.0 0.0001

Vigorous12 645 (10.4) 75.7 2.1 0.007Vigorous12 344 (5.5) 81.4 1.4 0.0001

Season

January to March 1,614 (25.9) 67.8 1.6 *April to June 1,649 (26.5) 73.2 1.2 0.014

July to September 1,516 (24.3) 81.7 1.5 0.0001October to December 1,449 (23.3) 74.7 1.5 0.002

Total 6,228 (100)Data are means SE unless otherwise indicated. Based on those who attended NHANES III morningexamination, fasted 8 h (excluding diagnosed diabetes), and age 20 years.*Reference category for Pvalue.

Scragg, Sowers, and Bell


7/30/2019 2813.full

4/6

groups, P values for coefficients amongnon-Hispanic blacks were all well above0.05, and vitamin D was not associatedwith HOMA -cell function in any ethnicgroup.

The relative contribution of ethnicdifferences in risk factors was evaluatedby assessing changes in the fasting diabe-tes OR for Mexican Americans comparedwith non-Hispanic whites, which was2.28 (95% CI 1.523.41), adjusting forage and sex. Further adjustment for BMIchanged the OR to 1.85 (1.192.88), andfor 25OHD it changed to 1.71 (1.072.71), while the OR decreased to 1.48(0.932.37) adjusting for both variables.

Adjusting for leisure time physical activitychanged the OR only slightly to 2.14(1.42- 3.22). A similar pattern was foundfor 2-h diabetes ORs. Thus, ethnic varia-tions in BMI and vitamin D status werethe main determinants of the increaseddiabetes risk in Mexican Americans com-

pared with non-Hispanic whites. We didnot repeat this calculation for non-Hispanic blacks because there was no as-sociation between serum 25OHD anddiabetes for this group.

CONCLUSIONS We have shownin a large sample representative of thea dult U.S. popula t ion t ha t se rum25OHD, a measure of vitamin D status, isinversely associated with diabetes riskand measures of insulin resistance moreso than with -cell function in non-Hispanic whites and Mexican Americansbut not in non-Hispanic blacks, aftercontrolling for known major diabetes riskfactors.

The inverse association between vita-min D and diabetes risk in non-Hispanicwhites and Mexican Americans is consis-tent with previous epidemiological stud-ies in other ethnic groups (7,8). Further,our findings for vitamin D and insulin re-

sistance are consistent with studies re-porting that insulin sensitivity measuredby glucose clamp is positively associatedwith serum 25OHD (1719) or increasedby vitamin D supplementation (20). Thelack of confounding between vitamin Dand leisure physical activity and also be-tween vitamin D and BMI suggests thatvitamin D affects diabetes risk by a mech-anism separate from those of the othertwo risk factors.

Vitamin D deficiency results in hyper-parathyroidism (2,15), through which itmay influence glucose metabolism. Pa-tients with hyperparathyroidism have anincreased prevalence of diabetes and in-sulin resistance, and parathyroidectomyimproves their glucose intolerance (21).Skeletal muscle, a key component of theinsulin resistance syndrome (22), mayalso be involved since vitamin D receptorshave been identified in that tissue (23).

The contrasting lack of any inverse

Table 2Ethnicity-specific diabetes ORs (95% CI) associated with quartiles of 25OHD, adjusted for age, sex, BMI, leisure time physicalactivity, and season

25OHD (nmol/l)

Non-Hispanic whites Non-Hispanic blacks Mexican Americans

Diabetes

OR (95% CI)

Diabetes

OR (95% CI)

Diabetes

OR (95% CI)Yes No Yes No Yes No

Fasting glucose

43.9* 25 245 1.00 27 838 1.00 27 387 1.00

44.060.4 28 542 0.35 (0.150.81) 28 451 2.83 (1.505.33) 26 479 0.52 (0.271.00)60.580.9 31 783 0.27 (0.120.64) 10 251 1.10 (0.502.45) 9 490 0.21 (0.080.52)

81.0 26 1086 0.25 (0.110.60) 7 124 3.40 (1.0710.86) 6 302 0.17 (0.080.37)2-h glucose

45.2 17 132 1.00 24 329 1.00 29 155 1.0045.360.4 26 288 0.45 (0.181.13) 17 164 1.52 (0.743.16) 28 154 0.85 (0.352.08)

60.579.9 37 373 0.42 (0.200.90) 12 96 1.32 (0.592.93) 21 155 0.76 (0.321.79)

80.0 29 484 0.28 (0.120.66) 3 55 0.76 (0.183.18) 9 108 0.45 (0.151.38)

Based on those who attended NHANES III morning examination and fasted 8 h (excluding diagnosed diabetes).*Quartiles for all ethnic groups aged 20 years.P 0.05, P 0.01 vs. quartile 1. Quartiles for all ethnic groups aged 4074 years.

Table 3Regression coefficient (SE) for association between plasma glucose, serum insulin, and HOMA of insulin resistance and -cellfunction as outcome variables and serum 25OHD (nmol l 1 10 1), adjusted for age, sex, BMI, leisure physical activity, and season

Outcome variables

Non-Hispanic white Non-Hispanic black Mexican American

(SE) P (SE) P (SE) P

Fasting glucose (mmol/l) 0.009 (0.009) 0.32 0.023 (0.015) 0.13 0.031 (0.012) 0.010Log

efasting insulin (pmol/l) 0.007 (0.004) 0.061 0.003 (0.006) 0.69 0.012 (0.005) 0.015

Loge

HOMA of insulin resistance 0.009 (0.004) 0.058 0.001 (0.007) 0.93 0.016 (0.005) 0.0024Log

eHOMA -cell function* 0.003 (0.003) 0.33 0.006 (0.007) 0.37 0.009 (0.005) 0.091

n (for fasting blood sample) 2,766 1,736 1,726

2-h glucose measurement (mmol/l) 0.035 (0.035) 0.32 0.068 (0.050) 0.18 0.135 (0.067) 0.047n (for 2-h blood sample) 1,386 700 659

Based on those who attended NHANES III morning examination and fasted 8 h (excluding diagnosed diabetes). *Also adjusted for HOMA of insulin resistance.



7/30/2019 2813.full

5/6

association of vitamin D with diabetes riskand insulin resistance in non-Hispanicblacks was unexpected, particularly giventheir lowvitamin D levels, which has beenreported previously (11). The lack of anassociation across vitamin D quartiles,compared with quartile 1, was consistent

for both measures of glucose metabolism(Table 2), aside from participants in the2nd and 4th quartiles who had raised fast-ing diabetes ORs. Their results are notpart of a consistent linear trend across vi-tamin D quartiles and possibly represent afalse-positive finding.

The reasons for the lack of an inverseassociation with vitamin D in non-Hispanic blacks are unclear, but an expla-nation of this ethnic variation in vitaminD effect may provide new insights intoany possible protective mechanisms re-

lated to vitamin D. One possibility is thatthere is a threshold effect that varies withethnicity. The number of non-Hispanicblack participants in the highest vitaminD quartile for the total sample was small(Table 2), and if there is an ethnicity-related threshold effect, our results can-not exclude a beneficial effect for non-Hispanic blacks at 25OHD levels of 80nmol/l. However, the nonsignificant mul-tiple regression coefficients from the totalnonHispanic black sample (Table 3) ar-gue against an association between vita-min D and glucose status among African

Americans in this study.Non-Hispanic blacks may have a de-

creased sensitivity to the effects of vitaminD and/or related hormones. Bone densityis increased in blacks compared withwhites (24), despite the former having el-evated parathyroid hormone blood levels(24,25), which should result in increasedbone resorption and decreased bone min-eral density. This suggests a decreasedsensitivity among non-Hispanic blacks tothe effects of parathyroid hormone (25).The latter hypothesis has been confirmed

by a recent study (26) which found thatblacks have a higher resistance thanwhites to the effect of parathyroid hor-mone on bone resorption. Cosman et al.(26) have proposed that whites may havedeveloped an increased skeletal sensitiv-ity to the effects of the parathyroid hor-mone in order to maintain calciumhomeostasis in general body tissues by in-creasing calcium supply from the skele-ton to compensate for increased urinarycalcium excretion. Perhaps some othertissues, such as skeletal muscle, also show

increased sensitivity to parathyroid hor-mone in non-Hispanic whites and inMexican Americans who have similarbone mineral density to whites (27).

This report has some limitations. Be-cause NHANES III is a cross-sectionalstudy, we cannot be certain that vitamin D

status affected glucose metabolism ratherthan vice versa. Further, there are no di-rect measures of insulin resistance and-cell function in this study. We are un-clear as to why a significant vitamin Deffect was found in non-Hispanic whitesfor diabetes ORs definedby 2-h glucose inTable 2 but not for 2-h glucose regressioncoefficients in Table 3, which may reflectrandom variation in the data influenced,perhaps, by the weighting variable.

The inverse association between vita-min D and diabetes we have observed in

non-Hispanic whites and Mexican Amer-icans but not non-Hispanic blacks, if con-firmed by further research, would haveimportant public health implications.Our findings would offer an explanation,in part, for the generally lower prevalenceof type 2 diabetes observed in Caucasianpopulations around the world comparedwith other ethnicities. In NHANES III,ethnic differences in serum 25OHD ex-plained much of the increased diabetesodds in Mexican Americans comparedwith non-Hispanic whites, as did ethnicdifferences in BMI. Simple and cheap pre-ventive strategies to increase vitamin Dlevels, such as increased sun exposure orvitamin D supplementation are available.However, further research is needed toconfirm our findings and to determinepossible mechanisms of any preventiveeffect from vitamin D against diabetes.

Acknowledgments C.B. was supported bythe Health Research Council of New Zealand.

References

1. Norman AW, Frankel BJ, Heldt AM,Grodsky: Vitamin D deficiency inhibitspancreatic secretion of insulin. Science209:823825, 1980

2. Zitterman A: Vitamin D in preventivemedicine: are we ignoring the evidence?Br J Nutr89:552572, 2003

3. Hypponen E, Laara E, Reunanen A, Jarve-lin M-R, Virtanen SM: Intake of vitamin Dand risk of type 1 diabetes: a birth-cohortstudy. Lancet 358:15001503, 2001

4. Ogunkolade BW, Boucher BJ, Prahl JM,Bustin SA, Burrin JM, Noonan K, NorthBV, Mannan N, McDermott MF, DeLuca

HF, Hitman GA: Vitamin D receptor(VDR) mRNA and VDR protein levels inrelation to vitamin D status, insulin secre-tory capacity, and VDR genotype in Ban-gladeshi Asians. Diabetes 51:22942300,2002

5. Ortlepp JR, Metrikat J, Albrecht M, von

Korff A, Hanrath P, Hoffmann R: The vi-tamin D receptor gene variant and physi-cal activity predicts fasting glucose levelsin healthy young men. Diabet Med20:451454, 2003

6. Oh J-Y, Barrett-Connor E: Association be-tween vitamin D receptor polymorphismand type 2 diabetes or metabolic syn-drome in community-dwelling olderadults: the Rancho Bernardo study. Me-tabolism 51:356359, 2002

7. Scragg R, Holdaway I, Singh V, Metcalf P,Baker J, Dryson E: Serum 25-hydroxyvi-tamin D

3levels decreased in IGT and di-

abetes. Diabetes Res Clin Prac 27:181

188, 19958. Baynes KCR, Boucher BJ, Feskens EJM,

Kromhout D: Vitamin D, glucose toler-ance and insulinaemia in elderly men.Diabetologia 40:344347, 1997

9. Wareham NJ, Byrne CD, Carr C, Day NE,Boucher BJ, Hales CN: Glucose toleranceis associated with altered calcium ho-meostasis: a possible link between in-creased serum calcium concentration andcardiovascular disease mortality. Metabo-lism 46:11711177, 1997

10. Harris MI, Flegal KM, Cowie CC, Eber-hardt MS, Goldstein DE, Little RR, Wied-meyer H-M, Bryd-Holt DD: Prevalence ofdiabetes, impaired fasting glucose, andimpaired glucose tolerance in U.S. adults:the Third National Health and NutritionExamination Survey, 19881994. Diabe-tes Care 21:518524, 1998

11. LookerAC, Dawson-Hughes B, Calvo MS,Gunter EW, Sahyoun NR: Serum 25-hy-droxyvitamin D status of adolescents andadults in two seasonal subpopulationsfrom NHANES III. Bone 30:771777,2002

12. National Center for Health Statistics:Third National Health and Nutrition Exam-ination Survey, 19881994, Reference Man-

uals and Reports (CD-ROM). Hyattsville,MD, Centers for Disease Control and Pre-vention, 1996

13. Crespo CJ, Keteyian SJ, Heath GW, Sem-pos CT: Leisure-time physical activityamong U.S.adults: results from the ThirdNational Health and Nutrition Examina-tion Survey. Arch Intern Med 156:9398,1996

14. Matthews DR, Hosker JP, Rudenski AS,Naylor BA, Treacher DF, Turner RL: Ho-meostasis model assessment: insulin re-sistance and -cell function from fastingplasma glucose and insulin concentra-

Scragg, Sowers, and Bell


7/30/2019 2813.full

6/6

tions in man. Diabetologia 28:412419,1985

15. Vieth R: Vitamin D supplementation, 25-hyrdoxyvitamin D concentrations andsafety. Am J Clin Nutr69:842856, 1999

16. Heaney RP,Dowell MS, Hale CA, BendichA: Calcium absorption varies within the

reference range for serum 25-hydroxyvi-tamin D. J Am Coll Nutr 22:142146,2003

17. Lind L, Pollare T, Hvarfner A, Lithell L,Sorenson OH, Ljunghall S: Long-termtreatment with active vitamin D (alphac-alcidol) in middle-aged men with im-paired glucose tolerance: effects oninsulin secretion and sensitivity, glucosetolerance and blood pressure. Diabetes Res11:141147, 1989

18. Lind L, Hanni A, Lithell L, Hvarfner A,Sorenson OH, Ljunghall S: Vitamin D isrelated to blood pressure and other car-diovascular risk factors in middle-aged

men. Am J Hypertens 8:894901, 199519. Chiu KC, Chu A, Go VLW, Saad MF: Hy-

povitaminosis D is associated with insulinresistance and cell dysfunction. Am JClin Nutr79:820 825, 2004

20. Borissova AM, Tankova T, Kirilov G, Da-kovska L, Kovacheva R: The effect of vita-

min D3 on insulin secretion andperipheral insulin sensitivity in type 2 di-abetic patients. Int J Clin Pract 57:258261, 2003

21. Taylor WH, Khaleeli AA: Coincident dia-betes mellitus and primary hyperparathy-roidism. Diabetes Metab Res Rev 17:175180, 2001

22. DeFronzo RA: The triumvirate: -cell,muscle, liver: a collusion responsible forNIDDM. Diabetes 37:667687, 1988

23. Simpson RU, Thomas GA, Arnold AJ:Identification of 1,25(OH)

2D3

receptorsandactivities in skeletal andheart muscle.J Biol Chem 260:88828891, 1985

24. Kleerekoper M, Nelson DA, Peterson EL,Flynn MJ, Pawluszka AS, Jacobson G, Wil-son P: Reference data for bone mass, calcio-tropic hormone, and biochemical markersof bone remodelling in older (5575) post-menopausal white and black women.J BoneMiner Res 9:12671276, 1994

25. Bell NH, Greene A, Epstein S, OexmannMJ, Shaw S, Shary J: Evidence for alter-ation of the vitamin D-endocrine systemin blacks. J Clin Invest 76:470 473, 1985

26. Cosman F, Morgan DC, Nieves JW, ShenV, Luckey MM, Dempster DW, Lindsay R,Parisien M: Resistance to bone resorbingeffects of PTH in black women. J BoneMiner Res 12:958966, 1997

27. Morton DJ, Barrett-Connor E, Kritz-Silver-stein D, Wingard DL, Schneider DL: Bonemineral density in postmenopausal Cauca-sian, Filipina, and Hispanic women. Int JEpidemiol 32:150156, 2003



Documents

2813.full