Abstract
OBJECTIVE
To describe baseline characteristics of the Vitamin D and Type 2 Diabetes (D2d) study, the first large U.S. diabetes prevention clinical trial to apply current American Diabetes Association (ADA) criteria for prediabetes.
RESEARCH DESIGN AND METHODS
This is a multicenter (n = 22 sites), randomized, double-blind, placebo-controlled, primary prevention clinical trial testing effects of oral daily 4,000 IU cholecalciferol (D3) compared with placebo on incident diabetes in U.S. adults at risk for diabetes. Eligible participants were at risk for diabetes, defined as not meeting criteria for diabetes but meeting at least two 2010 ADA glycemic criteria for prediabetes: fasting plasma glucose (FPG) 100–125 mg/dL, 2-h postload glucose (2hPG) after a 75-g oral glucose load 140–199 mg/dL, and/or a hemoglobin A1c (HbA1c) 5.7–6.4% (39–46 mmol/mol).
RESULTS
A total of 2,423 participants (45% of whom were women and 33% nonwhite) were randomized to cholecalciferol or placebo. Mean (SD) age was 60 (9.9) years and BMI 32.1 (4.5) kg/m2. Thirty-five percent met all three prediabetes criteria, 49% met the FPG/HbA1c criteria only, 9.5% met the 2hPG/FPG criteria only, and 6.3% met the 2hPG/HbA1c criteria only. Black participants had the highest mean HbA1c and lowest FPG concentration compared with white, Asian, and other races (P < 0.01); 2hPG concentration did not differ among racial groups. When compared with previous prediabetes cohorts, the D2d cohort had lower mean 2hPG concentration but similar HbA1c and FPG concentrations.
CONCLUSIONS
D2d will establish whether vitamin D supplementation lowers risk of diabetes and will inform about the natural history of prediabetes per contemporary ADA criteria.
Introduction
Prediabetes, typically defined as blood glucose concentrations above normal but below the threshold for diabetes, is a disease risk state that predicts an increased probability of developing diabetes and may itself be associated with health risks and complications (1). Although lifestyle changes can reduce the rate of progression to diabetes, achieving as well as maintaining sufficient lifestyle change is challenging (2). Therefore, simple, sustainable, and complementary prevention approaches are needed. The Vitamin D and Type 2 Diabetes (D2d) study is the largest clinical trial to examine the causal relationship between vitamin D supplementation and the development of diabetes in people at risk for diabetes (3). D2d is also the largest U.S.-based study to have assembled and followed a contemporary cohort of people at risk for diabetes, defined as meeting at least two prediabetes criteria by the American Diabetes Association (ADA) (4).
In 2003, the ADA revised the criteria for prediabetes to lower the threshold for fasting plasma glucose (FPG) from 110 to 100 mg/dL (5.6 mmol/L), and in 2010, hemoglobin A1c (HbA1c) of 5.7–6.4% (39–46 mmol/mol) was added as a criterion based on evidence of increased diabetes complications at these glycemic ranges (4). The 2-h postload glucose (2hPG) after a 75-g oral glucose load criterion was unchanged. The expanded criteria have been controversial (5,6). First, the natural history of prediabetes, based on the ADA’s current definition, has not been established in the modern era. Second, most data on the natural history of prediabetes are >20 years old (7–10) or were not conducted in populations generalizable to the current U.S. population (8–12). Third, some have argued that lowering the FPG threshold and adding the HbA1c criterion increase the prevalence of prediabetes without a clear association with clinically important outcomes and may lead to an unnecessary medicalization of prediabetes (13,14). Finally, there is evidence of interindividual variation in HbA1c relative to underlying glucose levels, with a tendency for black individuals to have higher HbA1c compared with whites with similar glucose levels (15).
In the current report, we describe the baseline characteristics of the D2d prediabetes cohort and compare them with prior diabetes prevention studies that used different enrollment criteria. As the largest clinical trial to enroll a contemporary cohort of American adults with prediabetes, D2d will fill important gaps in knowledge related to the current definition of prediabetes.
Research Design and Methods
Overview of Study Design
D2d is a U.S.-based, multicenter, randomized (1:1), double-blind, placebo-controlled, parallel-group, primary prevention clinical trial comparing oral administration of 4,000 IU/day of cholecalciferol (vitamin D3) versus placebo in people with prediabetes who are followed for incident diabetes for ∼3 years after randomization. Cancer and cardiovascular disease are key secondary outcomes. The design of D2d has been published (3). The study is approved and monitored by an independent data- and safety-monitoring board and the institutional review board of each collaborating clinical research site.
Study Population and Setting
Target participants were adults at risk for diabetes. At the baseline visit, eligible participants met at least two of three glycemic criteria for prediabetes established by the ADA in 2010 (4): FPG 100–125 mg/dL (5.6–6.9 mmol/L), or impaired fasting glucose (IFG); 2hPG after a 75-g glucose load 140–199 mg/dL (7.8–11.0 mmol/L), or impaired glucose tolerance (IGT); HbA1c 5.7–6.4% (39–46 mmol/mol), or—our designation—impaired A1c (iA1c). Other entry criteria included age ≥30 years (≥25 years for American Indians, Alaska Natives, Native Hawaiians, or other Pacific Islanders) and BMI 24–42 kg/m2 (22.5–42 kg/m2 for Asians). Key exclusion criteria included FPG, 2hPG, or HbA1c in the diabetes range; conditions affecting HbA1c, such as hemoglobinopathies; treatment with medications approved for treatment of diabetes; hyperparathyroidism; nephrolithiasis; bariatric surgery; use of supplements with vitamin D or calcium above study limitations (1,000 IU/day and 600 mg/day, respectively); regular use of tanning beds; medications or conditions that could interfere with absorption or metabolism of vitamin D; hypercalcemia; hypercalciuria; or chronic kidney disease (estimated glomerular filtration rate [eGFR] <50 mL/min per 1.73 m2).
D2d is being conducted at 22 U.S. collaborating clinical sites (www.d2dstudy.org/sites). Several sites serve populations with substantial racial diversity, while 12 sites are located at high latitudes (above 37° N) to include participants with lower ultraviolet B exposure. D2d is an event-driven trial that will continue until the required number of diabetes outcome events is reached. Results are expected in 2019.
Screening and Baseline Assessment
Prescreening procedures were site specific and included telephone prescreenings, medical chart reviews, and—at some sites—targeted laboratory testing with FPG and HbA1c. If potential participants met prescreening criteria, they were invited for in-person screening, which occurred in two steps. At screening visit 1, nonglycemic eligibility criteria (e.g., medical history, laboratory criteria for safety) were confirmed and glycemic criteria for prediabetes were preliminarily evaluated by measuring FPG and HbA1c. Algorithms using the screening visit 1 FPG and HbA1c results guided sites as to which participants should proceed to the next screening visit. At screening visit 2, a 75-g oral glucose tolerance test was performed after an 8-h overnight fast, and FPG, 2hPG, and HbA1c were analyzed by the D2d central laboratory to determine final eligibility. Screening visit 2 served as the baseline visit for participants who were randomized.
At screening visit 1, participants self-reported demographics such as age, race, ethnicity, employment, education, and household income. Racial and ethnic categorization followed National Institutes of Health (NIH) guidelines. Personal health history, including smoking history, family history of diabetes, medication use, and use of dietary supplements, were captured by targeted questionnaires. Height and weight were measured with a stadiometer and calibrated balance beam or electrical digital scale, respectively, following standardized procedures.
Laboratory Methods
Specimens were processed locally. Plasma for glucose measurement was shipped frozen to the D2d central laboratory at the University of Vermont’s Laboratory for Clinical Biochemistry Research (Colchester, VT). Whole blood for HbA1c measurement was shipped refrigerated. HbA1c was measured using an ion-exchange high-performance liquid chromatography method (Tosoh G8; Tosoh Bioscience, South San Francisco, CA). This method is certified by NGSP (formerly the National Glycohemoglobin Standardization Program), and the D2d central laboratory is certified by NGSP as a Level I Laboratory with documented traceability to the Diabetes Control and Complications Trial (DCCT) reference method (16). Plasma glucose was measured using a hexokinase method (Roche Glucose HK Gen.3 on the Cobas Integra 400 or Cobas c311 analyzer; Roche Diagnostics, Indianapolis, IN) and standardized against isotope dilution mass spectrometry. Creatinine was measured at each clinical site, and eGFR was calculated using the Chronic Kidney Disease Epidemiology Collaboration equation (17). Serum 25-hydroxyvitamin D from the baseline and yearly visits will be analyzed at the end of the study. Serum insulin will also be analyzed at the end of the study and oral glucose tolerance test–based indices of insulin secretion and sensitivity will be derived (3).
Statistical Analysis
Baseline characteristics are reported as mean (SD) unless otherwise specified. Characteristics by sex, race, and ethnicity were compared using ANOVA tests and χ2 tests as appropriate. Two-sided P values <0.05 were considered statistically significant. SAS (version 9.4; SAS Institute, Cary, NC) was used for all analyses.
Results
Recruitment
Recruitment occurred from October 2013 through December 2016. At baseline, 3,288 people met at least one of the three ADA glycemic criteria for prediabetes and all other nonglycemic criteria. After exclusion of 865 people who met only one of the three criteria, 2,423 participants were randomized (Supplementary Fig. 1). Median enrollment at each site was 88 participants (range 29–318).
Demographics Overall
Baseline characteristics are shown in Table 1. At baseline, mean age was 60.0 years and BMI 32.1 kg/m2. Waist circumference was 107.1 cm in men and 102.5 cm in women. A family history of diabetes was reported by 62.5% of participants, and 7.6% of women reported a history of gestational diabetes mellitus. D2d enrolled a diverse cohort: 33.3% of participants identified as nonwhite and 9.3% identified as Hispanic or Latino. The majority were employed at least part-time (58.6%), approximately half (50.8%) completed postsecondary education (i.e., bachelor’s degree or higher), and more than one-third (37.5%) had an annual income over $75,000. Few (6.5%) reported current smoking. The most prevalent comorbidities were hypercholesterolemia (reported by 55.5% of the cohort) and hypertension (53.5%). In terms of supplement use, 42.8% of participants were taking vitamin D supplements (mean 732 IU/day) and 33.2% were taking calcium supplements (mean 312 mg/day). Across all participants (including those who did not take supplements), mean intakes were 313 IU/day for vitamin D and 103 mg/day for calcium.
Table 1.
Demographics and clinical characteristics by sex in D2d
| Overall (n = 2,423) | Men (n = 1,337) | Women (n = 1,086) | P for men vs. women | |
|---|---|---|---|---|
| Diabetes risk factors | ||||
| Age, years, mean (SD) | 60.0 (9.9) | 60.4 (10.3) | 59.4 (9.5) | 0.01 |
| Age range, years, n (%) | <0.01 | |||
| 25–44 | 209 (8.6) | 122 (9.1) | 87 (8.0) | |
| 45–59 | 907 (37.4) | 457 (34.2) | 450 (41.4) | |
| ≥60 | 1,307 (53.9) | 758 (56.7) | 549 (50.6) | |
| BMI, kg/m2, mean (SD) | 32.1 (4.5) | 31.5 (4.3) | 32.7 (4.6) | <0.01 |
| Waist circumference, cm, mean (SD) | 105.0 (11.7) | 107.1 (11.6) | 102.5 (11.4) | <0.01 |
| Self-reported family history of diabetes, n (%) | 1,514 (62.5) | 774 (57.9) | 740 (68.1) | <0.01 |
| Self-reported gestational diabetes mellitus, n (%) women | NA | NA | 83 (7.6) | NA |
| Demographics, n (%) | ||||
| Race | 0.04 | |||
| Asian | 130 (5.4) | 84 (6.3) | 46 (4.2) | |
| Black or African American | 616 (25.4) | 318 (23.8) | 298 (27.4) | |
| White | 1,616 (66.7) | 902 (67.5) | 714 (65.7) | |
| Other1 | 61 (2.5) | 33 (2.5) | 28 (2.6) | |
| Hispanic or Latino ethnicity | 225 (9.3) | 89 (6.7) | 136 (12.5) | <0.01 |
| Residence above 37° N latitude | 1,790 (73.9) | 1,021 (76.4) | 769 (70.8) | <0.01 |
| Socioeconomic, n (%) | ||||
| Current employment | <0.01 | |||
| Homemaker | 81 (3.3) | 5 (0.4) | 76 (7.0) | |
| Employed at least part-time | 1,421 (58.6) | 779 (58.3) | 642 (59.1) | |
| Retired | 781 (32.2) | 479 (35.8) | 302 (27.8) | |
| Not employed | 63 (2.6) | 29 (2.2) | 34 (3.1) | |
| Other | 63 (2.6) | 40 (3.0) | 23 (2.1) | |
| Unknown or not reported | 14 (0.6) | 5 (0.4) | 9 (0.8) | |
| Education | <0.01 | |||
| No schooling or less than high school (no diploma or GED) | 126 (5.2) | 48 (3.6) | 78 (7.2) | |
| Completed high school | 268 (11.1) | 145 (10.8) | 123 (11.3) | |
| Some post–high school education, no certificate or degree | 420 (17.3) | 236 (17.7) | 184 (16.9) | |
| Some post–high school education, Associate’s degree | 379 (15.6) | 201 (15.0) | 178 (16.4) | |
| Bachelor’s degree | 644 (26.6) | 377 (28.2) | 267 (24.6) | |
| Graduate or professional degree | 574 (23.7) | 324 (24.2) | 250 (23.0) | |
| Unknown or not reported | 12 (0.5) | 6 (0.4) | 6 (0.6) | |
| Annual household income ($) | <0.01 | |||
| <35,000 | 373 (15.4) | 183 (13.7) | 190 (17.5) | |
| 36,000–50,000 | 353 (14.6) | 176 (13.2) | 177 (16.3) | |
| 51,000–75,000 | 405 (16.7) | 199 (14.9) | 206 (19.0) | |
| 75,000 or more | 909 (37.5) | 586 (43.8) | 323 (29.7) | |
| Unknown or not reported | 383 (15.8) | 193 (14.4) | 190 (17.5) | |
| Health history, n (%) | ||||
| Smoking | <0.01 | |||
| Never | 1,410 (58.7) | 711 (53.6) | 699 (64.9) | |
| Former | 838 (34.9) | 513 (38.7) | 325 (30.2) | |
| Current | 155 (6.5) | 102 (7.7) | 53 (4.9) | |
| Unknown or not reported | 20 (0.8) | 11 (0.8) | 9 (0.8) | |
| Medical conditions, n (%) | ||||
| Hypercholesterolemia | 1,344 (55.5) | 821 (61.4) | 523 (48.2) | <0.01 |
| Cancer2 | 262 (10.8) | 145 (10.8) | 117 (10.8) | 0.95 |
| Cardiovascular disease3 | 299 (12.3) | 198 (14.8) | 101 (9.3) | <0.01 |
| Hypertension | 1,296 (53.5) | 758 (56.7) | 538 (49.5) | <0.01 |
| Asthma | 204 (8.4) | 72 (5.4) | 132 (12.2) | <0.01 |
| Chronic obstructive pulmonary disease | 35 (1.4) | 18 (1.3) | 17 (1.6) | 0.65 |
| Sleep apnea | 299 (12.3) | 203 (15.2) | 96 (8.8) | <0.01 |
| Osteoarthritis or degenerative joint disease | 534 (22.0) | 242 (18.1) | 292 (26.9) | <0.01 |
| Osteoporosis or osteopenia | 78 (3.2) | 7 (0.5) | 71 (6.5) | <0.01 |
| Medication use, n (%) | ||||
| Hypercholesterolemia | 1,065 (44.0) | 681 (50.9) | 384 (35.4) | <0.01 |
| Hypertension | 1,263 (52.1) | 738 (55.2) | 525 (48.3) | <0.01 |
| Osteoporosis | 13 (0.5) | 0 (0.0) | 13 (1.2) | <0.01 |
| Dietary supplements4 | ||||
| Vitamin D | ||||
| Participants taking vitamin D, n (%) | 1,037 (42.8) | 506 (37.8) | 531 (48.9) | <0.01 |
| Vitamin D intake among all participants, IU/day, mean (SD) | 313 (399) | 265 (371) | 372 (423) | <0.01 |
| Vitamin D intake among participants using supplements, IU/day, mean (SD) | 732 (255) | 701 (242) | 761 (264) | <0.01 |
| Calcium | ||||
| Participants taking calcium, n (%) | 804 (33.2) | 393 (29.4) | 411 (37.8) | <0.01 |
| Calcium intake among all participants, mg/day, mean (SD) | 103 (176) | 70 (126) | 145 (215) | <0.01 |
| Calcium intake among participants using supplements, mg/day, mean (SD) | 312 (167) | 238 (117) | 382 (177) | <0.01 |
| Clinical testing, mean (SD) | ||||
| Systolic blood pressure, mmHg | 128.4 (13.4) | 129.9 (12.8) | 126.5 (13.9) | <0.01 |
| Diastolic blood pressure, mmHg | 77.0 (9.3) | 78.7 (8.9) | 74.9 (9.3) | <0.01 |
| Serum creatinine, mg/dL | 0.89 (0.19) | 0.99 (0.17) | 0.77 (0.14) | <0.01 |
| eGFR5, mL/min/1.73 m2 | 87.1 (15.7) | 85.8 (15.2) | 88.6 (16.2) | <0.01 |
| Serum calcium, mg/dL | 9.41 (0.37) | 9.38 (0.36) | 9.44 (0.38) | <0.01 |
| Urine calcium-to-creatinine ratio | 0.09 (0.06) | 0.08 (0.05) | 0.09 (0.06) | <0.01 |
| Glycemic testing, mean (SD) | ||||
| FPG, mg/dL | 107.9 (7.4) | 108.8 (7.4) | 106.9 (7.3) | <0.01 |
| 2hPG, mg/dL | 137.2 (34.3) | 135.3 (35.9) | 139.7 (32.1) | <0.01 |
| HbA1c, % | 5.9 (0.2) | 5.9 (0.2) | 5.9 (0.2) | 0.23 |
| Prediabetes categories, n (%) | <0.01 | |||
| iA1c + IFG | 1,184 (48.9) | 687 (51.4) | 497 (45.8) | |
| IFG + IGT | 152 (6.3) | 83 (6.2) | 69 (6.4) | |
| IGT + iA1c | 231 (9.5) | 103 (7.7) | 128 (11.8) | |
| iA1c + IFG + IGT | 856 (35.3) | 464 (34.7) | 392 (36.1) |
Health history and medication use were assessed by self-report. Self-reported gestational diabetes mellitus is among all women participants whether they were pregnant or not. Racial and ethnic categories follow NIH guidelines. “Hispanic” refers to ethnicity and includes any race. IFG defined as FPG 100–125 mg/dL (5.6–6.9 mmol/L). IGT defined as 2hPG glucose after a 75-g glucose load 140–199 mg/dL (7.8–11.0 mmol/L). iA1c defined as HbA1c 5.7–6.4% (39–47 mmol/mol).
1“Other” includes American Indian or Alaska Native, Native Hawaiian or other Pacific Islander, or other race.
2Cancer (except for basal cell skin cancer) within 5 years of randomization was an exclusion criterion. Prostate cancer or well-differentiated thyroid cancer not expected to require treatment over the next 4 years were not exclusions. Volunteers with history of squamous cell cancer of the skin, which was completely excised and with no evidence of metastases, were eligible.
3Cardiovascular disease included arrhythmias, chest pain, congestive heart failure, aortic or coronary artery disease, coronary artery bypass graft/percutaneous coronary intervention, myocardial infarction, palpitations, peripheral vascular disease.
4Data are derived from a direct question about medications and supplements—not from a food-frequency questionnaire.
5eGFR was calculated using the Chronic Kidney Disease Epidemiology Collaboration equation.
Demographics by Sex
Women comprised 44.8% of the cohort (Table 1). Compared with men, women were younger and had higher BMI and lower waist circumference. Women and men also differed in racial and ethnic categories, employment status, education, and annual household income. Compared with men, women were less likely to report a history of hypercholesterolemia, cardiovascular disease, hypertension, sleep apnea, and medication use for these conditions. Women were also less likely to report a history of smoking. Women were more likely than men to report a family history of diabetes (68.1 vs. 57.9%; P < 0.01) or personal history of asthma, osteoarthritis, or osteoporosis/osteopenia. At baseline, compared with men, a higher proportion of women reported taking vitamin D (48.9% for women vs. 37.8% for men; P < 0.01) and calcium (37.8% for women vs. 29.4% for men; P < 0.01) supplements. Women had lower baseline blood pressure and higher eGFR, serum, and urine calcium concentration.
Glycemic Profile Overall
Mean baseline HbA1c was 5.9% (41 mmol/mol), FPG 107.9 mg/dL, and 2hPG 137.2 mg/dL (Table 1). The baseline glycemic profile of the D2d cohort covers a wide spectrum of the prediabetes criteria (Table 1 and Supplementary Fig. 2). Overall, 35.3% of participants met all three prediabetes criteria (FPG, 2hPG, and HbA1c), 48.9% met the FPG/HbA1c criteria only, 9.5% met the 2hPG/FPG criteria only, and 6.3% met the 2hPG/HbA1c criteria only. Overall, 84% of participants met both FPG and HbA1c criteria, while 51.1% met the 2hPG criterion, which has been used as an inclusion criterion in many previous diabetes prevention trials.
Glycemic Profile by Sex
At baseline, mean FPG was lower among women than men (106.9 vs. 108.8 mg/dL respectively; P < 0.01; Table 1), while mean 2hPG was higher in women than men (139.7 vs. 135.3 mg/dL respectively; P < 0.01); HbA1c concentrations did not differ (5.9 vs. 5.9% [41 vs. 41 mmol/mol]; P = 0.23). The proportions of women who met the four different combinations of prediabetes (iA1c/IFG, IFG/IGT, IGT/iA1c or iA1c/IFG/IGT) differed compared with men.
Glycemic Profile by Race and Ethnicity
Diabetes risk factors and glycemic profile differed by races and ethnicity (Table 2). White participants were older than participants of other races, while participants of Hispanic ethnicity were younger than non-Hispanic participants. Asian participants had lower BMI than other racial groups, while blacks had higher baseline BMI than whites. Waist circumference was generally lower in Asian participants. BMI did not differ by ethnicity, but Hispanic men had a lower waist circumference than non-Hispanics. White participants were less likely to report a family history of diabetes than black participants (59.5 vs. 68.1%, respectively; P < 0.05). Fewer non-Hispanic participants reported a family history of diabetes compared with Hispanics (61.8 vs. 68.9%, respectively; P < 0.05). Gestational diabetes mellitus was reported more commonly among white than black women (9.5 vs. 2.0%, respectively; P < 0.05) and among Hispanic than non-Hispanic women (16.2 vs. 6.4%, respectively; P < 0.01). Other racial differences were not statistically significant. D2d does not have data on the number of pregnancies per woman.
Table 2.
Glycemic profile by race and ethnicity in D2d
| Race |
P for race | Ethnicity |
P for ethnicity | |||||
|---|---|---|---|---|---|---|---|---|
| White (N = 1,616) | Black (N = 616) | Asian (N = 130) | Other (N = 61) | Hispanic (N = 225) | Non-Hispanic (N = 2,198) | |||
| Diabetes risk factors | ||||||||
| Age, years, mean (SD) | 61.7 (9.5)1,2,3 | 56.8 (9.5)1 | 55.2 (11.1)2 | 54.7 (9.9)3 | <0.01 | 54.7 (9.8) | 60.5 (9.8) | <0.01 |
| BMI, kg/m2, mean (SD) | 32.1 (4.4)1,2 | 32.7 (4.5)1,4 | 28.2 (3.8)2,4,6 | 31.4 (4.3)6 | <0.01 | 32.3 (4.5) | 32.0 (4.5) | 0.47 |
| Waist circumference, cm, mean (SD) | ||||||||
| Men | 108.8 (11.5)1,2,3 | 105.3 (10.6)1,4 | 97.3 (11.8)2,4 | 102.4 (9.1)3 | <0.01 | 104.3 (11.1) | 107.3 (11.6) | 0.02 |
| Women | 103.3 (11.2)1 | 102.1 (10.9)1,4 | 93.7 (10.6)4 | 100.4 (14.0) | <0.01 | 101.1 (11.6) | 102.7 (11.3) | 0.14 |
| Self-reported family history of diabetes, n (%) | 962 (59.5)1 | 420 (68.1)1 | 86 (66.2) | 46 (75.4) | <0.01 | 155 (68.9) | 1,359 (61.8) | 0.04 |
| Self-reported gestational diabetes mellitus, n (%) women | 68 (9.5)1 | 6 (2.0)1 | 5 (10.9) | 4 (14.3) | <0.01 | 22 (16.2) | 61 (6.4) | <0.01 |
| Glycemic testing, mean (SD) | ||||||||
| FPG, mg/dL | 108.8 (7.3)1 | 105.8 (7.4)1 | 107.3 (7.1) | 107.3 (6.7) | <0.01 | 107.4 (7.3) | 108 (7.4) | 0.24 |
| 2hPG, mg/dL | 137.7 (35.0) | 135.7 (32.5) | 140.5 (33.7) | 135.3 (35.4) | 0.41 | 135.2 (32.5) | 137.5 (34.5) | 0.36 |
| HbA1c, % | 5.9 (0.2)1 | 6.0 (0.2)1,4,5 | 5.9 (0.2)4 | 5.9 (0.2)5 | <0.01 | 5.9 (0.2) | 5.9 (0.2) | 0.56 |
| Prediabetes categories, n (%) | <0.01 | 0.92 | ||||||
| iA1c + IFG | 768 (47.5)1 | 323 (52.4)1 | 64 (49.2) | 29 (47.5) | 112 (49.8) | 1,072 (48.8) | ||
| IFG + IGT | 126 (7.8)1 | 17 (2.8)1 | 6 (4.6) | 3 (4.9) | 16 (7.1) | 136 (6.2) | ||
| IGT + iA1c | 119 (7.4)1 | 91 (14.8)1 | 14 (10.8) | 7 (11.5) | 20 (8.9) | 211 (9.6) | ||
| iA1c + IFG + IGT | 603 (37.3)1 | 185 (30.3)1 | 46 (35.4) | 22 (36.1) | 77 (34.2) | 779 (35.4) | ||
The Tukey-Kramer test was used for post hoc pairwise comparisons between racial groups. Self-reported gestational diabetes mellitus is among all women participants whether they were pregnant or not. Racial and ethnic categories follow NIH guidelines. Hispanic refers to ethnicity and includes any race. IFG defined as FPG 100–125 mg/dL (5.6–6.9 mmol/L). IGT defined as 2hPG after a 75-g glucose load 140–199 mg/dL (7.8–11.0 mmol/L). iA1c defined as HbA1c 5.7–6.4% (39–47 mmol/mol).
1Difference between white and black significant at P < 0.05.
2Difference between white and Asian significant at P < 0.05.
3Difference between white and other significant at P < 0.05.
4Difference between black and Asian significant at P < 0.05.
5Difference between black and other significant at P < 0.05.
6Difference between Asian and other significant at P < 0.05.
Black participants had a higher mean HbA1c concentration than other races (6.0 vs. 5.9% [42 vs. 41 mmol/mol]; P < 0.05) despite having a lower mean FPG (105.8 vs. 108.8 mg/dL; P < 0.05) and a similar mean 2hPG concentration (135.7 vs. 137.7 mg/dL; P = 0.19). Glycemic concentrations did not differ between Hispanics vs. non-Hispanics. Compared with white participants, a higher percentage of black participants met both FPG and HbA1c criteria (52.4 for blacks vs. 47.5% for whites; P < 0.05) and 2hPG and HbA1c criteria (14.8 vs. 7.4%; P < 0.05), but a lower percentage met FPG and 2hPG criteria (2.8 vs. 7.8%; P < 0.05) or all three criteria (30.3 vs. 37.3%; P < 0.05).
Comparison With Other Prediabetes Trials
Glycemic eligibility criteria and the baseline characteristics of D2d are compared with other large diabetes prevention trials in Table 3 (7–12). The trials differ in several baseline characteristics (e.g., age, BMI) because they targeted different populations. D2d and the recently completed SCALE Obesity and Prediabetes trial (SCALE), which tested liraglutide versus placebo for diabetes prevention (11), are the only two large trials that used the 2010 ADA glycemic criteria to define a prediabetes cohort. SCALE required that participants meet at least one of the three ADA glycemic criteria (compared with D2d, which required at least two of the three criteria). The mean baseline 2hPG concentrations in D2d and SCALE were lower than in the older trials, which included the 2hPG as an inclusion criterion. Mean HbA1c and FPG concentrations were comparable among all prediabetes trials.
Table 3.
Key baseline characteristics of D2d versus other large prediabetes trials
| D2d | SCALE | NAVIGATOR | DREAM | DPP | STOP-NIDDM | ACE | |
|---|---|---|---|---|---|---|---|
| N | 2,423 | 2,254 | 9,518 | 5,269 | 3,234 | 1,429 | 6,522 |
| Years conducted (country) | 2013–2018 (estimated) (U.S.) | 2011–2015 (multiple countries worldwide)1 | 2002–2007 (multiple countries worldwide)1 | 2001–2006 (multiple countries worldwide) | 1996–2001 (U.S.)1 | 1995–2000 (multiple countries worldwide)1 | 2009–2015 (China) |
| Study design and intervention | Two arms: vitamin D3 vs. placebo | Two arms: liraglutide vs. placebo (2:1 ratio) | 2 × 2 factorial design: nateglinide and/or valsartan vs. placebos | 2 × 2 factorial design: ramipril and/or rosiglitazone vs. placebos | Three arms: metformin vs. intensive lifestyle vs. placebo | Two arms: acarbose vs. placebo | Two arms: acarbose vs. placebo |
| Glycemic criteria | At least two of three 2010 ADA criteria for prediabetes: 2hPG 140–199 mg/dL, FPG 100–125 mg/dL, HbA1c 5.7–6.4% | At least one of three 2010 ADA criteria for prediabetes: 2hPG 140–199 mg/dL, FPG 100–125 mg/dL, HbA1c 5.7–6.4% | 2hPG 140–199 mg/dL and FPG 95–125 mg/dL | 2hPG 140–199 mg/dL or FPG 110–125 mg/dL | 2hPG 140–199 mg/dL and FPG 95–125 mg/dL | 2hPG 140–199 mg/dL and FPG 101–139 mg/dL | 2hPG 140–199 mg/dL |
| Age, years, mean | 60.0 | 47.4 | 63.8 | 54.7 | 50.6 | 54.5 | 64.3 |
| Women, % | 45 | 76 | 51 | 59 | 68 | 51 | 27 |
| BMI, kg/m2, mean | 32.1 | 38.9 | 30.5 | 30.9 | 34.0 | 31 | 25.4 |
| Waist circumference, cm, mean | 105.0 | 116.6 | 101 | Not reported | 105.1 | 101 | 91.2 |
| White race, % | 66.7 | 83 | 83.1 | Not reported | 552 | 98 | 0 |
| Family history of diabetes, % | 62.5 | Not reported | 38 | Not reported | 69 | Not reported | Not reported |
| FPG, mg/dL, mean | 107.9 | 99.0 | 109.6 | 104.4 | 106.5 | 112.0 | 99.0 |
| 2hPG, mg/dL, mean | 137.25 | 133.2 | 164.9 | 156.6 | 164.6 | 166.7 | 167.4 |
| HbA1c, %, mean | 5.9 | 5.8 | 5.8 | Not reported | 5.9 | Not reported | 5.9 |
| Duration of follow-up, years, mean | 3 (estimated) | 3.0 | 5.0 | 3.0 | 2.8 | 3.3 | 5.0 |
| Cumulative incidence of diabetes in placebo group during mean follow-up, % | Not available | 11 | 34 | 19 | 29 | 42 | 15 |
Trials selected for having >1,000 participants and at least 1 year of follow-up. DREAM, Diabetes Reduction Assessment with Ramipril and Rosiglitazone Medication; NAVIGATOR, Nateglinide and Valsartan in Impaired Glucose Tolerance Outcomes Research; STOP-NIDDM, Study to Prevent Non-Insulin-Dependent Diabetes Mellitus.
1Study duration was estimated based on published information about the end of recruitment and follow-up time.
2Non-Hispanic white.
Comparison With Other Trials of Vitamin D Supplementation
There are two other large trials specifically designed to test the hypothesis that vitamin D supplementation reduces the risk of diabetes among patients at risk for diabetes (Table 4) (18,19). D2d is significantly larger (2,423 participants vs. 511 and 750) and uses a different vitamin D dosing regimen (daily D3 vs. weekly D3 vs. active vitamin D). The Tromsø Study (Norway) randomized 511 white adults with prediabetes to 20,000 IU/week (∼2,900 IU/day) of vitamin D3 or placebo and followed them for incident diabetes for an average of 3.3 years (19). Risk of developing diabetes was lower in the vitamin D versus placebo group throughout the study, but the difference was not statistically significant (hazard ratio 0.90 [95% CI 0.69–1.18]). The Diabetes Prevention with active Vitamin D (DPVD) trial has not reported findings despite concluding in 2013 (18). Three ongoing trials will explore the effect of vitamin D supplementation on diabetes risk or fasting glucose and insulin concentrations as secondary outcomes (Table 4). Two are very large (>20,000 participants) community-based trials with primary outcomes of cancer, cardiovascular disease, and mortality (20,21); the third is an efficacy trial with primary outcomes of nonvertebral fracture, functional and cognitive decline, blood pressure, and infection (22). In secondary analyses, D2d will assess the effect of vitamin D supplementation on indices of insulin secretion and insulin sensitivity after a 75-g oral glucose tolerance test.
Table 4.
Ongoing and completed randomized, placebo-controlled trials of vitamin D supplementation and diabetes risk
| D2d | Tromsø study | DPVD | VITAL | D-Health | DO-HEALTH | |
|---|---|---|---|---|---|---|
| N | 2,423 | 511 | 750 (target) | 25,874 | 21,315 | 2,157 |
| Years conducted (country) | 2013–2018 (estimated) (U.S.) | 2008–2015 (Norway) | 2013–unknown (estimated) (Japan) | 2010–2018 (estimated) (U.S.) | 2014–2025 (estimated) (Australia) | 2012–2017 (estimated) (Europe) |
| Diabetes outcome | Primary | Primary | Primary | Secondary1 | Secondary2 | Secondary (fasting glucose, insulin)3 |
| Glycemic inclusion criteria | At least two of three ADA criteria for prediabetes: 2hPG 140–199 mg/dL, FPG 100–125 mg/dL, HbA1c 5.7–6.4% | 2hPG 140–198 mg/dL and/or FPG 108–124 mg/dL4 | 2hPG 140–199 mg/dL and fasting glucose <126 mg/dL and HbA1c <6.5% | None | None | None |
| Active intervention | Two arms: 4,000 IU vitamin D3 daily vs. placebo | Two arms: 20,000 IU vitamin D3 weekly (∼2,900 daily) vs. placebo | Two arms: 0.75 μg eldecalcitol (1,25[OH]2D3) daily vs. placebo | 2 × 2 factorial design: 2,000 IU D3 daily, 1 g daily marine n-3 fatty acid vs. placebos | Two arms: 60,000 IU D3 monthly vs. placebo | 2 × 2 × 2 factorial design: 2,000 IU D3 daily, 1 g marine n-3 fatty acid daily, exercise program |
| Personal use of vitamin D from supplements, % of participants (maximum amount allowed) | 43 (1,000 IU/day) | 35 (<400 IU/day) | Not available | Not available (800 IU/day) | Not available (500 IU/day [2,000 IU/day if prescribed]) | Not available (800 IU/day) |
| Treatment duration, years | ∼3 (estimated) | 5 | 2.8 | 5 | 5 | 3 |
DO-HEALTH, Vitamin D3 – Omega3 – Home Exercise – HeALTHy Ageing and Longevity Trial; VITAL, VITamin D and OmegA-3.
1Primary outcomes: cancer, major adverse cardiovascular events.
2Primary outcome: all-cause mortality.
3Primary outcomes: nonvertebral fracture, functional decline, blood pressure, cognitive decline, infection.
4HbA1c added as inclusion requirement midway through recruitment; all participants had HbA1c between 5.8 and 6.9%.
Conclusions
D2d is a large randomized clinical trial testing the hypothesis that oral daily vitamin D3 lowers risk of diabetes in U.S. adults with prediabetes. The study’s large size, with recruitment from 22 sites across the U.S., ensures that the D2d cohort includes people with a wide spectrum of diabetes risk, appropriate for testing the underlying hypothesis, while the placebo group will provide information on the natural history of prediabetes in the current era.
When designing D2d, we used contemporary definitions to assemble a prediabetes cohort that follows the latest guidelines, reflects current diagnosis patterns, and identifies those at the highest risk of progression to diabetes. As such, D2d has assembled the largest cohort of U.S. adults with prediabetes based on the revised ADA criteria, which include HbA1c and lower FPG thresholds. The contemporary ADA criteria have been controversial, as they increase the prevalence of prediabetes by identifying people at lower risk for diabetes and cardiovascular complications, who may be less likely to benefit from interventions compared with people meeting earlier criteria (13,14). D2d’s inclusion criteria resulted in a cohort with lower mean 2hPG concentrations but similar HbA1c and FPG concentrations compared with previous cohorts. However, this hyperglycemia pattern likely matches more closely that of contemporary patients with prediabetes, since 8 of 10 D2d participants met both FPG and HbA1c criteria, which are the tests most commonly used in clinical practice to diagnose prediabetes. The 2hPG criterion is used less often in the clinical setting, likely because of its patient and provider burden as well as low reproducibility (23).
We currently have little knowledge of how the 2010 ADA criteria are related to future diabetes risk in the general U.S. population (24,25). The control arm of other large trials, such as the Diabetes Prevention Program (DPP) or Acarbose Cardiovascular Evaluation (ACE) (7,12), cannot answer this question because they required the 2hPG criterion, which is rarely available in clinical practice. Reanalyses of data from older cohorts applying the new HbA1c and FPG criteria retrospectively are limited because these criteria were not used to build the cohorts (26). In addition, many trials were conducted decades ago and changes in the background milieu (e.g., lifestyle changes, increased use of mobile technology, slowing in the growth of overweight and obesity) (27,28) make previous cohorts less representative of the current U.S. population at risk for diabetes. Because D2d used the current ADA criteria to identify people at risk for diabetes and will take into consideration all contemporary factors and influences, the study will help establish the natural history of prediabetes among U.S. adults followed in a clinical trial in the contemporary era; such information is important to make informed decisions about diabetes risk at both the personal and public health level. Although SCALE (liraglutide vs. placebo), conducted in 27 countries worldwide, also identified people with prediabetes using the 2010 ADA criteria, it only required that participants meet one of three glycemic criteria for prediabetes and had 50% loss to follow-up (11), limiting the ability to interpret the natural history of this lower-risk prediabetes cohort.
The D2d cohort is well balanced by sex, which makes possible the evaluation of vitamin D’s effect by sex given several sex-based differences in characteristics. Women participants are younger, use more vitamin D and calcium supplements, and have different health histories and baseline clinical testing results than men. Women also have lower FPG concentrations and higher 2hPG concentrations than men. This pattern has previously been noted (29,30), and it is not clear whether this is due to the same glucose load (75 g) being given to smaller individuals or to actual sex differences in glucose metabolism (31).
The D2d cohort is racially diverse, which allows testing for effect modification by race, especially among black individuals who have both higher diabetes risk and different vitamin D homeostasis (32,33). In the D2d cohort, key diabetes risk factors including age, BMI, waist circumference, family history of diabetes, history of gestational diabetes mellitus, and glycemic concentrations vary between racial and ethnic groups. Notably, black participants have a higher baseline mean HbA1c concentration than other races, despite having lower FPG and similar 2hPG concentrations. This finding, which is increasingly recognized, suggests that HbA1c overestimates mean glycemia in black compared with white individuals (16). Indeed, the Department of Veterans Affairs and Department of Defense 2017 guidelines recommend that HbA1c 6.5–6.9% (48–52 mmol/mol) alone not be used to diagnose diabetes in the absence of a confirmatory FPG measurement (34). D2d will provide valuable data in this controversial area.
Given the wide range of glycemic phenotypes within the current ADA definition of prediabetes (e.g., iA1c, IFG, or IGT alone or iA1c, IFG, and IGT), it is likely that the definition of prediabetes will continue to evolve. Beyond glycemic criteria, there are several distinct clinical phenotypes that may be important to consider when classifying future risk of diabetes and diabetes-specific complications (35). Given the large size of the well-characterized D2d cohort and long-term follow-up with frequent (twice yearly) evaluations of glycemic status and other clinical outcomes (e.g., cancer, cardiovascular disease), D2d is well positioned to examine how different phenotypes influence future risk, including how glycemia (assessed by FPG and 2hPG) and HbA1c are associated with future diabetes and cardiovascular risk and how risk varies by age, sex, race, and ethnicity.
Serum 25-hydroxyvitamin D values are not currently available. Per the study’s protocol and analytical plan, 25-hydroxyvitamin D will be analyzed at the conclusion of the study in pairs (before/after intervention) and in the same analytical run to reduce systematic error and interassay variability. We expect baseline levels to be similar between the two groups, as factors that impact vitamin D status (e.g., geographical location, racial/ethnicity [36,37]) are balanced between groups. Nevertheless, in prespecified subgroup analyses, heterogeneity of treatment effects by baseline 25-hydroxyvitamin D levels will be assessed (3).
In conclusion, in a contemporary cohort of U.S. adults at risk for diabetes, D2d is expected to address two important knowledge gaps: 1) whether vitamin D supplementation prevents diabetes and 2) how the 2010 expanded ADA criteria for prediabetes impact the natural history of prediabetes. The answers to these questions will have extensive implications for the many U.S. adults at risk for diabetes.
Supplementary Material
Article Information
Acknowledgments. The authors thank the D2d investigators, staff, and trial participants for their outstanding dedication and commitment to the study.
Funding. The planning phase of D2d was funded by the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) through a multicenter clinical study implementation planning grant (U34) to Tufts Medical Center in Boston, MA (U34-DK-091958 to principal investigator A.G.P.). Planning was also supported in part by the Intramural Research Program of the NIDDK. The conduct of D2d is primarily supported by NIDDK and the Office of Dietary Supplements of the NIH through the multicenter clinical study cooperative agreement (U01-DK-098245 to principal investigator A.G.P.) to Tufts Medical Center, where the D2d Coordinating Center is based. The U01 grant mechanism establishes the NIDDK project scientist (M.A.S.) as a member of the D2d Research Group. The study also received secondary funding from the ADA (1-14-D2d-01). Educational materials are provided by the National Diabetes Education Program. The D2d investigators and the NIDDK project scientist were responsible for the design and conduct of the study; collection, management, analysis, and interpretation of data; preparation, review, and approval of the manuscript; and decision to submit the manuscript for publication. Under the terms of the cooperative agreement funding mechanism used by the NIH, representatives from the sponsoring NIDDK participated in the design and conduct of the study; interpretation of data; preparation, review, and approval of the manuscript; and the decision to submit the manuscript for publication.
The sponsor did not have the right or ability to veto submission for publication. Study pills were purchased from an independent nutritional supplement manufacturing company that has no association with any members of the D2d Research Group.
Duality of Interest. No potential conflicts of interest relevant to this article were reported.
Author Contributions. E.S.L., A.G.P., R.E.P., B.D.-H., M.A.S., and P.R.S. contributed to the design, acquisition and interpretation of data, draft, revisions, and final approval of the manuscript. All other authors contributed to acquisition and interpretation of data and critical review and final approval of the manuscript. A.G.P. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Footnotes
Clinical trial reg. no. NCT01942694, clinicaltrials.gov.
This article contains Supplementary Data online at http://care.diabetesjournals.org/lookup/suppl/doi:10.2337/dc18-0240/-/DC1.
A complete list of the members of the D2d Research Group can be found in the Supplementary Data.
Contributor Information
Collaborators: D2d Research Group, Erin S. LeBlanc, Richard E. Pratley, Bess Dawson-Hughes, Myrlene A. Staten, Patricia R. Sheehan, Michael R. Lewis, Anne Peters, Sun H. Kim, Ranee Chatterjee, Vanita R. Aroda, Chhavi Chadha, Lisa M. Neff, Irwin G. Brodsky, Clifford Rosen, Cyrus V. Desouza, John P. Foreyt, Daniel S. Hsia, Karen C. Johnson, Philip Raskin, Sangeeta R. Kashyap, Patrick O’Neil, Lawrence S. Phillips, Neda Rasouli, Emilia P. Liao, David C. Robbins, and Anastassios G. Pittas
References
- 1.Huang Y, Cai X, Mai W, Li M, Hu Y. Association between prediabetes and risk of cardiovascular disease and all cause mortality: systematic review and meta-analysis. BMJ 2016;355:i5953. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Balk EM, Earley A, Raman G, Avendano EA, Pittas AG, Remington PL. Combined diet and physical activity promotion programs to prevent type 2 diabetes among persons at increased risk: a systematic review for the Community Preventive Services Task Force. Ann Intern Med 2015;163:437–451 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Pittas AG, Dawson-Hughes B, Sheehan PR, et al.; D2d Research Group . Rationale and design of the Vitamin D and Type 2 Diabetes (D2d) study: a diabetes prevention trial. Diabetes Care 2014;37:3227–3234 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.American Diabetes Association Standards of medical care in diabetes--2010 [published correction appears in Diabetes Care 2010;33:692]. Diabetes Care 2010;33(Suppl. 1):S11–S61 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Yudkin JS. “Prediabetes”: are there problems with this label? Yes, the label creates further problems! Diabetes Care 2016;39:1468–1471 [DOI] [PubMed] [Google Scholar]
- 6.Cefalu WT. “Prediabetes”: are there problems with this label? No, we need heightened awareness of this condition! Diabetes Care 2016;39:1472–1477 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.The Diabetes Prevention Program Research Group The Diabetes Prevention Program: baseline characteristics of the randomized cohort. Diabetes Care 2000;23:1619–1629 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Holman RR, Haffner SM, McMurray JJ, et al.; NAVIGATOR Study Group . Effect of nateglinide on the incidence of diabetes and cardiovascular events. N Engl J Med 2010;362:1463–1476 [DOI] [PubMed] [Google Scholar]
- 9.Gerstein HC, Yusuf S, Bosch J, et al.; DREAM (Diabetes REduction Assessment with ramipril and rosiglitazone Medication) Trial Investigators . Effect of rosiglitazone on the frequency of diabetes in patients with impaired glucose tolerance or impaired fasting glucose: a randomised controlled trial. Lancet 2006;368:1096–1105 [DOI] [PubMed] [Google Scholar]
- 10.Chiasson JL, Josse RG, Gomis R, Hanefeld M, Karasik A, Laakso M; STOP-NIDDM Trial Research Group . Acarbose for prevention of type 2 diabetes mellitus: the STOP-NIDDM randomised trial. Lancet 2002;359:2072–2077 [DOI] [PubMed] [Google Scholar]
- 11.le Roux CW, Astrup A, Fujioka K, et al.; SCALE Obesity Prediabetes NN8022-1839 Study Group . 3 years of liraglutide versus placebo for type 2 diabetes risk reduction and weight management in individuals with prediabetes: a randomised, double-blind trial. Lancet 2017;389:1399–1409 [DOI] [PubMed] [Google Scholar]
- 12.Holman RR, Coleman RL, Chan JCN, et al.; ACE Study Group . Effects of acarbose on cardiovascular and diabetes outcomes in patients with coronary heart disease and impaired glucose tolerance (ACE): a randomised, double-blind, placebo-controlled trial. Lancet Diabetes Endocrinol 2017;5:877–886 [DOI] [PubMed] [Google Scholar]
- 13.De Caterina R, Madonna R. Impaired fasting plasma glucose and long-term cardiovascular risk: still a foggy relationship. Eur Heart J 2010;31:1159–1162 [DOI] [PubMed] [Google Scholar]
- 14.Shahraz S, Pittas AG, Kent DM. Prediabetes risk in adult Americans according to a risk test. JAMA Intern Med 2016;176:1861–1863 [DOI] [PubMed] [Google Scholar]
- 15.Bergenstal RM, Gal RL, Connor CG, et al.; T1D Exchange Racial Differences Study Group . Racial differences in the relationship of glucose concentrations and hemoglobin A1c levels. Ann Intern Med 2017;167:95–102 [DOI] [PubMed] [Google Scholar]
- 16.The DCCT Research Group Feasibility of centralized measurements of glycated hemoglobin in the Diabetes Control and Complications Trial: a multicenter study. Clin Chem 1987;33:2267–2271 [PubMed] [Google Scholar]
- 17.Levey AS, Stevens LA, Schmid CH, et al.; CKD-EPI (Chronic Kidney Disease Epidemiology Collaboration) . A new equation to estimate glomerular filtration rate. Ann Intern Med 2009;150:604–612 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Kawahara T, Suzuki G, Inazu T, et al. . Rationale and design of Diabetes Prevention with active Vitamin D (DPVD): a randomised, double-blind, placebo-controlled study. BMJ Open 2016;6:e011183. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Jorde R, Sollid ST, Svartberg J, et al. . Vitamin D 20,000 IU per week for five years does not prevent progression from prediabetes to diabetes. J Clin Endocrinol Metab 2016;101:1647–1655 [DOI] [PubMed] [Google Scholar]
- 20.Manson JE, Bassuk SS, Lee IM, et al. . The VITamin D and OmegA-3 TriaL (VITAL): rationale and design of a large randomized controlled trial of vitamin D and marine omega-3 fatty acid supplements for the primary prevention of cancer and cardiovascular disease. Contemp Clin Trials 2012;33:159–171 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Neale RE, Armstrong BK, Baxter C, et al. . The D-Health Trial: a randomized trial of vitamin D for prevention of mortality and cancer. Contemp Clin Trials 2016;48:83–90 [DOI] [PubMed] [Google Scholar]
- 22.DO HEALTH Home page [Internet], 2017. Available from http://do-health.eu/wordpress/. Accessed 2 April 2018.
- 23.Ko GT, Chan JC, Woo J, et al. . The reproducibility and usefulness of the oral glucose tolerance test in screening for diabetes and other cardiovascular risk factors. Ann Clin Biochem 1998;35:62–67 [DOI] [PubMed] [Google Scholar]
- 24.Chamberlain JJ, Rhinehart AS, Shaefer CF Jr., Neuman A. Diagnosis and management of diabetes: synopsis of the 2016 American Diabetes Association Standards of Medical Care in Diabetes. Ann Intern Med 2016;164:542–552 [DOI] [PubMed] [Google Scholar]
- 25.Vijayakumar P, Nelson RG, Hanson RL, Knowler WC, Sinha M. HbA1c and the prediction of type 2 diabetes in children and adults. Diabetes Care 2017;40:16–21 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Christophi CA, Resnick HE, Ratner RE, et al.; Diabetes Prevention Program Research Group . Confirming glycemic status in the Diabetes Prevention Program: implications for diagnosing diabetes in high risk adults. J Diabetes Complications 2013;27:150–157 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Ford ES, Dietz WH. Trends in energy intake among adults in the United States: findings from NHANES. Am J Clin Nutr 2013;97:848–853 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Ogden CL, Carroll MD, Kit BK, Flegal KM. Prevalence of obesity among adults: United States, 2011-2012. NCHS Data Brief 2013;Oct:1–8 [PubMed] [Google Scholar]
- 29.Williams JW, Zimmet PZ, Shaw JE, et al. . Gender differences in the prevalence of impaired fasting glycaemia and impaired glucose tolerance in Mauritius. Does sex matter? Diabet Med 2003;20:915–920 [DOI] [PubMed] [Google Scholar]
- 30.King H, Rewers M; WHO Ad Hoc Diabetes Reporting Group . Global estimates for prevalence of diabetes mellitus and impaired glucose tolerance in adults. Diabetes Care 1993;16:157–177 [DOI] [PubMed] [Google Scholar]
- 31.Janghorbani M, Amini M. Effects of gender and height on the oral glucose tolerance test: the isfahan diabetes prevention study. Rev Diabet Stud 2008;5:163–170 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Institute of Medicine of the National Acadamies. Dietary Reference Intakes for Calcium and Vitamin D. Washington, DC, National Academies Press, 2011 [PubMed]
- 33.Scragg R, Sowers M, Bell C; Third National Health and Nutrition Examination Survey . Serum 25-hydroxyvitamin D, diabetes, and ethnicity in the Third National Health and Nutrition Examination Survey. Diabetes Care 2004;27:2813–2818 [DOI] [PubMed] [Google Scholar]
- 34.Conlin PR, Colburn J, Aron D, Pries RM, Tschanz MP, Pogach L. Synopsis of the 2017 U.S. Department of Veterans Affairs/U.S. Department of Defense clinical practice guideline: management of type 2 diabetes mellitus. Ann Intern Med 2017;167:655–663 [DOI] [PubMed] [Google Scholar]
- 35.Vas PRJ, Alberti KG, Edmonds ME. Prediabetes: moving away from a glucocentric definition. Lancet Diabetes Endocrinol 2017;5:848–849 [DOI] [PubMed] [Google Scholar]
- 36.Harris SS. Vitamin D and African Americans. J Nutr 2006;136:1126–1129 [DOI] [PubMed] [Google Scholar]
- 37.Jain RB. Variability in the levels of vitamin D by age, gender, and race/ethnicity: data from National Health and Nutrition Examination Survey 2007-2010. J Nutr Health Sci 2016;3:203 [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.