Table 3.
Observational longitudinal cohort studies of vitamin D status (plasma or serum 25[OH]D concentration, predicted 25[OH]D concentration or self-reported vitamin D intake) and incident type 2 diabetes
| Study, Year (reference) Cohort [Country] | Male, % | Mean baseline age (range), y | White, % | n*/N (incidence) | Vitamin D measure; comparison† | Mean follow-up, y (start-end) | Results, Adjusted RR, OR, or HR (95% CI) P for trend | Outcome (ascertainment method) | Adjustments |
|---|---|---|---|---|---|---|---|---|---|
| Liu et al, 2005 Women’s Health Study [US] | 0 | 52 (45–75) | 95 | 805/10,066 (8.0%) | Vitamin D intake (total); ≥511 vs. ≤159 IU/d | 9 (ND) | 0.73 (0.54, 0.99) § P=0.02 |
Type 2 diabetes (validated self-report) | Age |
| Pittas et al, 2006 [US] | 0 | 46 (30–55) | 98 | 4,843/83,779 (5.8%) | Vitamin D intake (total); >800 vs. ≤200 IU/d | 20 (1980–2000) | 0.87 (0.69, 1.09) P=0.67 |
Type 2 diabetes (validated self-report) | Age, BMI, exercise, residence, family history of diabetes, hypertension. calcium intake, smoking, alcohol, coffee, other diet |
| Knekt et al, 2008 Finnish Mobile Clinic Health Examination Survey [Finland] | 100 | ND (40–74) | 100 | 105/1,628 (6.4%); nested case-control study with 206 control participants | 25OHD concentration; 30 vs. 10 ng/mL (means) | 22 (1973–1994) | 0.49 (0.15, 1.64) P=0.06 |
Type 2 diabetes (medication-treated, registry-based) | Age, BMI, exercise, season, smoking, education, medications |
| 0 | ND (40–74) | 100 | 125/1,699 (7.4%); nested case-control study with 246 control participants | 25OHD concentration; 25 vs. 9 ng/mL (means) | 0.91 (0.37, 2.23) P=0.66 |
Age, BMI, exercise, season, smoking, education, medications | |||
| Knekt et al, 2008 Mini-Finland Health Survey [Finland] | 100 | 53 (40–69) | 100 | 83/1,948 (4.3%); nested case-control study with 245 control participants | 25OHD concentration; 31 vs. 9 ng/mL (means) | 17 (1978–1994) | 0.17 (0.05, 0.52) P<0.001 |
Type 2 diabetes (medication-treated, registry-based) | Age, BMI, exercise, season, smoking, education, medications |
| 0 | ND (40–69) | 100 | 99/2228 (4.4%); nested case-control study with 289 control participants | 25OHD concentration; 25 vs. 8 ng/mL (means) | 1.45 (0.58, 3.62) P=0.83 |
Age, BMI, exercise, season, smoking, education, medications | |||
| Kirii et al, 2009 Japan Public Health Center-based Prospective Study [Japan] | 100 | 57 (40–69) | NR (~100% Japanese) | 634/25,877 (2.4%) | Vitamin D intake (total); 720 vs. 188 IU/d (means) | 5 (1990–1998) | 0.96 (0.74, 1.23) P=0.35 |
Type 2 diabetes (validated self-report) | Age, BMI, exercise, family history of diabetes, smoking, diet, hypertension |
| 0 | 57 (40–69) | NR (~100% Japanese) | 480/33,919 (1.4%) | Vitamin D intake (total); 696 vs. 192 IU/d (means) | 5 (1990–1998) | 0.88 (0.67, 1.16) P=0.67 |
Type 2 diabetes (validated self-report) | ||
| Liu et al, 2010 Framingham Offspring Study [US] | 54 | 60 | ~100 | 133/2,956 (4.4%) | Predicted 25OHD score; 22 vs. 17 ng/mL (median) | 7 (1991–2001) | 0.60 (0.37, 0.97) P=0.03 |
Type 2 diabetes (medication-treated, laboratory-based) | Age, sex, waist circumference, ** family history of diabetes, hypertension, low HDL-cholesterol, high triglycerides, impaired fasting glucose, diet |
| Pittas et al, 2010 Nurses Health Study [US] | 0 | 46 (30–55) | 98 | 608/32,826 (1.8%); nested case-control study with 569 control participants | 25OHD concentration; 33 vs. 14 ng/mL (median) | 14 (1990– 2004) | 0.52 (0.33, 0.83) P=0.008 |
Type 2 diabetes (validated self-report) | Age, BMI, exercise, season, race, fasting status, latitude, hypercholesterolemia, hypertension, family history of diabetes, smoking, physical activity, alcohol, multivitamin use, diet |
| Anderson et al, 2010 Intermountain Healthcare system [US] | 25 | 55 | NR | NR/41,497 (NR); | 25OHD concentration; ≤15 vs. >30 ng/mL (median) | 1.3 (2000 –2009) | 1.89 (1.54, 2.33) P<0.001 |
Diabetes (physician-diagnosed based on ICD-9 code) | Age, gender, hypertension, hyperlipidemia, heart failure, infection, depression, renal failure |
| Grimnes et al, 2010 Tromso study [Norway] | NR NR |
60 (Non smokers) 57 (smokers) |
NR (majority Caucasians) NR (majority Caucasians) |
183/4157 64/1962 |
25OHD concentration; quartiles 25OHD concentration; quartiles |
11 (1994–2005) 11 (1194–2005) |
HR=0.95 (0.86–1.0) p NS HR=0.96(0.83–1.12) p NS |
Type 2 diabetes (self-report verified by A1C and hospital discharge diagnosis) Type 2 diabetes (self-report verified by A1C and hospital discharge diagnosis) |
Age, Sex, BMI, physical activity, month (stratified by smoking status) Age, Sex, BMI, physical activity, month (stratified by smoking status) |
| Bolland et al, 2010 Community dwelling women [Australia] | 0 | 74 (>55) | 100 | 15/1,471 | 25OHD concentration; <20 vs. ≥20 ng/mL | 5 (1998–2003) | 0.90 (0.4, 1.9) P=NS |
Type 2 diabetes (self-reported) | Age, weight, smoking, season, treatment allocation |
| Gagnon et al, 2011 [Australia] | 45 | 51 (xx) | 92 | 199/6,537 (3.8%); | 25OHD concentration; ≤19 vs. ≥32 ng/mL | 5 (1999, 2005) | 0.68 (0.43, 1.07) P=0.02 |
Type 2 diabetes (medication-treated, FPG or OGTT) | Age, gender, waist, exercise, race, |
| Robinson et al, 2011 WHI [US] | 0 | (50–79) | 317/5,140 (6.2%); nested case-control study | 25OHD concentration; < 20 vs. ≥30 ng/mL | 7.3 | 1.14 (0.68, 1.90) P=0.873 |
Diabetes (self-report, medication-treated) | Age, BMI, season, race, others | |
| Thorand et al, 2011 [Germany] | 53 | (35–74) | 100% | 416/1683 (25%) case-control | 25OHD concentration; 11 vs. 68 ng/mL (median) | 11 | 0.63 (0.44, 0.90) P=0.01 |
Type 2 diabetes (validated self-report) | Age, sex, survey, season, BMI, smoking, alcohol, physical activity, systolic blood pressure, total cholesterol/HDL cholesterol, parental history of DM. |
| Pittas et al, 2012 Diabetes Prevention Program [USA] | 33 | 51 | 57% | /2040 | 25OHD concentration; ≤13 vs. ≥30 ng/mL | 2.7 yr | 0.72 (0.56, 0.90) p=0.0054 |
Type 2 diabetes (OGTT) | Age, gender body mass index, race, family history of diabetes, personal history of hypertension at baseline, smoking status at baseline, alcohol consumption, C-reactive protein, kidney function, self-reported physical activity, calcium intake and treatment arm |
| Deleskog et al, 2012 [Denmark] | 100 | (35–56) | 100% | 145/1011 | 25OHD concentration; >28 vs. < 18 ng/mL (means) | 10 | 0.38 (0.21, 0.71) | Type 2 diabetes (OGTT) | Age, gender, BMI, exercise, season, BP, family history of diabetes |
| Forouhi et al, 2012 EPIC-Norfolk [Europe] | 52% | 58 | >90% | 621/826 nested case-control study | 25OHD concentration; >32 vs. <20 ng/mL (means) | 10 | 0.50 (0.32, 0.76) | Type 2 diabetes (validated self-report) | Age, gender, BMI, exercise, season, family history of diabetes, cholesterol, alcohol, smoking, education, supplement use |
| Afzal et al, 2013 Copenhagen City Heart Study [Europe] | 43% | 56 | 100% | 810/9841 | 25OHD concentration Quartiles value NR | 29 | 1.35 (1.09–1.66) | Type 2 diabetes (Self report, medication treated, non fasting glucose, registry based) | sex, age, smoking status (never/ever), BMI, income, and duration and intensity of leisure time physical activities. |
*
number of cases if nested case-control study
†
highest/lowest risk category versus reference category
§
estimated from reported data
25OHD, plasma or serum 25-hydroxyvitamin D; BMI, body mass index; BP, blood pressure; CVD, cardiovascular disease; HR, hazard ratio; IU, international units; ND, no data; OR, odds ratio; RR, relative risk
To convert 25OHD concentration from ng/mL to nmol/L multiply by 2.459