Characteristics of randomized controlled trial studies included in the meta-analysis
| Author (year) | Study location | N subjects (Tx/Px) | Sex | Ethnicity | BMI (kg/m2) | Health status | Vitamin D levels at baseline (nmol/L) | Vitamin D levels after intervention (nmol/L) | Type of vitamin D; mode of delivery; dose (IU/day) | Vitamin D analytical measures | Duration (months) | Outcome measured |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Mitri et al. (2011) | U.S. | 23/24 | M/F | Multiethnic | ≥25 (≥23 if Asian) | Prediabetes (IGT, IFG, HbA1c ≥5.8%) | T 59.9 ± 2.7,§ P 60.4 ± 2.7§ | T 15.7 ± 3.7§ (change from baseline), P −20.5 ± 3.5§ | Vitamin D3; oral; 2,000 | LC-MS | 4 | Insulin sensitivity index (FIVGTT), 2-h glucose (OGTT) |
| Harris et al. (2012) | U.S. | 43/46 | M/F | African American | 25–39.9 | Prediabetes or early T2D (HbA1c 6.7–7%), baseline 25(OH)D <50 nmol/L | T 39.6 ± 12.9, P 38.2 ± 15.5 | T 41.6 ± 3.1§ (change from baseline), P 0.9 ± 2.9§ | Vitamin D3; oral; 4,000 | RIA | 3 | Matsuda index, 2-h glucose (OGTT) |
| Iraj et al. (2012) | Iran | 20/20 | M/F | Middle Eastern | ≥25 | First-degree relatives with T2D, 25(OH)D <75 nmol/L, prediabetes | T 27.5 ± 15.0, P 28.2 ± 15.0 | T 87.4 ± 49.9, P 38.7 ± 54.9 | Vitamin D3; injection; 10,000 | CLIA | 2 | Matsuda index, AUCglucose (OGTT) |
| Davidson et al. (2013) | U.S. | 56/53 | M/F | Multiethnic (Hispanic and African American) | ≥25 | Prediabetes, 25(OH)D <75 nmol/L | T 54.9 ± 12.5, P 54.9 ± 12.5 | T 169.6 ± 37.4, P 54.9 ± 17.4 | Vitamin D3; oral; 12,695 | LC-MS | 3 | Matsuda index, 2-h glucose |
| Oosterwerff et al. (2014) | Netherlands | 53/57 | M/F | Multiethnic (Morocco, Suriname, Turkey) | ≥25 | IFG or IGT, 25(OH)D <50 nmol/L | T 25.0 ± 10.8, P 22.0 ± 11.0 | T 60.0 ± 16.0, P 23.0 ± 15.0 | Vitamin D3; oral; 1,200 | LC-MS | 4 | Insulin sensitivity index, 2-h glucose |
| Gagnon et al. (2014) | Australia | 35/45 | M/F | Multiethnic | 25–40 | 25(OH)D <50 nmol/L, prediabetes | T 47.0 ± 13.0, P 43.2 ± 13.0 | T 89.1 ± 16.5, P 41.4 ± 16.3 | Vitamin D3; oral; 2,000 | CLIA | 6 | Matsuda index |
| Kampmann et al. (2014) | Denmark | 7/8 | M/F | NR | ≥25 | T2D, 25(OH)D <50 nmol/L | T 31.0 ± 4.9,§ P 34.8 ± 3.8§ | T 104.9 ± 19.0,§ P 32.1 ± 3.8§ | Vitamin D3; oral; 6,400 | ELISA | 3 | M value derived from clamp |
| Mitchell et al. (2015) | U.S. | 40/50 | M/F | Multiethnic | Median 25 | 25(OH)D <50 nmol/L | T 44.9 ± 17.5, P 44.9 ± 17.5 | T 107.3 ± 29.9, P 49.9 ± 24.9 | Vitamin D2; oral; 7,142 | LC-MS | 3 | Insulin sensitivity index (FIVGTT) |
| Tuomainen et al. (2015) | Finland | T1/T2/P (24/21/21) | M/F | NR | ≥25 | Prediabetes, 25(OH)D <75 nmol/L | Mean baseline T + P 57.0 ± 11.0 | T1 27.7 ± 17.2 (change from baseline) | (T1) Vitamin D3; oral; 1,600 | HPLC-CEAD | 5 | Matsuda index |
| T2 45.0 ± 23.4 (change from baseline), P 4.1 ± 17.3 | (T2) Vitamin D3; oral; 3,200 | |||||||||||
| Wagner et al. (2016) | Sweden | 21/22 | M/F | NR | ≤32 | Prediabetes or drug-naive diabetes, HbA1c ≤7.9%, FPG <9 mmol/L, 25(OH)D <75 nmol/L | T median 43.0 (IQR 36.0–50.0), P 43.0 (37.0–54.0) | Median change T 42.0 (IQR 32.0–50.0), P 0.0 (−7.0 to 11.0) | Vitamin D3; oral; 4,285 | CLIA | 2 | GIR derived from clamp |
| Yeow et al. (2015) | Malaysia | 13/13 | F | Asian | 23–31 | History of GDM during last pregnancy 6–48 months postpartum; hypovitaminosis (25(OH)D 15–50 nmol/L) | T median 35.6 (IQR 25.6–43.9), P 35.1 (21.6–40.7) | Median change T 51.1 (IQR 39.9–76.1), P −0.2 (−10.18 to 11.8) | Vitamin D3; oral; 4,000 | ECLIA | 6 | Insulin sensitivity index, 2-h glucose |
| Barengolts et al. (2015) | U.S. | 87/86 | M | African American | 28–39 | Fasting glucose 5.3–6.9 mmol/L, HbA1c 5.7–6.4%, 25(OH)D 12.5–75 nmol/L | T 36.7 ± 11.7, P 34.9 ± 12.0 | T 120.1 ± 45.9, P 49.7 ± 18.2 | Vitamin D2; oral; 7,142 | CLIA | 12 | Matsuda index |
| Garg et al. (2015) | India | 15/17 | F | Asian | ≥23 | PCOS, 25(OH)D <50 nmol/L | P 16.9 ± 6.1 | T 78.6 ± 34.6, P 16.7 ± 5.8 | Vitamin D3; oral; 4,000 | CLIA | 6 | Matsuda index, AUCglucose |
| Mousa et al. (2017) | Australia | 28/26 | M/F | Multiethnic | ≥25 | 25(OH)D <50 nmol/L | T 31.4 ± 12.6, P 34.2 ± 10.0 | T 88.4 ± 21.0, P 36.1 ± 15.3 | Vitamin D3; oral; 4,000 | CLIA | 4 | M value derived from clamp |
| Gulseth et al. (2017) | Norway | 33/29 | M/F | Multiethnic (Nordic and South Asian) | <45 | T2D, 25(OH)D <50 nmol/L | T 38.0 ± 11.9, P 36.8 ± 12.6 | T 53.7 ± 9.2, P 38.2 ± 12.9 | Vitamin D3; injection and oral; 3,333 | RIA | 6 | GIR derived from clamp |
| Lerchbaum et al. (2017) | Austria | 49/49 | M | NR | Median 25 | 25(OH)D <75 nmol/L | T median 52.0 (IQR 42.0–65.0), P 51.0 (43.0–68.0) | Median T 107.0 (IQR 89.0–119.0), P 69.0 (46.0–79.0) | Vitamin D3; oral; 2,857 | LC-MS | 3 | Matsuda index, AUCglucose |
| Moreira-Lucas et al. (2017) | Canada | 36/35 | M/F | Multiethnic | <40 | 25(OH)D ≤65 nmol/L, HbA1c 5.4%–6.4% | T 48.1 ± 14.3, P 47.6 ± 14.3 | Mean change T 50.6 (95% CI 36.7, 64.6), P −2.11 (−6.11, 1.89) | Vitamin D3; oral; 4,000 | LC-MS | 6 | Matsuda index, AUCglucose |
| Cefalo et al. (2018) | Italy | 9/9 | M/F | NR | >25 | BMI ≥30 kg/m2, 25(OH)D <75 nmol/L | T 36.7 ± 13.2, P 34.7 ± 21.1 | T 74.8 ± 18.7, P 41.7 ± 7.7 | Vitamin D3; oral; 3,571 | NR | 3 | Insulin-mediated glucose uptake, AUCglucose (from OGTT) |
CLIA, chemiluminescence immunoassay; ECLIA, electrochemiluminescence immunoassay; FPG, fasting plasma glucose; FIVGTT, frequently sampled intravenous glucose tolerance test; HPLC-CEAD, high-performance liquid chromatography-coulometric electrode array detector; IFG, impaired fasting glucose; IGT, impaired glucose tolerance; IQR, interquartile range; LC-MS, liquid chromatography–mass spectrophotometry; M, male; F, female; NR, not reported; P, placebo; RIA, radioimmunoassay; Tx/Px, treatment/placebo group; T, treatment; T2D, type 2 diabetes.
↵§ Data are means ± SE.