Table 2.
Observational studies on B vitamins and bone health.
| References | Study Populations/Follow-up Time | Exposures | Outcomes | Effect of B Vitamins | Main Findings |
|---|---|---|---|---|---|
| Cross-Sectional Studies | |||||
| Cagnacci et al. 2003 [20] | 161 Italian postmenopausal women | Serum folate, B12, and homocysteine | BMD | Yes: Folate. No: B12 | Folate predicted BMD; No association found in B12 or homocysteine. |
| Dhonukshe-Rutten et al. 2003 [21] | 194 free-living Dutch frail elderly people 70 years+ | Plasma B12 | BMC, BMD, osteoporosis prevalence | Yes: B12 in women | B12 predicted BMC and BMD in women only. Women with marginal or deficient B12 increased risk of osteoporosis substantially. |
| Golbahar et al. 2004 [22] | 271 postmenopausal Iranian women | Plasma homocysteine, folate, and B12 ; MTHFR C667T polymorphism | Femoral neck and lumbar spine BMD | Yes: Folate. No: B12 | Folate predicted BMD. No correlation between MTHFR and folate or B12. Folate but not MTHFR predicted plasma homocysteine. |
| Abrahamsen et al. 2005 [23] | 1700 Danish postmenopausal women | Dietary intake of B2, B6, folate and B12 | BMD | Yes: B2, B6, folate, and B12 | B2 is the only significant predictor among the B complex for FN BMD in the TT genotype. Lowest quartile of B2, B6, folate and B12 intake reduced BMD in the TT genotype. |
| Golbahar et al. 2005 [24] | 366 postmenopausal Iranian women | RBC 5-MTHFR | BMD | Yes: Folate | RBC 5-MTHF predicted BMD, but not plasma 5-MTHF. |
| Morris et al. 2005 [25] | 1500 U.S. White men and women | Serum and RBC folate, serum B12 and homocysteine | BMD; Osteoporosis prevalence risk | Yes: B12; No: Folate | No association was found between folate and BMD or osteoporosis. Significant risk observed in the lowest quartile of B12 vs. the highest quartile. A positive relationship between B12 and BMD when B12 < 220 pmol/L. |
| Baines et al. 2007 [26] | 328 postmenopausal British women | Plasma homocysteine, serum folate, B6, B12 and MTHFR genotypes | BMD | Yes: Folate; No: B6 and B12 | Folate significantly related to BMD. B6, folate and B12 significantly related to homocysteine level. Homocysteine appeared to be related to BMD. |
| Holstein et al. 2009 [27] | 94 German men and women | Fasting serum folate, B6, B12 | BTM: OC, TRAP; BMD, trabecular thickness, number, and area. | Yes: B6, folate, B12 | OC is lower in those with low level of B vitamins. Trabecular thickness and area are lower in those with low folate. Trabecular number is lower in those with low B6. No association between B vitamins and BMD or homocysteine. |
| Bozkurt et al. 2009 [28] | 178 postmenopausal Turkish women | Serum homocysteine, folate and B12 | BMD at femoral neck and lumbar spine | Yes: B12 ; No: Folate | Homocysteine level was higher in osteoporotic patients vs. those with normal or osteopenia. Only B12 predicted osteoporosis at the lumbar spine and femur. |
| Halıloglu et al. 2010 [29] | 120 Turkish postmenopausal women | Homocysteine, and serum folate, and B12 | BMD, BTM: BAP and CTx | No: Folate and B12 | Folate, B12 not related to BMD or BTM. But homocysteine was related to BTM. |
| Rumbak et al. 2011 [30] | 131 Croatian women | Plasma homocysteine, serum and red blood cell folate, and B12 | BMD from lumbar spine, femoral neck, total femur and distal radius | No: Folate, B12 | No relation found for homocysteine, folate or B12 with BMD |
| Longitudinal Studies | |||||
| Macdonald et al. 2004 [31] | 1241 Scottish women aged 45–54 years/6.6 years | Dietary intake of B2, B6, folate and B12 | BMD; BMD change; BTM, fPYD/Cr and fDPD/Cr and serum P1NP | Yes: B2; No: B6, folate and B12 | B2 intake significantly related to BMD in subjects with MTHFR TT homozygotes. No associations found in BMD change or BTM, nor in CC or CT genotypes. No associations were found in other B vitamins. |
| Stone et al. 2004 [32] | 83 U.S. White women 65+ years from a subset/3.5 and 5.9 years | Serum B12 | BTM: BAP, osteocalcin; hip BMD and calcaneal bone mass. | Yes: B12 | Women at lower serum B12 ≤ 280 pg/mL had higher rate of bone loss from the hip than those at B12 > 280 pg/mL. No association was found with site BMD or BTM. |
| Dhonukshe-Rutten et al. 2005 [33] | 615 men and 652 women aged 76 (SD 6.6) years/3 years | Homocysteine, B12 status and the combined effect | Broadband ultrasound attenuation; BTM: OC and of DPD/Cr and fracture risk | Yes: B12 | Women with B12 < 200 pM and homocysteine > 15 µM had lower BUA, higher DPD/Cr, and higher OC. No differences in men between the different level of homocysteine and B12. High level of homocysteine and/or low level of B12 increased risk by 2.8 and 3.8 folds in men and women, respectively. Lowest quartile of B12 increased fracture risk in women only. Highest quartile of homocysteine significantly increased fracture risk in men. |
| Ravaglia et al. 2005 [34] | 702 Italians aged 65–94 years/4 years | Serum folate, B12 and homocysteine | Fractures | Yes: Folate; No: B12 | Higher level of homocysteine increased fracture risk. Lowest folate quartile had 2-fold increased risk vs. higher quartiles, but no dose-respondent relationship. No association between B12 and fracture risk. |
| Tucker et al. 2005 [35] | 2576 U.S. White men and women 30–87 years | Plasma B12 | BMD at total hip, trochanter, Ward’s area, and femoral neck and at lumbar spine (cross-sectional analysis) | Yes: B12 | A positive relationship between plasma B12 and BMD at hip in men, and at the lumbar spine for women. |
| Gjesdal et al. 2007 [36] | 4766 Norwegian men and women 65–67 years/12.6 years | Plasma homocysteine, folate, B12 and 677C→T and 1298A→C polymorphisms for MTHFR genotypes | Hip fracture | Yes: Folate; No: B12 | Homocysteine increased hip fracture in both genders. Folate inversely related to hip fracture in women only. No association was found for B12 or MTHFR genotypes. |
| Yazdanpanah et al. 2007 [37] | 5305 Dutch men and women aged 55+ years/6–7 years | Dietary intake of B2, B6, folate and B12 | BMD (cross-sectional analysis); fracture risk | Yes: B2, B6; No: Folate and B12 | B2 and B6 positively related to BMD, where B2 is the strongest predictor. Compared to the lowest 3 quartiles of B6, Q4 was related to 23% lower risk in vertebral fracture and 45% lower risk in non- fragility fracture. |
| Cagnacci et al. 2008 [38] | 161 healthy postmenopausal women aged 54 years/5 years | Serum folate, homocysteine and vitamin B12. | BMD (cross-sectional and 5 years follow-up) | Yes: Folate No: B12 | Initial and 5-year follow-up assessments, as well as annual change of lumbar spine BMD was significantly related to serum folate. |
| McLean et al. 2008 [39] | 1002 U.S. White men and women aged 75 years/4 years | Plasma B6, folate and B12, homocysteine | Femoral BMD at baseline and hip fracture | Yes: B6, B12 | B6 inversely related to bone loss. B6 and B12 inversely associated with hip fracture risk, and risk remained elevated after adjusted for BMD and homocysteine. |
| Rejnmark et al. 2008 [40] | 1869 Danish perimenopausal women 43–58 years/10 years | Dietary intake and supplemented folate, B2, and B12 | BMD, fracture risk | Yes: Folate and BMD at 5 years; No: Folate, B2 or B12 | No association was found in folate, B2, or B12 with BMD (cross-sectional) or fracture risk. Folate predicted BMD at year 5 significantly. |
| Yazdanpanah et al. 2008 [41] | 5305 Dutch men and women aged 55+ years/6–7 years | Dietary intake of B2 and folate | BMD fracture risk | Yes: B2 No: Folate | The lowest quartile of B2 in women with TT genotype, bone loss was higher and had 2-fold greater risk in fractures vs. CC type. B2 modified MTHFR C677T variant on fracture risk. No association was found with folate. |
| Dai et al. 2013 [42] | 63154 Chinese men and women 45–74 years/13.8 years | Dietary intake of B1. B2, B3, B6, folate, and B12 | Hip fracture risk | Yes: B6; No: other B vitamins | Dose-dependent inverse relationship between B6 and risk of hip fracture in women only. This association was modified by history of diabetes, where the association was present in those without diabetes prevalence. |
Abbreviations: BAP: Bone-specific alkaline phosphatase; BMC: Bone mineral content; BMD: Bone mineral density; BTM: Bone turnover biomarker; (f) DPD/Cr: (Free) deoxypyridinoline cross-links/Creatinine; FN: Femoral neck; fPYD/Cr: Free pyridinoline cross-links/Creatinine; MTHFR: Methylenetetrahydrofolate reductase; OC: Osteocalcin; RBC 5-MTHFR: Red blood cell 5-methylenetetrahydrofolate; TRAP: Tartrate-resistant acid phosphatase.