- Click here for explanation of classification scheme.

To review and synthesize the literature regarding the association of serum 25(OH)D concentrations with bone mineral density.

• Randomized controlled trials (RCTs) regarding serum 25(OH)D and bone density
• Single prospective cohort trials
• Case-control studies.

Study did not evaluate bone mineral density (BMD).

Recruitment

The following databases were searched

• MEDLINE (1996 to the third week of June 2006)
• Embase (2002 to the 25th week of 2006)
• CINAHL (1982 to the fourth week of June 2006)
• AMED (1985 to June 2006)
• Biological Abstracts (1990 to February 2005)
• The Cochrane Central Register of Controlled Trials (CENTRAL; second quarter of 2006).

No language restrictions were applied, and studies were restricted to human subjects. 

Design

Studies for inclusion were limited, whenever possible, to randomized, controlled trial (RCTs) to limit bias. Prospective cohorts, case-control and before-after studies were included due to the lack of studies addressing the association between serum 25(OH)D concentrations and bone health outcomes. 

The results of the search were assessed using a three-step process:

• First, bibliographic records, including title, keywords and abstract, were screened by one reviewer
• Potentially relevant records were then screened independently by two reviewers using the full-text report and strict eligibility criteria
• Conflicts were discussed and resolved through consensus or adjudication by a third reviewer
• Relevant studies were subsequently assessed for study design and categorized by question. 

Statistical Analysis

• Meta-analysis of RCTs that assessed interventions, populations and outcomes (e.g., fractures or falls) was concluded using a random effects model, with an assessment of statistical heterogeneity
• For continuous outcomes (e.g., serum 25(OH)D concentrations and Bone Mineral Density (BMD), the difference in means between treatment groups was used for the meta-analyses: The absolute change in 25(OH)D concentrations was used for quantitative pooling. A weighted average was used to calculate the 25(OH)D values for the combined treatment and placebo group. The difference in means was then calculated using the weighted averages for the two combined groups.
• To avoid differences in the reporting of units for 25(OH)D concentrations, (i.e., nmol per L, ng per mL), all values were converted to nmol per L that was the unit used for data synthesis
• For the pooling of BMD results, the percent change in BMD from baseline in the treatment vs. control or placebo was used as the unit of analysis.

Timing of Measurements

The results of the search were assessed using a three-step process:

• First, bibliographic records, including title, keywords and abstract, were screened by one reviewer
• Potentially relevant records were then screened independently by two reviewers using the full-text report and strict eligibility criteria
• Conflicts were discussed and resolved through consensus or adjudication by a third reviewer
• Relevant studies were subsequently assessed for study design and categorized by question. 

Dependent Variables

Bone mineral density.

Independent Variables

Serum 25(OH)D.

Control Variables

• BMI or weight
• Age
• Physical activity
• Calcium use
• Smoking status
• Hormone levels.

• 19 studies evaluated the association between serum 25(OH)D and BMD
• Six were RCT
• Seven were single-prospective cohorts
• Six were case control studies
• All RCTs included postmenopausal women.
• Four cohorts included females only
• Three cohorts included both genders
• Three case-control studies included females only
• Two case-control studies included both genders
• One case control included 100% male subjects
• Four of the seven cohorts adjusted for either BMI or weight
• Three cohorts adjusted for age
• Two cohorts adjusted for physical activity, calcium use, smoking status, or levels of other hormones. 

Findings

Three cohorts evaluated the relationship between 25(OH)D levels and BMD and five examined the relation between 25(OHD levels and changes in BMD.

• Four of seven cohorts reported an association between serum 25(0H)D and femoral neck BMD
• One cohort found a positive association between changes in serum 25(OH)D and lumbar spine (but not femoral neck) BMD
• Stone et al found that women in the highest quartile of serum 25(OH)D (greater than 80nmol per L) had a mean annual loss in total hip BMD of -0.1%, compared to a -0.7% loss in the lower quartile (less than 52.5 nmol per L)
• Two cohorts found a positive association between serum 25(OH)D and BMD of the femoral neck
• Of the six case-control studies that evaluated the relationship between 25(OH)D and BMD, one reported a weak association between 25(OH)D and BMC of the femoral neck
• Five RCTs and three cohort studies did not find an association between serum 25(OH)D and BMD or bone loss
• Four cohort studies found a significant association between 25(OH)D and lumbar spine BMD which was weak
• Six case-control studies suggested an association between 25(OH)D and BMD and the association was the most consistent at the femoral neck BMD
• In some studies, it was unclear whether the effect of serum 25(OH)D on bone loss was mediated by serum PTH.

There is fair evidence to support an association between serum 25(OH)D and BMD of changes in BMD at the femoral neck.

Government:

NIH

This particular aspect of the review was not addressed as thoroughly as others. Some aspects are unclear.