Table 4. Random-Effects Models Illustrating Change in MMSE-Measured Cognitive Function by Serum 25(OH)D Levela.
Serum 25(OH)D Level, nmol/L | All Participants (n=858) |
Nondemented Participants Only (n=829) |
||
---|---|---|---|---|
Unadjusted Modelb | Fully Adjusted Modelc | Unadjusted Modelb | Fully Adjusted Modelc | |
≥75 | 0.0 [Reference] | 0.0 [Reference] | 0.0 [Reference] | 0.0 [Reference] |
≥50 to <75 | −0.085 (0.169) | −0.111 (0.115) | −0.117 (0.174) | −0.130 (0.114) |
≥25 to <50 | −0.139 (0.146) | −0.035 (0.095) | −0.154 (0.153) | −0.051 (0.096) |
<25 | −0.664 (0.146) | −0.321 (0.109) | −0.684 (0.153) | −0.310 (0.109) |
P value for linear trend | <.001 | .03 | <.001 | .04 |
Abbreviations: 25(OH)D, 25-hydroxyvitamin D; MMSE, Mini-Mental State Examination29 (range, 0-30; higher score represents better function).
All data are reported as estimated (SE) β values.
Adjusted for baseline cognitive score only.
Adjusted for age, sex, education, baseline cognitive score, season tested, body mass index, impaired mobility, diabetes, and stroke.