Deficiency studies in animal models and epidemiologic investigations have supported a role of vitamin D in neuropsychiatric and neurodegenerative disorders. The behavioral characterization of the VDR knock-out (VDR-KO) mouse revealed changes consistent with diminished musculoskeletal development and motor impairment (reduced stride length, hyperlocomotion, and reduced habituation in the open field test) (Burne, JJ et al. 2005
). While some studies reported no observed impairments in working memory or anxiety in the VDR-KO model (Burne, JJ et al. 2005
), others showed anxiety-like behavior and behavioral impairment (Kalueff, Lou et al. 2004
; Kalueff, Keisala et al. 2006
). In embryonic animal models, vitamin D deficiency during fetal development resulted in morphological brain changes (Eyles, Brown et al. 2003
), motor impairments (Burne, Becker et al. 2004
) and memory and learning impairments (McGrath, Eyles et al. 2003
; Becker, Eyles et al. 2005
While in vitro and animal models suggest neuroprotective benefits from vitamin D upon exposure, there are inconsistencies in the clinical literature related to vitamin D and cognitive function in the elderly. In a small case control study (n=84) in subjects in the Tromso study, subjects with secondary hyperparathyroidism (SHPT) (n=21) performed worse on cognitive tests associated with working memory (digit span)(Wechsler 1987
), processing speed (Stroop test)(Golden 1978
), language (controlled oral word association) (Spreen and Strauss 1998
) and mood (Becks Depression Inventory) compared to subjects without SHPT (n=63). While the authors also revealed that low serum 25(OH)D concentrations were significantly associated with mood, they were unable to detect an association with 25(OH)D and cognitive function; this may in large part be due to the limited cognitive battery implemented as well as the small sample size under study.
In a much larger cross-sectional investigation of NHANES III data, McGrath et al. found no association between 25(OH)D and cognitive function in adolescence (16–19y) and adults (20–59y), but revealed an inverse association between 25(OH)D and a test of learning and memory in older adults (60–90y) (McGrath, Scragg et al. 2007
). This finding was somewhat surprising, but should be carefully evaluated in the context of the detected difference. Subjects in the highest quintile of vitamin D had the lowest score on this test (6.5 (0.1), compared to scores in the highest quintile (6.4 (0.1) (n=4809). This detected difference is very small and the clinical relevance is questionable. Additionally, the evaluation of an association between 25(OH)D and one test of cognition is not representative of cognitive function.
More recent positive associations between vitamin D and cognitive function in older adults have been revealed. In a cross-sectional study of eighty ambulatory elders (40 with mild dementia and 40 non-demented elders >60yrs), Wilkins et al. reported that vitamin D deficiency (<50nmol/l) was associated with poorer performance on global tests of cognitive function such as the Short Blessed Test (Katzman, Brown et al. 1983
) and a higher score (indicative of poorer cognition) on the Clinical Dementia Rating Sum of Box Scores (CDR)(Morris 1993
) (Wilkins, Sheline et al. 2006
). The authors were not able to detect an association with vitamin D and the cognitive factor score, which is likely due to the small sample size under study (Wilkins, Sheline et al. 2006
). Another positive association between vitamin D and a measure of global cognition was reported by Pryzbesky et al. In a retrospective chart review of thirty-two subjects over 60 years of age that were being evaluated in a memory clinic, the authors reported positive correlations between 25(OH)D concentrations and performance on the Mini-Mental State Examination (MMSE) (r=0.23;P=0.01) (Przybelski and Binkley 2007
In the cases above, the studies were limited in the sample size as well as the outcome evaluation tools used to explore 25(OH)D and cognitive function in older adults. For example, McGrath et al. evaluated cognition in the elderly group using only one cognitive tool, a learning and memory tool. Given the preponderance of evidence of vitamin D in hippocampal health, a more sophisticated cognitive evaluation battery designed to evaluate multiple facets of cognitive performance is necessary. While thestudy by Jorde et al. had a more detailed cognitive battery, the study was limited by its size (n=84). Additionally, the authors did not adjust for multiple testing in their analyses and are thereby limited in the interpretation of their findings.
Contrary to the findings in the NHANES III and Tromso study, and consistent with the findings from Wilkins et al. and Przybelski et al., we recently reported findings of a positive association between 25(OH)D and cognitive function. Using data from the NAME study, (Scott, Peter et al. 2006
), a large cross-sectional study (n=1200) with a comprehensive neuropsychological battery, 25(OH)D was associated with both global and specific aspects of cognitive function (Buell JS et al, manuscript in review). These associations were not consistent across all cognitive domains. Our results showed positive associations between 25(OH)D and primarily measures of executive functioning. These findings suggest that 25(OH)D may play a role in subcortical health and are consistent with vasculoprotective mechanisms of vitamin D. Further, these results may help to explain the null association with 25(OH)D and cognition in the earlier studies (Jorde, Waterloo et al. 2006
; McGrath, Scragg et al. 2007
) and help elucidate a potential mechanism through with vitamin D may exert neuroprotective effects.