Neurology. 2014 (Sep 2); 83 (10): 920–928 ~ FULL TEXT
Thomas J. Littlejohns, MSc, William E. Henley, PhD, Iain A. Lang, PhD,
Cedric Annweiler, MD, PhD, Olivier Beauchet, MD, PhD, Paulo H.M. Chaves, MD, PhD,
Linda Fried, MD, MPH, Bryan R. Kestenbaum, MD, MS, Lewis H. Kuller, MD, DrPH,
Kenneth M. Langa, MD, PhD, Oscar L. Lopez, MD, Katarina Kos, MD, PhD,
Maya Soni, PhD, and David J. Llewellyn
From the University of Exeter Medical School,
Department of Internal Medicine and Geriatrics (C.A., O.B.),
Angers University Hospital,
A new Vitamin D study, published online in Neurology, tracked 1,658 elderly men and women for 5 years. At the start of the study, none of them suffered from dementia. The researchers controlled for many dementia risk factors — including age, education, sex, body mass index, smoking, alcohol use, diabetes and hypertension. They found that those individuals with vitamin D levels BELOW 50 (nanomoles per liter) experienced a 53 percent increased risk for all-cause dementia, and a 69 percent increased risk for Alzheimer’s disease. People with readings of 25 or less were more than twice as likely to have Alzheimer’s or another form of dementia.
OBJECTIVE: To determine whether low vitamin D concentrations are associated with an increased risk of incident all-cause dementia and Alzheimer disease.
METHODS: One thousand six hundred fifty-eight elderly ambulatory adults free from dementia, cardiovascular disease, and stroke who participated in the US population-based Cardiovascular Health Study between 1992-1993 and 1999 were included. Serum 25-hydroxyvitamin D (25(OH)D) concentrations were determined by liquid chromatography-tandem mass spectrometry from blood samples collected in 1992-1993. Incident all-cause dementia and Alzheimer disease status were assessed during follow-up using National Institute of Neurological and Communicative Disorders and Stroke/Alzheimer's Disease and Related Disorders Association criteria.
RESULTS: During a mean follow-up of 5.6 years, 171 participants developed all-cause dementia, including 102 cases of Alzheimer disease. Using Cox proportional hazards models, the multivariate adjusted hazard ratios (95% confidence interval [CI]) for incident all-cause dementia in participants who were severely 25(OH)D deficient (<25 nmol/L) and deficient (?25 to <50 nmol/L) were 2.25 (95% CI: 1.23-4.13) and 1.53 (95% CI: 1.06-2.21) compared to participants with sufficient concentrations (?50 nmol/L). The multivariate adjusted hazard ratios for incident Alzheimer disease in participants who were severely 25(OH)D deficient and deficient compared to participants with sufficient concentrations were 2.22 (95% CI: 1.02-4.83) and 1.69 (95% CI: 1.06-2.69). In multivariate adjusted penalized smoothing spline plots, the risk of all-cause dementia and Alzheimer disease markedly increased below a threshold of 50 nmol/L.
CONCLUSION: Our results confirm that vitamin D deficiency is associated with a substantially increased risk of all-cause dementia and Alzheimer disease. This adds to the ongoing debate about the role of vitamin D in nonskeletal conditions.
From the FULL TEXT Article:
Recent meta-analyses confirm that low serum vitamin D concentrations are associated with prevalent Alzheimer disease (AD) dementia and cognitive impairment. [1–3] This is cause for concern given the high rates of vitamin D deficiency in older adults  and continued uncertainty about the causes of AD and other forms of dementia.  Both the 1,25-dihydroxyvitamin D3 receptor and 1?-hydroxylase, the enzyme responsible for synthesizing the bioactive form of vitamin D, are found throughout the human brain.  In vitro, vitamin D increases the phagocytic clearance of amyloid plaques by stimulating macrophages [7, 8] and reduces amyloid-induced cytotoxicity and apoptosis in primary cortical neurons.  Vitamin D deficiency has also been linked to vascular dysfunction and ischemic stroke risk  as well as brain atrophy.  However, reverse causation is also possible, as the onset of dementia may lead to dietary changes and reduced outdoor activity, which in turn result in lower vitamin D concentrations. 
Previous prospective studies have established that low vitamin D concentrations in elderly adults are associated with an increased risk of cognitive decline. [3, 13–15] Furthermore, it has been hypothesized that the risk of cognitive decline markedly increases below a threshold between 25 and 50 nmol/L.  However, preliminary prospective studies of vitamin D and dementia risk have been discordant. In a small study of 40 high-functioning elderly women, severe vitamin D deficiency (<25 nmol/L) was associated with a higher risk of non-AD dementias but not AD over 7 years.  In contrast, in 10,186 individuals, severe vitamin D deficiency was associated with medical records indicating AD but not vascular dementia over 30 years of follow-up.  The discrepancy in these findings may be due to a lack of statistical power  or use of unstandardized dementia diagnoses from medical records, which may result in considerable misclassification.  We therefore conducted what is to our knowledge the first large, prospective, population-based study incorporating a comprehensive adjudicated assessment of dementia and AD to examine their relationship with vitamin D concentrations.
We have conducted what is to our knowledge the first large, prospective, population-based study to examine vitamin D concentrations in relation to a comprehensive adjudicated assessment of dementia and AD. We observed a strong monotonic association between 25(OH)D concentrations and the risk of both incident all-cause dementia and AD. This association was robust to adjustment for a range of potential confounders and the exclusion of dementia cases that occurred within a year of baseline.
The 2 previous studies that have investigated vitamin D and incident dementia have produced conflicting results. The first found that severe vitamin D deficiency was associated with non-AD dementia but not AD risk.  The second found that severe vitamin D deficiency was associated with AD but not vascular dementia risk.  However, the first study incorporated a small sample of high-functioning women (n = 40), and the lack of association with AD may reflect limited statistical power. The second study relied on registry data for dementia diagnoses, which may have resulted in considerable misclassification. Our results establish that low 25(OH)D concentrations are linked to an increased risk of incident all-cause dementia and AD, and they are consistent with studies suggesting a link with cognitive impairment [1, 3, 12, 26, 28] and cognitive decline. [13–15] Few studies have examined potential mediators of this association, although there was no evidence in the present study or the InCHIANTI  study for mediation by diabetes or hypertension.
A threshold below which the risk of dementia increases markedly has previously been hypothesized to lie in the 25–50 nmol/L range.  The optimal level of vitamin D for general health remains controversial, with the Institute of Medicine recommending 50 nmol/L and the Endocrine Society recommending 75 nmol/L. [29, 30] A post hoc analysis of the Women's Health Initiative randomized controlled trial discovered that a relatively low dose of vitamin D (400 IU) in combination with calcium (1,000 mg) did not protect against dementia over a mean follow-up period of 7.8 years in women who had relatively high serum vitamin D levels at baseline (mean of 49 nmol/L in a small subsample).  Our results clarify that the threshold above which older adults are unlikely to benefit from supplementation with regard to dementia risk is likely to lie in the region of 50 nmol/L when 25(OH)D concentrations are measured using LC-MS. This therefore adds to the ongoing debate regarding optimal vitamin D levels for different health outcomes.
A number of potential mechanisms linking low vitamin D levels with the risk of dementia have been identified.  Vitamin D receptors are expressed throughout the brain, including areas involved in memory such as the hippocampus and dentate gyrus.  Similarly, the enzyme that synthesizes the active form of vitamin D, 1?-hydroxylase, is produced in several cerebral regions. The active form of vitamin D, 1,25dihydroxy-vitamin D3 (1,25-D3), regulates neurotrophin expression, such as nerve growth factor, neurotrophin 3, and glial-derived neurotrophic factor,  and the survival, development, and function of neural cells.  In vitro, vitamin D stimulates macrophages, which increases the clearance of amyloid plaques, a hallmark of AD. [7, 8] Vitamin D also reduces amyloid-induced cytotoxicity and apoptosis in primary cortical neurons.  A recent study found that amyloid-? induction of induced nitric oxide synthase, part of the inflammatory process of AD, is dependent on the disruption of the vitamin D-vitamin D receptor pathway.  Vitamin D supplementation ameliorates age-related decline in learning and memory in aged rats.  In addition, vitamin D deficiency has been linked to cerebrovascular pathology. Meta-analyses establish that 25(OH)D deficiency is associated with an increased risk of incident stroke,  particularly ischemic stroke.  A cross-sectional study of 318 elderly adults found that 25(OH)D deficiency was associated with increased white matter hyperintensity volume and a greater number of large vessel infarcts.  In summary, low vitamin D concentrations may increase the risk of dementia and AD through both neurodegenerative and vascular mechanisms.
Our study has a number of strengths. The study sample was relatively diverse as it was population-based and included white and African-American men and women. A recent systematic review raised the possibility that the consistent observational associations between vitamin D levels and a wide range of health conditions may simply reflect reverse causation.  However, in this study reverse causation is made less likely by the fact that participants were ambulatory and relatively healthy at baseline (their outdoor activity was not likely to be limited by impaired function linked to the onset of dementia). The long follow-up and exclusion of prevalent dementia and incident dementia occurring within a year of baseline also make reverse causation less likely. All-cause dementia and AD in the CHS were diagnosed by a committee of neurologists and psychiatrists using a comprehensive range of data, including neuroimaging, according to international criteria (NINCDS-ADRDA).  Our study also has several limitations. While the CHS is multiethnic, it did not incorporate people of Hispanic or other ethnicities. Due to the exclusion of participants with cardiovascular disease and stroke at baseline, there were few cases of incident vascular dementia (n = 15). It was therefore not possible to investigate the relationship between vitamin D concentrations and incident vascular dementia due to a lack of statistical power, and further research is necessary to investigate generalizability to older adults with vascular dysfunction. In a cohort with a greater burden of vascular and metabolic dysfunction it would also be interesting to investigate these factors as time-varying covariates. The representativeness of our final sample may have been reduced due to the inability to include participants with insufficient serum volume for 25(OH)D measurement (n = 945) as well as those lost to follow-up (n = 596). It is possible that the delay between obtaining the blood samples in 1992–1993 and measuring 25(OH)D concentrations in 2008 could have introduced measurement error; however, this is unlikely to have introduced systematic bias. Despite the wide range of information (including repeat neuroimaging) available to the committee diagnosing all-cause dementia and AD, a degree of misclassification is still likely. In particular, many cases of AD may actually reflect a mixture of pathologies, so caution should be exercised when considering potential mechanisms. As with all observational studies, unmeasured confounding is possible, and our findings do not in themselves demonstrate a causal relationship.
We found a strong association between baseline vitamin D concentrations and the risk of incident all-cause dementia and AD over a mean of 5.6 years of follow-up in ambulatory older adults free from vascular conditions at baseline. Further studies are necessary to replicate our findings and extend them to more diverse populations. It would be useful to conduct prospective studies to investigate the association between vitamin D concentrations and incident vascular dementia and neuroimaging abnormalities. Our findings support the hypothesis that vitamin D may be neuroprotective and that “sufficiency” in the context of dementia risk may be in the region of 50 nmol/L. This information is likely to prove useful in improving the design and reducing the cost of randomized controlled trials investigating whether vitamin D supplements can be used to delay or prevent the onset of dementia and AD in older adults.