vitamin d deficiency and dementia

i  found this recent study suggesting a relationship between vitamin d deficiency and cognitive impairment (see vit d and dementia neuro 2014 in dropbox, or doi.org/10.1212/WNL.0000000000000755#sthash.O7EPM0di.dpuf). as noted in prior blogs and summarized in older reviews (eg, see vit d review nejm 2007 in dropbox, or N Engl J Med 2007;357:266-81), there are vitamin D receptors throughout the body, with likely importance in immune function (and inverse relationship between vitamin d levels and several immunologic diseases, and an intriguing study finding that supplementing vitamin d led to improved outcomes in treatment of tuberculosis), cancer development, diabetes, all-cause mortality, as well as bone and muscle function -- see the dropbox for a myriad of articles. on scanning the contents of the journal Neurology over the past 2 years, there have also been a slew of articles associating vitamin D deficiency with increased risk of multiple sclerosis exacerbations, gait disturbances, cognitive decline in community dwelling women > 75yo (the EPIDOS study), and a systematic review/meta-analysis -- see vit d and dementia metaanal neuro 2012 in dropbox, or  DOI 10.1212/WNL.0b013e31826c197f​ -- confirming many neurologic benefits (but no good RCTs cited). in the current  prospective observational study involving the Cardiovascular Health Study in the 1990s:

--1658 elderly ambulatory adults (mean age 74, 69% women, 88% white,  BMI 27, 9% smokers, 8% heavy drinkers, 10% diabetic, 58% hypertensive), all initially free from dementia, cardiovascular disease and stroke had initial 25(OH)D levels checked and were followed 5.6 years
 --171 people developed dementia, with 102 cases of Alzheimer's (assessment of dementia and type done by neurologists, based on annual cognitive assessments, repeat MRIs, interviews and questionnaires)

--risk of dementia in those 25(OH)D severely deficient (<25 nmol/L, or <10 ng/ml): 2.25 (1.23-4.13)
 --risk of dementia in those 25(OH)D deficient (25-50 nmol/L, or 10-20 ng/ml): 1.53 (1.06-2.21)
--risks of Alzheimer's was similar to above: 2.22 for severely deficient and 1.69 for deficient

so, intriguing. makes sense that the existence of vitamin d receptors throughout the body (and the brain) has some evolutionary advantage. and vitamin d deficiency is extraordinarily common in our northern climates (i have routinely checked patients in boston for their vitamin d levels for the past 15 years and only rarely find someone who is replete, and pretty frequently find those with undetectable levels). and it is cheap, easy to replete, and no evidence of adverse effects (unless serum levels are dramatically high, eg over 150 ng/ml). reasonable in vitro mechanisms for preventing dementia: vit D increases phagocytic clearance of amyloid plaques by stimulating macrophages. also vitamin d regulates nerve growth factor, the synthesis of neurotransmitters,  and the function of neural cells, and decreases amyloid-induced neurotoxicity/apoptosis. vit d deficiency is also assoc with ischemic strokes.  the issue with vitamin d is that most of the studies done are observational ones, and one might think that those with higher vitamin d levels may be different than those with lower levels -- reverse causation (eg, those with higher levels might be fitter: going outside more, perhaps more exercise, etc, and are therefore at lower risk and having nothing to do with vitamin d levels -- and I for one have been led astray by observational studies showing, for instance, that estrogen replacement therapy or decreasing homocysteine levels would be cardioprotective...).  that all being said, there are only a few RCTs with patients' randomized to vitamin D/placebo finding benefit (benefit has been found in decreasing falls in elderly, and some data on decreasing fracture risk, leading the Institute of Medicine to publish their 2011 report suggesting testing and treating vitamin d deficiency -- see for example the older blog appended below). there was one review in BMJ this year confirming benefit of higher vitamin D levels for cardiovascular mortality, cancer mortality, non-cancer mortality and all-cause mortality in observational studies. the intervention studies, however, did  find an 11% relative decrease in all cause mortality but only for vitamin D₃ supplementation  (see vit d and mortality bmj 2014 in dropbox, or doi: http://dx.doi.org/10.1136/bmj.g1903.)