|Home | About | Journals | Submit | Contact Us | Français|
Vitamin D may be of interest in the prevention of cognitive impairment, though previous findings are inconclusive. Participants were 1766 adults aged 65 years and older from the Health Survey for England 2000, a nationally representative population-based study. Cognitive impairment was assessed using the Abbreviated Mental Test Score. The cross-sectional relation of serum 25-hydroxyvitamin D quartiles to cognitive impairment was modeled using logistic regression. In all, 212 participants (12%) were cognitively impaired. Odds ratios (95% confidence intervals) for cognitive impairment in the first (8–30 nmol/L), second (31–44 nmol/L), and third (45–65 nmol/L) quartiles of serum 25-hydroxyvitamin D compared with the fourth (66–170 nmol/L) were 2.3 (1.4–3.8), 1.4 (0.8–2.4), and 1.1 (0.6–1.9), after adjustment for age, sex, education, ethnicity, season of testing, and additional risk factors for cognitive impairment (P for linear trend = .001). Our data suggest low serum 25-hydroxyvitamin D is associated with increased odds of cognitive impairment.
Vitamin D deficiency is common in older adults as a result of restricted sunlight exposure, reduced capacity of the skin to produce vitamin D, and reduced dietary intake.1 Vitamin D deficiency in elderly adults is associated with fracture risk, mortality, and various chronic conditions such as type 2 diabetes, although oral supplementation appears to be effective, well tolerated, and economical.2–4 Vitamin D may also be of interest in the prevention of neurodegenerative diseases as ample evidence from in vitro and animal experiments suggests an important role in the expression of neurotrophic factors, the stimulation of adult neurogenesis, calcium homeostasis, and detoxification.5–7 Cognitive impairment is common in elderly adults, although its etiology remains unclear,8,9 and a causal relationship between vitamin D status and brain dysfunction would have major public health implications.5
A number of small clinical studies suggest that serum 25-hydroxyvitamin D [25(OH)D] concentration, an effective indicator of vitamin D status,10 may be associated with dementia and cognitive function. For example, serum 25(OH)D concentrations were lower in 20 female mild dementia cases in comparison with 40 cognitively normal female controls.11 Similarly, a monotonic decrease in serum 25(OH)D was observed when 58 patients with severe Alzheimer’s disease (AD) and 42 mild AD cases were compared with 100 controls.12 However, no significant difference in levels of serum 25(OH)D was observed when comparing 16 dementia cases and 16 controls. 13 Levels of serum 25(OH)D were positively associated with Mini-Mental State Examination (MMSE) scores in 225 outpatients with AD,14 32 patients attending a memory clinic,15 and 2 of 4 measures of cognitive performance in 40 patients with AD and 40 controls.16 However, no association was observed between MMSE scores and serum 25(OH)D concentration in 44 self-neglecting older adults,17 or between serum 25(OH)D and cognitive function across a range of neuropsychological tests in 148 nonsmokers.18
To our knowledge, only 1 previous analysis on this topic has incorporated a large representative population-based sample. McGrath and colleagues observed no linear association between serum 25(OH)D and verbal memory performance in 4809 noninstitutionalized older adults.19 Evidence from well-designed trials is also lacking, though vitamin D supplementation was associated with a significant improvement in choice reaction time after 6 months in 139 ambulatory participants with a history of falls.20 Following vitamin D supplementation, a modest increase in clock drawing test performance, though not verbal fluency, was also observed over 4 weeks in 25 nursing home residents with low 25(OH)D status at baseline.21 Given the paucity of reliable existing evidence, we examined whether low levels of serum 25(OH)D were associated with increased odds of cognitive impairment in a large representative sample of elderly adults.
Participants were from the Health Survey for England 2000 (HSE), a nationally representative population-based study of the adult and child population living in private households plus adults aged 65 years and older resident in institutions.22 For community residents, a stratified sample of 360 postcode sectors (geographic areas) were selected as primary sampling units, and 19 addresses were selected from each of these, giving a total selected sample of 6840 addresses, of which 91% contained private households. All people within eligible households were invited to participate, and 7988 interviews were conducted, of which 1677 were aged 65 years or older (the response rate for this group was 71%). For the institutionalized sample, care homes were selected through multilevel stratified sampling by geographic area. A total of 6 residents aged 65 years and older were randomly selected from each care home. Interviews were conducted in person or using proxies where appropriate. Of the 607 care homes sampled, 544 (90%) took part. Of the 2493 interviews, 1220 (49%) were conducted in person and 1273 (51%) by proxy. A blood sample was obtained from 1818 participants who gave written informed consent, and a valid serum 25(OH)D sample was obtained from 1766 participants (708 men and 1058 women). The HSE was approved by the London Multicentre Research Ethics Committee and performed in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki.
Nonfasting blood samples were collected in a standardized fashion into 6-mL plain tubes (no anticoagulant) by nurses during the HSE.22 Specimens were then dispatched to the Royal Victoria Infirmary in Newcastle, England, for analysis. Analyses were carried out according to standard operating procedures by State Registered Medical Laboratory Scientific Officers under the supervision of the senior officer. Analyses were carried out by the supraregional assay service through a manual 2-step procedure using the DiaSorin Kit (Stillwater, Minn). Step 1 involved rapid extraction of 25(OH)D and related metabolites from serum using acetonitrile. Following extraction, the sample was assayed using an equilibrium radiolabeled antibody procedure whereby an antibody specific for 25(OH)D was incubated with the sample and radiolabeled 125I for 90 minutes at 20°C to 25°C. Separation of bound from free 25(OH)D was achieved by a second 20-minute incubation with an antibody-precipitating complex (donkey antigoat complex/polyethylene glycol). Finally, nonspecific binding (NSB)/addition buffer was added prior to centrifugation and decanting to reduce the interference by NSB. Monthly internal quality control results and external quality assessment results for serum 25(OH)D were acceptable.22
Cognitive function was measured using the Abbreviated Mental Test (AMT), a widely used neurocognitive screening instrument composed of 10 items primarily covering attention, orientation in time and space, and memory.23 Cognitive impairment was defined on the basis of 3 or more incorrect responses of 10. Such a cut-point has been used in previous studies and is a sensitive (81%) and specific (85%) marker of cognitive impairment.24–26 The AMT is also highly correlated with the Mini-Mental State Examination (r = .86–.87).27
Two-sample t tests (equal variances not assumed) and Fisher exact tests were used as appropriate to compare unadjusted continuous and categorical characteristics of the study population by cognitive status. Multivariable logistic regression models were used to determine the relationship of serum 25(OH)D to cognitive impairment. We divided levels of serum 25(OH)D (nmol/L) into quartiles to aid clinical interpretation. In basic adjusted models, we adjusted for age in years, sex, education, and ethnicity (collapsed into white and ethnic minorities) as these are established potential confounders in cognitive research28 and season during which testing occurred to account for seasonal variation in sunlight exposure. In fully adjusted models, we also adjusted for additional correlates of cognitive impairment: current smoking status, alcohol consumption (g/d), psychiatric morbidity (>3 symptoms on the 12-item General Health Questionnaire),29 hypoalbuminemia (serum albumin <3.5 g/dL), and self-reported medical history (diagnosed diabetes, cardiovascular disease, stroke, and hypertension). We also adjusted for impaired mobility (those reportingmoderate or severe difficulty walking) as reduced outdoor activity may result in lowered serum 25(OH)D concentrations.30
Possible 2-way interactions between serum 25(OH)D and age or sex were tested by including product terms in a fully adjusted logistic regression model. As a secondary analysis, we examined whether the same pattern of associations was observed when men and women were analyzed separately or whether institutionalized participants were excluded. An alternative cut-point of 4 or more incorrect responses is sometimes used to define cognitive impairment using the AMT,23,31 so we examined the effect of changing the cut-point used to define cognitive impairment in a sensitivity analysis. Information on body mass index (BMI [kg/m2]) was available for a subsample of 1279 participants, and we examined whether the same pattern of associations was observed with additional adjustment for BMI as vitamin D status is known to vary on the basis ofBMI,32 and BMI is a potential confounder in cognitive research.28 We also examined the odds ratios for cognitive impairment in participants who were severely 25(OH)D deficient (defined as <25 nmol/L), 25(OH)D deficient (defined as ≥25 nmol/L and <50 nmol/L), 25(OH)D insufficient (defined as ≥50 nmol/L and <75 nmol/L) compared with participants with sufficient 25(OH)D (defined as ≥75 nmol/L).33 Clusters, strata, and survey weights derived using inverse probability weighting were used to adjust for the survey design, sampling strategy, and possible nonresponse bias. We used Stata SE version 9.2 for all analyses (StataCorp, College Station, Tex).
The characteristics of the study population are described in Table 1. Unadjusted serum 25(OH)D levels were higher in the cognitively normal than the cognitively impaired, and about half of the people with cognitive impairment had levels of serum 25(OH)D in the first quartile of the distribution (8–30 nmol/L). The cognitively normal participants were also generally younger than the cognitively impaired, more likely to have educational qualifications, consume alcohol, have a higher BMI, and less likely to have impaired mobility, hypoalbuminemia, or a history of stroke.
In an unadjusted logistic regression model, those in the lowest quartile of serum 25(OH)D concentration were more than 4 times more likely to be cognitively impaired, and there was clear evidence of a monotonic relationship (Table 2). This association was attenuated following adjustment for age, sex, education, ethnicity, and season tested, though remained significant. Adjustment for additional correlates of cognitive impairment had little effect, and in the fully adjusted model, those with the lowest levels of serum 25(OH)D were more than twice as likely to be cognitively impaired (Figure 1). No significant 2-way interaction between serum 25(OH)D and age was observed (P = .558).
A different pattern of associations was observed for women and men when examined separately using logistic regression models (Table 3), and there was a 2-way interaction between serum 25(OH)D and sex of borderline statistical significance (P = .048). There was evidence for a linear increase in the odds of cognitive impairment by serum 25(OH)D quartiles for men (P for linear trend < .001), but a smaller association that became nonsignificant after full adjustment was observed for women (P for linear trend = .139). When the sample was restricted to noninstitutionalized participants (n = 1050), a monotonic increase in the odds of cognitive impairment by serum 25(OH)D quartiles was observed (P for linear trend < .001; Table 4). Using an alternative cut-point of 4 or more incorrect responses on the AMT to define cognitive impairment had little effect on the pattern of results observed (results not shown). Odds ratios (95% confidence intervals) for cognitive impairment in participants who were severely 25(OH)D deficient (<25 nmol/L), 25(OH)D deficient (≥25 nmol/L and <50 nmol/L), and 25(OH)D insufficient (≥50 nmol/L and <75 nmol/L) compared with participants with sufficient 25(OH)D (≥75 nmol/L) were 2.72 (1.48–5.00), 1.37 (0.81–2.31), and 0.89 (0.50–1.56) in fully adjusted models (P for linear trend <.001). Additional adjustment for BMI did not change the pattern of associations observed for men, though the association became stronger for women (P for linear trend = .082), and there was no longer a suggestion of a 2-way interaction between serum 25(OH)D and sex (P = .512).
The aim of this study was to examine the association between serum 25(OH)D and cognitive impairment in a large population-based sample of elderly adults. Our results suggest that levels of serum 25(OH)D were generally lower in the cognitively impaired general population, and there was evidence for a monotonic relationship. The association between serum 25(OH)D and cognitive impairment was stronger in men, which appeared to be partly attributable to differences in BMI. Thus, our results suggest that high levels of serum 25(OH)D are associated with lower odds of cognitive impairment.
To date, evidence for the beneficial effect of vitamin D on cognitive function comes mainly from animal and in vitro experiments.5–7 A number of studies suggest that low levels of serum 25(OH)D are associated with poor cognitive function14–16 and dementia,11,12 whereas others have reported no such association.13,17,18 However, these studies incorporated relatively small highly selected clinical samples, were not population based, and did not adjust for important potential confounders. McGrath and colleagues recently reported no linear association between serum 25(OH)D and verbal memory performance in 4809 noninstitutionalized older adults.19 However, given that cognitive impairment and dementia are highly prevalent in those residing in institutions,34,35 it is not clear whether the lack of association in their study reflects cognitive homogeneity within their sample. Alternatively, the briefmeasure of verbal memory incorporated in their analysis may not provide a sensitive measure of cognitive function, or vitamin D may be primarily associated with cognitive domains other than memory. Two trials also suggest that vitamin D supplementation may be associated with improved choice reaction time over 6 months,20 and clock drawing test performance over 4 weeks.21 Taken together with the results from this study, it appears that low serum 25(OH)D may be a risk factor for cognitive impairment.
The bioactive form of 25(OH)D, calcitriol (1,25-dihydroxycholecalciferol), has long been known for its important role in regulating levels of calcium, phosphate homeostasis, and bone mineralization. More recently, it has become clear that receptors for vitamin D are present in a wide variety of cells, including neurons and glial cells, and genes encoding the enzymes involved in the metabolism of calcitriol are also expressed in brain cells.5,36 Calcitriol reduces cellular calcium, inhibits the synthesis of inducible nitric oxide synthase, and increases levels of the antioxidant glutathione, suggesting a role in brain detoxification pathways.6,37 Calcitriol stimulates neurogenesis and regulates the synthesis of neurotrophic factors such as nerve growth factor (NGF), neurotrophin (NT) 3, NT 4, and glial cell line–derived neurotrophic factor (GDNF), which are important for cell differentiation and survival.5,38,39 Calcitriol is also an effective immunosuppressor and may be neuroprotective by inhibiting autoimmune damage to the nervous system.6,38 Low circulating levels of 25(OH)D may be associated with a higher risk of multiple sclerosis,40 and 25(OH)D has also recently been hypothesized to play a role in the pathogenesis of Parkinson disease.41 Accumulating data therefore provide evidence for previously unsuspected roles for vitamin D in brain development and neuroprotection.
A number of methodological issues should be considered when interpreting our findings. Although information on cognitive impairment from a widely used neurocognitive screening test was available, a clinical examination was not incorporated in the HSE. The cross-sectional design of this study means that we cannot determine whether low levels of serum 25(OH)D cause cognitive impairment. However, a causal relation between these factors has been suggested in animal and in vitro experiments5–7 and clinical trials.20,21 Reverse causality also appeared unlikely as additional adjustment for factors that may lead to reduced outdoor activity (notably psychiatric morbidity19 and impaired mobility30) had little effect, though a direct measure of sunlight exposure was not available. There is a potential for nonresponse bias, though the HSE response rate was acceptable and survey weights were used to adjust for possible bias. Given that the elderly English population is predominantly of white ethnic origin, further research is necessary to examine the degree to which our findings generalize to more diverse populations. Although unlikely, a genetic predisposition to both cognitive impairment and low serum25(OH)D might confound the association observed in our study.
In conclusion, we provide new evidence to suggest that serum 25(OH)D is related to cognitive impairment in the elderly population and a potential diagnostic aid for screening or differential diagnosis.42 This is important because serum 25(OH)D may play an important role in the expression of neurotrophic factors, the stimulation of adult neurogenesis, calcium homeostasis, and detoxification.5–7 Furthermore, the association between serum 25(OH)D levels and cognitive impairment underlines the importance of micronutrients in the elderly population. Further research is warranted to investigate if vitamin D supplementation is a cost-effective and safe way of reducing the incidence of cognitive impairment in the growing elderly population around the world.
The HSE is funded by the UK Department of Health. Dr KML was supported by grants from the NIA (K08 AG019180 and R01 AG027010) and a Paul Beeson Physician Faculty Scholars in Aging Research award. Dr IAL is an Academic Speciality Registrar in Public Health supported by the NHS South-West Region Public Health Training Scheme. The sponsors played no part in the study design, the collection, analysis, and interpretation of data, the writing of the report, or the decision to submit the article for publication.