INTRODUCTION
Brain regulation of Vitamin D is possible by neuronal cytochromes (CYP27B1 and CYP24A1) involved in the conversion of 25OHD3 to 1,25-OH2D3, and 1,25-OH2D3 to 24,25-OH2D31. Vitamin D receptors are located in diverse brain areas: cerebellum, thalamus, hypothalamus, basal forebrain, hippocampus, olfactory system, cingulate gyrus, and temporal, and orbital cortex2. It has been proposed that Vitamin D helps to reduce the accumulation of cerebral β-amyloid-42 by innate immune stimulation and phagocytosis activation. Vitamin D also regulates neurotrophic factors and has antioxidant properties3. Its deficiency is common in the older adult. In the United States, the National Health and Nutrition Examination Survey found a combined Vitamin D deficiency and insufficiency (<30 ng/mL) of 42% in the general population4, and, in Mexico, the reported prevalence is 37%. However, other studies have shown a 70-90% deficiency in older adults with cognitive impairment5. An association has been previously established between low levels of Vitamin D and Alzheimers dementia (AD) and all-type dementia6-8. Different studies show that Vitamin D deficiency is associated with other health problems, such as lower general health status. As such, Vitamin D deficiency increases susceptibility and accelerates the disease, rather than being a direct cause of dementia9. Nevertheless, studies have also shown a directly proportional relationship between serum Vitamin D levels and cognitive performance, particularly concerning executive function10. In Mexico, a community study of older adults did not find an association between cognitive performance and Vitamin D levels11,12. Therefore, this study aimed to determine the association of serum Vitamin D levels with mild cognitive impairment (MCI) and AD in older Mexican adults.
METHODS
Study population
This cross-sectional study was conducted at the memory clinic of a tertiary-level university hospital in Mexico City. Each subject completed clinical and neuropsychological evaluations between January 2018 and January 2019. Subjects who were 65 years or older with available serum Vitamin D levels (in the past 6 months) were included and assigned to one of three groups: MCI, probable AD, and normal cognition (NC). For this study, we excluded subjects with a diagnosis of: major depressive disorder without treatment, non-AD dementias, other neurological disorders including structural cerebral lesions which could affect cognitive functions (i.e., acute stroke, brain tumor or normal-pressure hydrocephalus). We also included subjects with malabsorption syndrome (Crohns or celiac disease), chronic kidney disease, osteoporosis, sarcoidosis, tuberculosis, histoplasmosis or an active granulomatous disease. Subjects who were currently receiving Vitamin D supplements or other medications (bisphosphonates, anticonvulsants, antimycotics) for any cause were also excluded. Sociodemographic variables and health status included information about the subjects sex, age, years of education, and presence or absence of diabetes, dyslipidemia, hypertension, hypothyroidism, atrial fibrillation, and polypharmacy13. The local Ethics Committee approved the study (GER-2738-18-19-1).
Neuropsychological evaluation
Subjects assigned to the MCI group met the following criteria: Petersen MCI proposed criteria14 and the Statistical Manual of Mental Disorders-Fifth Edition (DSM-5) criteria for minor neurocognitive disorder15; clinical dementia rating (CDR) scale16 score of 0.5 points; a mini-mental state examination (MMSE)17 score of 24-27; and preserved functional capacity measured by Katz Index (Basic Activities of Daily Living)18 and Lawton-Brody (Instrumental Activities of Daily Living) scale19. The diagnosis of probable AD was made according to the National Institute of Aging and the Alzheimers Association (NIA-AA-2011) and DSM-5 Criteria15,20. A CDR ≥ 1 and an MMSE score < 24 were also considered for diagnosis. The cognitive evaluation included the following tests: verbal fluency test21 (considered abnormal if the subjects did not produce a certain number of words according to their educational level); the Frontal Assessment Battery (FAB) (score < 11 was considered as executive dysfunction)22,23; and the clock-drawing test24 (1 point was assigned for each mistake; greater executive dysfunction was considered with higher scores). The NC group included subjects without a memory complaint and with NC tests results, adjusted for age and educational level.
Serum Vitamin D determination
Serum 25-hidroxyvitamin D [25(OH)D] levels were measured by the ARCHITECT i2000 SR system 5P02 (Abbott). A chemiluminescent microparticle immunoassay was used25. This assay showed precision with a coefficient of variance (CV%) <5%. For low, average, and high Vitamin D levels, precision remained accurate. At an average of 25 OH-D of 5.3 ng/mL, 20.6 ng/mL, and 72.2 ng/mL, the intra-assay CVs were 3.9%, 2.1%, and 2.3%, respectively. The corresponding inter-assay CVs were 1.6%, 1.3%, and 1.2%. The following categories for this analysis were assigned: >30 ng/mL, sufficiency; 21-29 ng/mL, insufficiency; and <20 ng/mL, deficiency12,26.
Statistical analysis
When appropriate, arithmetic means, standard deviations and frequencies or proportions were used to describe variables. Chi-square test was used to compare qualitative data and ANOVA for continuous variables. The cognitive domains evaluated through MMSE were compared after their transformation to z-scores. To determine the strength of association between Vitamin D levels and the subjects cognitive status (MCI or probable AD), univariate and multivariate logistic regression models adjusted for age, sex, and education level were performed. p < 0.05 value was considered significant. Statistical analysis was performed using SPSS software for Windows (SPSS Inc., Chicago. IL, version 22).
RESULTS
A total of 208 subjects were included. Mean age was 79 ± 1 year; 65% (n = 136) were female, and mean educational level was 6.7 ± 2.3 years. Eighty-nine subjects (43%) had AD; 88 (42%) had MCI; and 31 (14%) had NC. Sociodemographic variables and health status are shown in Supplementary Table S1; 75% (n = 157) had polypharmacy, 61% (n = 128) had hypertension diagnosis, 39% (n = 81) had dyslipidemia, 36% (n = 74) had diabetes, 16% (n = 55) had hypothyroidism, and 7% (n = 14) had atrial fibrillation. The overall prevalence of severe Vitamin D deficiency was 54%, being more frequent in the AD group (64%), followed by the MCI (59%) and the NC (13%) (p < 0.001) groups. Compared to the NC and the MCI groups, subjects in the AD group were older (81 vs. 73 and 79 years, respectively; p < 0.001), had less years of education (5.7 vs. 9.9 and 6.3 years, respectively; p < 0.001), and had lower Vitamin D levels (18.2 vs. 26 and 19.2 ng/mL, respectively; p < 0.001).
Cognitive performance scores, according to Vitamin D levels, are shown in Table 1. Subjects with severe deficiency of Vitamin D had a lower MMSE score when compared with the groups of sufficiency and insufficiency (22.7 ± 5.5 vs. 27 ± 3 and 25 ± 4.7 points, respectively; p < 0.001). The Vitamin D deficiency group also had worse performance in all MMSE domains measured: orientation, attention, and evocation. Executive function was measured by the FAB, and statistically significant differences were observed when comparing the deficiency group with the sufficiency of Vitamin D group (10.1 ± 3.2 vs. 13.3 ± 5.5; p < 0.01). The same differences were noted between these groups in the clock-drawing test scores (3.36 ± 2.8 vs. 1.69 ± 2 points; p = 0.002), phonemic verbal fluency number of words produced (6.9 ± 4.4 vs. 12 ± 4 words, respectively; p < 0.001), and semantic verbal fluency (11.2 vs. 12.5 words; p = 0.20).
Cognitive status | Sufficiency (n = 13) | Insufficiency (n = 83) | Deficiency (n = 112) | p value |
---|---|---|---|---|
MMSE, mean (SD) | 27(3) | 25 (4.7) | 22.7 (5.5) | <0.001 |
MMSE domains | ||||
z-scores (SD) | ||||
Orientation | 0.58 (0.59) | 0.15 (0.85) | −0.17 (1.09) | < 0.001 |
Attention | 0.37 (0.88) | 0.15 (0.99) | −0.15 (0.99) | |
Memory (Coding) | 0.42 (1.05) | 0.12 (0.59) | −0.14 (1.19) | |
Memory (Evocation) | 0.75 (0.75) | 0.10 (1.05) | −0.16 (0.93) | < 0.001 0.10 < 0.001 |
Executive function mean (SD) | ||||
Clock-drawing test | 1.69 (2.01) | 2.78 (2.87) | 3.36 (2.81) | 0.02 |
FAB | 13.38 (5.5) | 11.96 (3.82) | 10.18 (3.28) | < 0.001 |
Verbal fluency | 12.08(4.05) | 9.14 (4.8) | 6.99 (4.46) | < 0.001 |
Serum Vitamin D levels: sufficiency: > 30 ng/mL, Insufficiency: 21-29 ng/mL, Deficiency: ≤ 20 ng/mL. MMSE: mini-mental state examination, FAB: Frontal assessment battery. Means and SD presented. z-scores shown on MMSE domains.
*KruskalWallis.
SD: standard deviations.
In the univariate logistic regression analysis, the Vitamin D deficiency and insufficiency groups were significantly associated with MCI (HR: 22.7, 95% confidence interval [CI] 4.6-112; p < 0.001, and HR 2.88, 95% CI 0.75-11.11; p < 0.001, respectively). In the group with dementia diagnosis, when compared with the sufficiency of Vitamin D control group, significant associations were also observed for the deficiency and insufficiency states (HR: 49, 95% CI 7.5-318; p < 0.01 and HR: 5.2, 95% CI 0.9-27.8; p = 0.05, respectively). After adjusting for age, sex, and years of education, the association remained significant in the MCI group in both the deficiency (HR: 25, 95% CI 4.4-139; p < 0.001) and insufficiency states (HR: 4.1, 95% CI 0.9-18.1; p < 0.01). In the AD group, the association also remained significant in both the deficiency (HR: 41.7, 95% CI 5.7-301; p < 0.001) and insufficiency (HR: 6, 95%CI 1.0-36.2; p = 0.04) states (Table 2).
MCI | Model 1 HR (95% CI) | p value | Model 2 HR (95% CI) | p value |
---|---|---|---|---|
Vitamin D | ||||
Sufficiency | 1 | 1 | ||
Insufficiency | 2.88 (0.750-11.11) | 0.05 | 4.15 (0.95-18.14) | 0.05 |
Deficiency | 22.75 (4.61-112) | < 0.001 | 25.02 (4.48-139) | < 0.001 |
Dementia | ||||
Vitamin D | ||||
Sufficiency | 1 | 1 | ||
Insufficiency | 5.25 (0.988-27.89) | 0.05 | 6.07 (1.01-36.20) | 0.04 |
Deficiency | 49 (7.54-318) | < 0.001 | 41.7 (5.76-301) | < 0.001 |
Model 1: univariate analysis, Model 2: multivariate analysis, both adjusted for age, sex and years of education. Vitamin D sufficiency: > 30 ng/mL, Insufficiency: 21-29 ng/mL, Deficiency: ≤ 20 ng/mL.
DISCUSSION
Our study showed that Vitamin D deficiency was independently associated with MCI and AD. Previous studies have established an association between Vitamin D deficiency, all-type dementia, and AD27. A meta-analysis that included 20,750 participants from different countries found a significant association between Vitamin D status and cognitive impairment (HR 1.24, 95% CI 1.14-1.35; p < 0.001)9. A cohort study which included 1200 older Chinese adults and had a 2-year follow-up, observed an association between low Vitamin D levels and worsened global cognitive function (HR: 2.89, 95% CI 1.36-6.14; p = 0.004)28. In another multi-ethnic older adult longitudinal study, with a 4.8-year follow-up, cognitive impairment was more frequent in those subjects with Vitamin D deficiency. The study with the longest follow-up was the Cardiovascular Health Study in the United States. This study included 658 cognitively healthy participants and the authors found a higher risk of all-type dementia (HR: 2.25, 95% CI 1.2-4.1, p = 0.002) and AD (HR: 2.2, 95% CI 1-4.8, p = 0.008) associated with Vitamin D deficiency5.
The pathophysiological mechanisms that link low Vitamin D levels with a greater dementia risk have been explained through the Vitamin D-receptor hypothesis, in which the location of these receptors in many different areas of the brain, including memory areas such as the hippocampus and dental gyrus, is proposed as a possible contributing factor to cognitive impairment2. The 1α-hydroxylase enzyme, which is responsible for the hydroxylation of [25(OH)D] to its active form 1,25 dihydroxyVitamin D3 (1,25-D3), contributes to the homeostasis in calcium signaling, neurotrophic factors, and synaptic plasticity29. In vitro, Vitamin D stimulates macrophages, which, in turn, increase β-amyloid plaque clearance3.
Regarding cognitive domains, our study demonstrated that subjects with Vitamin D deficiency had worse performance in attention, evocation, orientation, and mainly executive function domains. A meta-analysis established a strong association between executive dysfunction, memory, and Vitamin D deficiency30,31. This could be explained by the integrity of frontal-subcortical circuits, in which an injury, including vascular lesions, could produce executive dysfunction. For this reason, low serum levels of Vitamin D could explain the loss of its neuroprotective effects32. In preclinical stages or MCI, only a French study has evaluated Vitamin D deficiency; a positive association between low Vitamin D levels and any type of MCI was demonstrated (HR 25.4, 95% CI 3.2-201.2, p = 0.002)7. This evidence supports the notion that cognitive impairment progression could be modified in the early stages of the disease. The effect of Vitamin D in MCI is caused through the same mechanisms as in dementia since it acts as a neurosteroid (i.e., crosses the blood-brain barrier and binds to its receptor in neurons). It is possible that hypovitaminosis D decreases defense mechanisms, participating in brain dysfunction and cognitive decline. Therefore, it seems reasonable to consider that low levels of Vitamin D in the early or preclinical stages (MCI) participate in the progression of cognitive disorders7,33.
The main limitation of this study is related to its cross-sectional nature. The relationship between Vitamin D deficiency and dementia determined through a single time-point of measurement may be susceptible to bias, as well as differences in age and level of education between groups. Another fact that must be taken into account is that exposure to sunlight was not measured. However, the results of this study open the possibilities for other trials, such as those with a longitudinal design. The strength of our study is that it included Mexican subjects with MCI and AD diagnoses, who were selected on the same season of the year and had complete clinical and neuropsychological evaluations, leading to a comprehensive cognitive domain analysis performed in subjects with Vitamin D deficiency. This could represent an opportunity to influence a potentially modifiable risk factor to improve the course of the disease.
SUPPLEMENTARY DATA
Supplementary data are available at Revista de Investigación Clínica online (www.clinicalandtranslationalinvestigation.com). These data are provided by the corresponding author and published online for the benefit of the reader. The contents of supplementary data are the sole responsibility of the authors.