点击显示 收起
1 From the Gerontological Research Programme, Faculty of Medicine, National University of Singapore, Singapore (LF, T-PN, LC, and MN); the Department of Psychological Medicine, National University of Singapore, Singapore (LF, T-PN, and E-HK); and the Cognitive Neuroscience Laboratory, Singapore Health Service, Singapore (LC)
2 Supported by a research grant (no. 03-1-21-17-214) from the Biomedical Research Council, Agency for Science, Technology and Research (ASTAR). 3 Reprints not available. Address correspondence to T-P Ng, Gerontological Research Programme, National University of Singapore, Department of Psychological Medicine, National University Hospital, 5 Lower Kent Ridge Road, Singapore 119074. E-mail: pcmngtp{at}nus.edu.sg.
ABSTRACT
Background: The relations of elevated homocysteine, low folate, and vitamin B-12 with cognitive performance in nondemented elderly are not well established. Limited research data suggest differential effects of homocysteine and folate on specific cognitive domains.
Objective: The aim was to examine the independent associations of homocysteine, folate, and vitamin B-12 with cognitive performance in high-functioning elderly Chinese.
Design: Homocysteine, folate, and vitamin B-12 concentrations were measured in fasting blood samples of 451 Chinese aged 55 y with Mini-Mental State Examination scores 24 and who were considered fully independent based on Activities of Daily Living score. Cognitive functions were assessed by a neuropsychological test battery. Independent associations (standardized ß) were determined in multiple linear regression models that simultaneously controlled for potential confounders.
Results: Log-transformed homocysteine was inversely associated with performance on Block Design (ß = –0.319, P = 0.006) and the written Symbol Digit Modality Test (ß = –0.129, P = 0.031). Log-transformed folate was significantly associated with Rey Auditory Verbal Learning Test delayed recall (ß = 0.139, P = 0.010), verbal learning (ß = 0.112, P = 0.038), percentage of forgetting (ß = –0.139, P = 0.013), and the Categorical Verbal Fluency test (ß = 0.104, P = 0.042). Vitamin B-12 was not significantly associated with any cognitive test score.
Conclusions: In this high-functioning elderly Chinese population, elevated homocysteine is associated with deficits in constructional ability and processing speed and folate is associated with measures of episodic memory and language. Our results provide support for differential effects of homocysteine and folate on specific cognitive functions.
Key Words: Homocysteine • folate • vitamin B-12 • cognitive function • Chinese
INTRODUCTION
Recent research has investigated the link between plasma homocysteine, folate, and vitamin B-12 and the risk of dementia and cognitive impairment in older adults. The thiol-containing amino acid homocysteine links the methionine cycle with the folate cycle and is a sensitive marker for folate and vitamin B-12 deficiency (1, 2). To date, studies of the associations of elevated concentrations of homocysteine and decreased folate and vitamin B-12 with cognitive function in the elderly have yielded inconsistent findings.
Previous studies have observed that dementia patients have elevated homocysteine or decreased folate and vitamin B-12 concentrations compared with control subjects (3-5). Longitudinal studies have reported positive findings linking elevated concentrations of homocysteine or lowered concentrations of B vitamins with cognitive decline and dementia (6-11); however, negative findings have also been reported (12-14). Randomized controlled trials have shown that supplements of folic acid or vitamin B-12 have no apparent efficacy in improving cognitive function (15, 16).
In studies conducted in nondemented, healthy elderly populations, some investigators have reported an association of elevated homocysteine or decreased B vitamins with lowered cognitive performance (17-31). Limited data from a few studies also suggest that homocysteine and folate may be associated differentially with specific cognitive domains (17, 25, 30). However, many studies had a limited sample size (17, 18, 21, 23, 27) or measured only a limited range of cognitive domains (20, 24, 26, 31). Furthermore, some studies did not measure B vitamins and homocysteine simultaneously (18, 19, 21, 22, 28), and thus could not examine the differential contributions of each of these closely associated micronutrients. In a triethnic population, Wright (26) reported that an association between homocysteine and Mini-Mental State Examination (MMSE) score was observed only in white and Hispanic subjects but not in black subjects. This suggests that ethnicity may also have a modulating effect, either by itself or as a surrogate of ethnicity-related factors. In the present study, we aimed to determine the independent association between homocysteine, folate, and vitamin B-12 and cognitive function in a group of community-dwelling Singapore Chinese elderly with generally good cognitive and physical functioning.
SUBJECTS AND METHODS
Subjects
The subjects in the present study were identified from participants of the Singapore Longitudinal Ageing Study, a community-based epidemiologic study of aging and health. Briefly, a whole population of all residents (Singapore citizens and permanent residents) in contiguous precincts within 5 districts in the South East region of Singapore who were aged 55 y were identified from a door-to-door census and invited to participate in the study. The study was approved by National University of Singapore Institutional Review Board. All participants signed an informed consent letter before participating. The response rate was 78.2%.
An exhaustive neuropsychological assessment was performed on a 1-in-3 random sample of the respondents who had the sufficient visual, language, and motor abilities that are required to complete the tests. The study was conducted on Chinese elderly subjects, because the numbers of Malay and Indian subjects were small. Because demented elderly persons are grossly impaired on cognition and may hence be prone to malnutrition, their inclusion may overestimate the potential associations and render the interpretations of cause-and-effect difficult. Thus, only respondents with a MMSE score of 24 and who were fully independent based on the Activity of Daily Living (ADL) test were selected for the study.
Laboratory measurements
Venous blood samples after overnight fasting were drawn from 0900 to 0930 according to standard procedures. Blood samples were transported to Singapore National University Hospital Referral Laboratory on ice within 2 h, and the plasma or serum was isolated and stored at –80°C before analysis. Plasma total homocysteine was measured by automated chemiluminescent enzyme immunoassay method (Diagnostic Products Corporation, Los Angeles, CA); the CV ranged from 4.1% to 10.4%. Folate and vitamin B-12 was measured in serum by radioassay with an Elecsys Folate II reagent kit (CVs ranged from 6.1% to 13.8%) and an Elecsys Vitamin B-12 reagent kit (CVs ranged from 3.2% to 7.6%; Roche Diagnostic, Indianapolis, IN), respectively.
Cognitive tests
Trained research assistants administered a comprehensive neuropsychological test battery that assesses a wide range of cognitive domains, including attention and working memory (Digit Span and Spatial Span), verbal learning and memory (Rey Auditory Verbal Learning Test; RAVLT), visual memory (Visual Reproduction), language (Categorical Verbal Fluency), executive function (Design Fluency), information processing speed (Symbol Digit Modality Test; SDMT), visual conceptual ability, visual-motor tracking and mental flexibility (Trail Making test A and B), and visual spatial and constructional ability (Block Design). The assessment was administered in English, Mandarin, or Chinese dialects according to the subject's habitual language. The entire battery took 1.5–2 h to complete.
Digit Span
In Digit Span Forward (32), the examiner reads strings of numbers in series with increasing length, and the examinee is asked to repeat the string in the exact order. In Digit Span Backwards, the examinee is asked to say the strings in reverse order.
Spatial Span
A computerized version of the Spatial Span test (33) was used; Spatial Forward was performed by having the examinee touch the same square on the computer screen in the same order as they were flashed up (changing from blue to red). In Spatial Backwards, the subject is asked to touch the squares in the reverse order.
Rey Auditory Verbal Learning Test
In the RAVLT (34), the examiner reads a semantically unrelated word list (list A) to the examinee in a series of 5 trials. After each learning trial, the examinee is asked to repeat all the words he or she can remember. A second distracter word list (list B) is then presented. Thirty minutes later, the examinee is asked to recall all the words he or she can remember from list A again.
Visual Reproduction
In the Visual Reproduction test (33), the examinee is allowed 10 s to observe each of 5 design cards (3 cards with one design and 2 cards with 2 designs) and reproduce them from memory. Thirty minutes later, the examinee is asked to reproduce the designs again.
Categorical Verbal Fluency
In this test, the examinee is asked to produce as many words as possible in one minute from 3 categories (animal, vegetable, and fruit) respectively.
Design Fluency
For the Design Fluency test (35), the examinee is asked to connect dots to make different designs with 4 straight lines in 60 s. In condition 1, there are only black dots presented. In condition 2, there are black and empty dots and the subject is asked to only use the empty dots. In condition 3, the subject is asked to alternate between black and empty dots.
Symbol Digit Modality Test
In the written version of the SDMT(36), the examinee is asked to write as many numbers as he or she can in the boxes below a series of symbols according to the key provided at the top of the page within 90 s. In the oral version, the examiner records the numbers spoken by the subjects.
Trail Making Test A and B
In part A of this test (37), the examinee is timed to connect the numbers in sequence as fast as possible with a pencil. In Part B, the subjects is asked to switch between number and letters.
Block Design
For the Block Design test (32), the examinee is asked to replicate models or pictures of two-color designs with blocks. The designs progress in difficulty from simple two-block designs to more complex, nine-block designs.
Related data
A detailed interview conducted by research nurses at the first visit collected information on related variables that included sociodemographic information (age, sex, and education), substance use (cigarette smoking and alcohol consumption), medical conditions, weight, and height. The Geriatric Depression Scale (38) was administered as a measure of depression. The MMSE (39) was administered as a global measure of cognitive function. Functional status was assessed by the respondent's level of dependency in performing basic ADL found in the Barthel Index (40). Only subjects fully independent on all 10 ADL items were included.
Statistical analysis
After omitting a small number of respondents with missing or extreme outlier data for smoking (1 respondent), alcohol consumption (1 respondent), Geriatric Depression Scale (1 respondent), homocysteine (1 respodent), folate (2 respondents), vitamin B-12 (1 respodent), and creatinine (1 respondent), data were analyzed for the remaining 451 elderly Chinese subjects aged 55 y who had MMSE scores 24, who were fully independent based on ADL scores, and who completed cognitive assessment and homocysteine, folate, and vitamin B-12 measurements between September 2003 and June 2005. Because not all subjects completed the whole test battery and some tests (ie, Trail Making Test and Block Design) were subsequent additions to the test battery, the results for different neuropsychological measures were based on varying numbers of subjects (ranging from 72 to 451).
Multiple linear regression was used to examine the relations between neurocognitive performance and homocysteine, folate, and vitamin B-12 concentrations, with control for potential confounding variables to evaluate whether the relations were altered by these other variables. We first modeled the relation between cognitive test scores and homocysteine, folate, or vitamin B-12 individually in base models (model 1), which controlled for age, sex, and the number of years of education. In a higher-order model (model 2), we added, as covariates, cardiovascular disease risk factors (including cigarette smoking, alcohol consumption, body mass index, and total cholesterol); chronic diseases including hypertension, diabetes, and cardiovascular disease (heart attack, heart failure, atrial fibrillation, or stroke); renal function (by measuring the serum creatinine concentration); and depression (measured by using the Geriatric Depression Scale). In the final model (model 3), we included homocysteine, folate, and vitamin B-12 into the regression models simultaneously with all the covariates tested in model 2. Because homocysteine, folate, vitamin B-12, and creatinine data were highly skewed to the right, they were natural logarithm–transformed before data analysis. To facilitate meaningful comparisons of results across different cognitive tests with varying metrics, standardized ß coefficient estimates are presented. All data analyses were done with SPSS version 13.0 (SPSS Inc, Chicago, IL).
RESULTS
The characteristics of the study population are summarized in Table 1 and Table 2. The mean (±SD) age of the respondents was 64.2 ± 6.7 y, and a larger proportion were women (60.2%). The mean (±SD) number of years of education was 6.9 ± 4.4 y, and there were only small proportions of regular daily drinkers (10.3%) and smokers (exsmokers and current smokers: 15.2%). Nearly one-half of the population had a history of hypertension (45.7%), and 12.3% of them had diabetes. Altogether, 10.5% of the respondents had a history of heart failure, heart attack, atrial fibrillation, or stroke.
View this table:
TABLE 1. Demographic and laboratory test characteristics of the study subjects1
View this table:
TABLE 2. Cognitive performance of the study subjects1
Homocysteine was significantly correlated with folate (Pearson correlation r = –0.235, P < 0.001), vitamin B-12 (r = –0.210, P < 0.001), and creatinine (r = 0.494, P < 0.001); the correlation (r) between folate and vitamin B-12 was 0.161 (P < 0.001).
In base models (model 1), log-transformed homocysteine was inversely and significantly associated with Design Fluency condition 2 (ß = –0.143), SDMT written version (ß = –0.115), and Block Design total raw score (ß = –0.296) (Table 3). In multivariable models that controlled for other covariates, homocysteine was significantly associated with 2 cognitive measures: Block Design (ß = –0.319) and SDMT written version (ß = –0.129), independent of a wide range of potential confounders as well as folate and vitamin B-12 (model 3, Table 4).
View this table:
TABLE 3. Associations between natural log-transformed homocysteine, folate, vitamin B-12 and cognitive test scores (model 1)1
View this table:
TABLE 4. Associations between natural log-transformed homocysteine and cognitive test scores, adjusted for confounding in multivariate models1
Log-transformed folate was significantly associated with RAVLT immediate recall, RAVLT delayed recall, RAVLT verbal learning, RAVLT percentage of forgetting, and Categorical Verbal Fluency in both base model 1 and model 2 (Table 3 and Table 5). In model 3, which controlled additionally for homocysteine and vitamin B-12, independent associations of folate with these cognitive scores remained significant, except for RAVLT immediate recall [RAVLT delayed recall ß = 0.139, RAVLT verbal learning ß= 0.112, RAVLT percentage of forgetting ß = –0.139, and Categorical Verbal Fluency ß = 0.104 (Table 5)]. Log-transformed vitamin B-12 was not associated significantly with any cognitive measure in all 3 regression models.
View this table:
TABLE 5. Associations between natural log-transformed folate and cognitive test scores, adjusted for confounding in multivariate models1
DISCUSSION
Our results, which were based on a community sample of older Chinese adults with generally well global cognitive and physical functioning, suggest that homocysteine and folate may have independent and differential associations with specific domains of cognitive function. Homocysteine appeared to be specifically associated with constructional ability and information processing speed; folate appeared to be associated specifically with episodic memory and language ability. These associations were independent of a wide range of confounding factors and also of each other.
With regard to executive control function measured with Design Fluency, no clear statistically significant associations with homocysteine or folate were observed, although the marginally significant associations shown in several models do not rule out this possibility. No significant association with Categorical Verbal Fluency was observed. With regard to vitamin B-12, there was no observable significant association with performance on any cognitive test.
Our findings in a Chinese population support similar associations of homocysteine and folate with cognitive performance reported in Western studies, even though there may be subtle differences between different ethnic populations. A few previous studies have also reported patterns of specific cognitive domains that are differentially associated with homocysteine and folate. In a small group of 70 male elderly subjects, the Normative Aging Study found that a higher concentration of homocysteine was specifically associated with poorer spatial copying skills, whereas B vitamins were more strongly associated with memory performance (17). The Sacramento Area Latino Study on Aging also found that, in elderly Latinos, homocysteine was inversely associated with performance on the modified mini-mental state examination (3MSE), picture association, verbal attention span, and pattern recognition, whereas folate was associated with 3MSE and delayed recall (25, 30). However, individual neuropsychological tests may assess overlapping cognitive domains to varying extents, and studies that use different test batteries make cross comparison of results difficult.
Because of our strict exclusion criteria and rigorous control for potential confounding, our results may possibly be biased by overadjustment toward showing null associations. Hence, marginally significant associations should not be dismissed. On the other hand, with numerous neuropsychological variables and multiple significance testings, several statistical tests may be expected to result in spurious significance at the conventional level of P < 0.05. Lastly, the cross-sectional design of our study cannot firmly establish cause-and-effect relation. Additional longitudinal studies are warranted.
In our study, the lack of an observable association between vitamin B-12 and performance on any cognitive test suggests a less important role for vitamin B-12 than for homocysteine and folate. However, the assessment of cobalamin status by using serum vitamin B-12 may have limitations. Methylmalonic acid and holotranscobalamin are proposed to be more sensitive markers for vitamin B-12 deficiency (2, 41). In the study conducted by Refsum et al (42), low holotranscobalamin was associated with Alzheimer disease even though total cobalamin was not. Hence, our study does not exclude a possible association with vitamin B-12.
It is interesting to discuss why homocysteine and folate can exert separate effects on cognitive function. Although homocysteine is viewed as a marker of folate deficiency, its concentration is affected by multiple factors (1). In our study population, the correlation between homocysteine and folate was moderate (r = –0.235), making it possible to analyze their differential effects. Folate provides the methyl group for the conversion of methionine to S-adenosylmethionine, which is the major methyl donor for most methyltransferase reactions; thus, folate deficiency can cause hypomethylation (43). Elevated homocysteine is postulated to affect the brain through vascular or neurotoxic mechanisms (44, 45). Folate deficiency, on the other hand, may affect brain function through high homocysteine (as an intermediary), through decreased S-adenosylmethionine, or both (45). Our results favor the involvement of both pathways and the differential effect of each mechanism. Protein synthesis in the brain is an essential process for the encoding of information in episodic memory, and it is conceivable that the B vitamin–related memory deficits could be due to changes in brain protein metabolism (46). We noted that, with a wide range of cognitive domains measured, the association between folate and cognition appears confined to episodic memory performance; thus, our results suggest that this is plausible.
Although conclusive evidence from randomized controlled trials is still lacking, it is nevertheless possible that modification of homocysteine and B vitamin status could delay cognitive decline or even enhance cognitive function in elderly persons. Observational studies of community-dwelling elderly showed that the rate of change of homocysteine concentrations was significantly correlated with the rate of change of the Stroop scores (47). More recently, the Hordaland Homocysteine Study found that a "favorable" change in folate or homocysteine concentrations over time was associated with better memory performance (31). Large scale clinical trials with a longer interventional period are needed
In conclusion, we found that in high-functioning Chinese older adults, there were differential associations of homocysteine, folate, and vitamin B-12 with specific domains of cognitive performance. Homocysteine was specifically associated with constructional ability and information processing speed, whereas folate was associated specifically with episodic memory and language ability. Vitamin B-12 did not appear to be associated with any cognitive performance, but additional studies are warranted.
ACKNOWLEDGMENTS
LF formulated the hypothesis, performed the literature review and statistical data analysis, and drafted and revised the manuscript. T-PN conceptualized and designed the study, formulated the hypothesis, reviewed statistical analysis and results, and drafted and revised the manuscript. LC designed and formulated the neuropsychological testing and reviewed the manuscript. MN participated in the study design, statistical analysis, interpretation of the results, and review of the manuscript. E-HK participated in the study design, interpretation of the results, and review of the manuscript. No commercial company sponsored or played any role in the design and methods of the study, subject recruitment, data collection and analysis, and preparation of this article.
REFERENCES