点击显示 收起
1 From the Departments of Gerontology and Geriatrics (SPM, JG, RGJW, and AJMdC) and Clinical Chemistry (MF), Leiden University Medical Center, Leiden, Netherlands; the Department of Psychiatry and Neuropsychology, Maastricht University, Maastricht, Netherlands (JJ); and the Academic Section of Geriatric Medicine, University of Glasgow, Glasgow, Scotland (DJS)
2 The Leiden 85-Plus Study was partly funded by the Dutch Ministry of Health, Welfare and Sports. Homocysteine measurements were funded in part by an unrestricted grant from Abbott Laboratories. 3 Address reprint requests to AJM de Craen, Department of Gerontology and Geriatrics, C-2-R, Leiden University Medical Center, PO Box 9600, 2300 RC, Leiden, Netherlands. E-mail: craen{at}lumc.nl.
ABSTRACT
Background: High concentrations of homocysteine and low concentrations of vitamin B-12 and folic acid are frequently observed in subjects with dementia.
Objective: We assessed whether serum concentrations of homocysteine, vitamin B-12, or folic acid predict cognitive decline in old age.
Design: This was a prospective, population-based, longitudinal study of 599 subjects (Leiden 85-Plus Study, Netherlands). The subjects were administered a battery of cognitive tests (including the Mini Mental State Examination, the Stroop test, a letter digit coding test, and a word recall test) at 85 y of age and yearly thereafter until 89 y of age. Serum concentrations of homocysteine, vitamin B-12, and folic acid were measured at 85 and 89 y of age. Cross-sectional associations between serum concentrations and cognition were assessed at 85 and 89 y of age. The association between baseline serum concentrations and subsequent longitudinal cognitive decline was assessed with the use of mixed linear models.
Results: In the cross-sectional analyses, serum concentrations of homocysteine and folic acid were significantly associated with cognitive performance, but serum concentrations of vitamin B-12 were not. In the longitudinal analyses, there were no significant associations of serum concentrations of homocysteine, vitamin B-12, or folic acid with rate of cognitive decline.
Conclusions: Elevated serum concentrations of homocysteine and reduced folic acid are associated with cognitive impairment in elderly persons but do not predict an increased rate of cognitive decline. The association of high serum concentrations of homocysteine and low folic acid with cognitive impairment in old age is likely to be a consequence of disease and not a contributory cause.
Key Words: Homocysteine • folic acid • vitamin B-12 • elderly persons • cognitive decline
INTRODUCTION
Deficiencies of vitamin B-12 or folic acid are common in elderly persons (1). Determination of serum vitamin B-12 and folic acid status is often included in the diagnostic workup in patients undergoing assessment for dementia (2). Moreover, treatment of low concentrations of vitamin B-12 or folic acid in early dementia is common practice (3). However, data on the efficacy of correction of these biochemical abnormalities in preventing cognitive decline or dementia are scarce. Six relatively small randomized trials have been conducted (4–9). In all those trials, subjects were supplemented for <3 mo with vitamin B-12, folic acid, or both, and the overall results were negative (10, 11).
There are 2 biological mechanisms by which low concentrations of vitamin B-12 or folic acid might result in dementia. First, vitamin B-12 and folic acid are essential factors for the methylation of homocysteine into methionine. A shortage of vitamin B-12 or folic acid causes an increase in homocysteine concentrations. Increased plasma concentrations of homocysteine have been associated with an increased prevalence of poor cognitive function (12) and an increased risk of development of dementia and Alzheimer disease (13). Second, low serum concentrations of vitamin B-12 or folic acid might result in a reduced availability of methyl groups in the brain. This can ultimately result in the impaired formation of myelin, various neurotransmitters, and membrane phospholipids (14).
Although a biological, causal connection seems plausible, the association between homocysteine, vitamin B-12, and folic acid on the one hand and subsequent cognitive decline on the other remains controversial. Using data from the Leiden 85-Plus Study, a population-based follow-up study in the elderly, we assessed whether plasma concentrations of homocysteine, vitamin B-12, or folic acid are associated with cognitive impairment and increased risk of cognitive decline in old age.
SUBJECTS AND METHODS
The Leiden 85-Plus Study is a population-based longitudinal study of inhabitants of Leiden, Netherlands. Between September 1997 and September 1999, 705 inhabitants of Leiden reached the age of 85 y and were therefore eligible to participate. There were no selection criteria for health or demographic characteristics. Fourteen inhabitants died before they could be enrolled. Of the remaining 691 subjects, 599 subjects participated (response rate of 87%). There were no significant demographic differences between the 599 respondents and the source population (15). All subjects gave informed consent. For cognitively impaired subjects, informed consent was obtained from a guardian. The Medical Ethical Committee of the Leiden University Medical Center approved the study.
All subjects, who were either free-living or living in institutions, were visited at their place of residence, where face-to-face interviews were administered, cognitive testing was performed, and a venous blood sample was drawn. Of the 599 participants, 7 subjects died before a blood sample could be obtained and 33 subjects refused to give a blood sample. Hence, a venous blood sample was available for 559 subjects. All subjects were revisited annually at their place of residence until the age of 89 y.
Measurements
All blood samples were drawn in the morning before 1100. The median time between venipuncture and centrifugation in our laboratory was 50 min (interquartile range: 30–70 min). Serum samples were stored at –80°C. After collection of all samples, serum homocysteine concentrations were measured in batches after reduction to the free form with a fluorescence polarization immunoassay on an IMx analyzer (Abbott, Abbott Park, IL.). Serum concentrations of vitamin B-12 and folic acid were measured in batches by using the Dual Count Solid Phase No Boil Assay (Diagnostic Products Corporation, Los Angeles, CA). The interassay CVs at various concentrations varied between 2.2% and 2.5% for homocysteine, between 4.3% and 9.8% for vitamin B-12, and between 4.7% and 7.3% for folic acid.
We administered the Mini Mental State Examination (MMSE) to assess global cognitive function (16). To further investigate the various domains of cognitive function, we used a neurocognitive test battery: the Stroop test, the letter digit coding test, and the word learning test (17). These tests have been used in the assessment of large samples of elderly subjects and have been shown to be reliable and sensitive (18). The Stroop test was used to assess attention. On a sheet of paper, 40 color words are printed in ink of different color from the word. For example, green is printed in blue letters, and the subject has to say blue. For analysis, we used the total number of seconds to complete the third Stroop card containing the 40 words. The letter digit coding test was used to assess cognitive processing speed. In this test, subjects are asked to fill in digits near letters according to a key presented at the top of the test sheet. For analysis, we used the total number of correct entries that were filled in after 60 s. The 12-word learning test was used to assess immediate and delayed memory. Twelve pictures were successively presented at a rate of one per 2 s. Next, the person was asked to recall as many pictures as possible. This procedure was carried out 3 times. The sum of recalled words of the 3 procedures was the measurement of immediate memory. After 20 min, delayed recall was tested. The number of pictures that was recalled was the measurement of delayed memory. A compound cognitive index was calculated of the 4 cognitive domains (attention, processing speed, immediate memory, and delayed memory) by averaging the 4 individual test scores, which were first transformed into a standardized z score (z score = [test score –mean test score]/SD).
The 4 neuropsychological tests were administered only to subjects with an MMSE score 19 points because the tests cannot be accurately administered in severely cognitively impaired subjects (ie, those with an MMSE score <19 points). The MMSE and all neuropsychological tests were administered at baseline and each year thereafter.
Additional measurements
Sex, level of education, use of vitamin B-12 and folic acid supplements, and living arrangement were considered as possible confounders. Subjects were classified into 2 levels of education: those with a lower education level, which included participants without schooling or with primary school education only (with a maximum of 6 y of schooling), and those with a higher education level (15). To determine use of vitamin B-12 and folic acid supplements, subjects’ pharmacy records were examined and subjects were asked about their use of vitamin supplements. Subjects were classified as being in institutional care when they resided in a nursing home or a home for the elderly. Depressive symptoms were assessed with the 15-item Geriatric Depression Scale (19). Subjects with a Geriatric Depression Scale score >4 points were considered to be depressed.
Statistical analysis
The cross-sectional associations between homocysteine, vitamin B-12, and folic acid and the scores on the global cognitive functioning test were assessed by using the nonparametric Jonckheere-Terpstra test for trend. To assess the longitudinal association, we used linear mixed models (20) that were controlled for sex, level of education, use of vitamin B-12 and folic acid supplements, and living arrangement. Individual concentrations of homocysteine, vitamin B-12, and folic acid were first logarithmically transformed and were then recalculated by use of the following formula to normalize the data and provide comparable risk estimates per 1-SD increase: (individual serum level –mean level)/SD. Next to homocysteine, vitamin B-12, or folic acid, each model included time and the interaction term of time and homocysteine, vitamin B-12, or folic acid. In the linear mixed model, the term for homocysteine, vitamin B-12, or folic acid represents the cross-sectional association between homocysteine, vitamin B-12, or folic acid and the scores on the selected cognitive test administered at baseline. The term for time indicates the linear deterioration per year on the cognitive test. The term for the interaction between time and homocysteine, vitamin B-12, or folic acid represents the additional annual deterioration on the selected cognitive test for each 1-SD increase in homocysteine, vitamin B-12, or folic acid.
RESULTS
Characteristics of the 559 participants are presented in Table 1. Two-thirds (66%) of the subjects were female, 65% had a low education level, 82% still lived independently, and depressive symptoms were observed in 110 (23%). Folic acid supplements were used nearly twice as often as vitamin B-12 supplements, by 11% and 6% of the participants, respectively.
View this table:
TABLE 1. Demographic and clinical characteristics of the study population
In the cross-sectional analyses at age 85 y, higher serum concentrations of homocysteine were related to poor global cognitive function as assessed with the MMSE (Table 2). The association between homocysteine and cognitive performance remained after the exclusion of institutionalized and vitamin-supplemented subjects. Serum concentrations of vitamin B-12 were inversely related to global cognitive function. However, this association disappeared after the exclusion of institutionalized and supplemented subjects. Serum concentrations of folic acid increased with better cognitive function. This association persisted when institutionalized and supplemented subjects were excluded. The associations of homocysteine, vitamin B-12, and folic acid with cognitive function were similar in men and women (data not shown). That is, there was no significant interaction with sex for any of the variables.
View this table:
TABLE 2. Serum concentrations of homocysteine, vitamin B-12, and folic acid according to Mini Mental State Examination (MMSE) score at age 85 y
Serum homocysteine, vitamin B-12, and folic acid concentrations were measured for a second time in 341 subjects who survived until age 89 y (Table 3). The associations between homocysteine, vitamin B-12, and folic acid on one the hand and cognitive function on the other were similar to those observed at age 85 y, except for homocysteine, which was no longer significantly associated with cognitive function in the unsupplemented, noninstitutionalized subjects.
View this table:
TABLE 3. Serum concentrations of homocysteine, vitamin B-12, and folic acid according to Mini Mental State Examination (MMSE) score at age 89 y1
Next, we used linear mixed models to examine the influence of concentrations of homocysteine, vitamin B-12, and folic acid at age 85 y on the change in the various cognitive performance tests during follow-up (Table 4). The domains of cognitive performance were prospectively measured at ages 85, 86, 87, 88, and 89 y. The cross-sectional association between homocysteine, vitamin B-12, and folic acid at age 85 y and global cognitive functioning at age 85 y obtained from these models was in line with the crude cross-sectional results from Table 2. Moreover, in all models, the term for time was highly significant, indicating that all domains of cognitive performance deteriorated from age 85 y onward. In the prospective analysis, a higher baseline concentration of homocysteine, vitamin B-12, or folic acid was not related to additional deterioration as measured on any of the cognitive tests. Therefore, concentrations of homocysteine, vitamin B-12, or folic acid at age 85 y were not predictive of the rate of cognitive decline during 4 y of follow-up. The interpretation of the significant positive association between baseline folic acid and decline in global cognitive function is that for each 1-SD increase in folic acid at age 85 y, the subsequent decrease in global cognitive decline over the years was accelerated. When we used the compound cognitive index, calculated by averaging the 4 individual cognitive test scores that were first transformed into a standardized z score, the analysis showed a similar picture as when the 4 cognitive tests were analyzed separately.
View this table:
TABLE 4. Association between baseline serum concentrations of homocysteine, vitamin B-12, and folic acid with annual rate of decline on various cognitive performance tests1
DISCUSSION
We assessed whether serum concentrations of homocysteine, vitamin B-12, and folic acid are predictive of cognitive decline in old age. To do so, we measured various domains of cognitive performance for 5 consecutive years in a large cohort of elderly subjects. Serum concentrations of homocysteine, vitamin B-12, and folic acid were measured at enrolment, when all subjects were 85 y old, and at the end of follow-up, when the subjects were 89 y old. In the cross-sectional analyses at 85 and 89 y of age, we found no significant association between serum concentrations of vitamin B-12 and cognitive performance, but serum concentrations of homocysteine and folic acid were significantly associated with cognitive performance. In the longitudinal analyses, however, serum concentrations of homocysteine, vitamin B-12, and folic acid did not predict deterioration in any of the cognitive domains assessed.
The initial cross-sectional associations we found between high concentrations of vitamin B-12 and low cognitive function both at 85 and again at 89 y of age probably reflect the supplementation policy in institutionalized subjects. When we excluded the institutionalized subjects from the analysis, the inverse association between concentrations of vitamin B-12 and cognitive function disappeared. However, the cross-sectional association between concentrations of homocysteine and folic acid on the one hand and cognitive functioning on the other remained after we excluded all institutionalized subjects. This cross-sectional association may be caused by the poor eating habits of cognitively impaired subjects. The fact that plasma concentrations of homocysteine, vitamin B-12, and folic acid were not risk factors for longitudinal cognitive decline, in combination with a significant cross-sectional association at both 85 and 89 y of age, strongly suggests that increased concentrations of homocysteine and a reduction of vitamin B-12 and folic acid identify only those individuals who have developed cognitive impairment over time.
Few longitudinal studies have assessed whether low serum concentrations of vitamin B-12 and folic acid are risk factors for subsequent cognitive decline and dementia. Crystal et al (21) found similar incidences of dementia and Alzheimer disease in subjects with low serum concentrations of vitamin B-12 compared with subjects with normal vitamin B-12 concentrations. Wang et al (22) found no significant association between vitamin B-12 at baseline and incidence of dementia, but reported that the combination of low vitamin B-12 and low folic acid was significantly related to poor cognitive function. Five nonrandomized vitamin B-12 supplementation studies have been reported (23–27), 4 of which showed beneficial results on cognitive function (23–26), whereas 1 study showed no effect (27). However, all studies were seriously flawed because none used a randomized design and some did not have a control group. Two small placebo-controlled randomized trials of vitamin B-12 have been reported (4, 5). Both studies found no beneficial effect of vitamin B-12 supplementation on cognitive function. Moreover, 4 placebo-controlled randomized trials of folic acid supplementation, with or without vitamin B-12, have been reported (6–9). Although folic acid plus vitamin B-12 was effective in reducing serum homocysteine concentrations, none of the trials found a beneficial effect of folic acid supplementation on any measure of cognition for healthy or cognitively impaired or demented persons (10, 11). The negative outcomes of all the randomized trials (4–9), however, might also be explained by their short duration or by their limited number of participants.
Several studies have investigated the longitudinal association between serum concentrations of homocysteine and cognitive decline. In a 3-y follow-up of 43 patients with Alzheimer disease, Clarke et al (28) found more radiologic disease progression on brain computed tomography among subjects with higher homocysteine concentrations at baseline. However, they did not find an association with rate of cognitive decline. McCaddon et al (29) showed that high concentrations of homocysteine at baseline were associated with more rapid cognitive decline in 32 healthy elderly individuals. Moreover, Seshadri et al (13) found that that elevated homocysteine at baseline was a strong risk factor for incident dementia up to 11 y later. In that study, however, no evidence was given that the subjects who developed dementia had comparable cognitive abilities at baseline as did the subjects who did not develop dementia. Finally, Dufouil et al (30) showed that the odds of cognitive decline in subjects with homocysteine concentrations >15 µmol/L was 2.8-fold that in subjects with homocysteine concentrations <10 µmol/L. Hence, the longitudinal evidence of the association between elevated concentrations of homocysteine and subsequent cognitive decline is not overwhelming: 2 studies were very small and therefore likely never to be published in case there was no association, and 1 study might have been flawed by a possible unequal distribution of baseline cognitive function.
The present work has several strengths. First, it combines a repeated cross-sectional analysis with a longitudinal design. This allowed us to disentangle the temporal relation between concentrations of homocysteine, vitamin B-12, and folic acid in combination with cognitive performance. We found that baseline concentrations of homocysteine, vitamin B-12, or folic acid were not predictive of subsequent cognitive decline but merely followed the cognitive decline. Second, our study was population based. This means that it included a wide range of vitamin concentrations and cognitive abilities, which allows direct extrapolation to older persons in the population at large. Third, because we documented the use of vitamin supplements and the living situation of all the study participants, we were able to assess the possible confounding effects of these variables.
A potential weakness is that we studied persons from age 85 y onward. Although this age is generally considered rather extreme, this is not the case. On the basis of official data from the Dutch Bureau of Statistics, 15% of men and 36% of women born between 1912 and 1914 actually survived until the age of 85 y. Additionally, the highest incidence of dementia is in those aged 85 y (31). Thus, studying risk factors for cognitive decline in 85-y-olds is relevant. Hence, we think that understanding the biological risk factors for cognitive impairment at very old age to be important, both from a scientific and from a clinical point of view.
We realize that serum vitamin B-12 concentrations may not be the best indicator of metabolic vitamin B-12 deficiency. About 20–40% of elderly persons with low serum vitamin B-12 concentrations have normal metabolic status, whereas high serum concentrations do not always indicate sufficient vitamin B-12 availability at the tissue level. In clinical practice, however, serum concentrations are routinely used in the assessment of dementia and are included in the clinical guidelines. Our results indicate that serum concentrations of vitamin B-12 are not useful as a tool for identifying subjects at risk of cognitive decline. Moreover, subjects with metabolic vitamin B-12 deficiency who present with normal serum vitamin B-12 concentrations are likely to have higher serum concentrations of homocysteine. A high serum concentration of homocysteine was also not found to be a risk factor for cognitive decline.
In summary, in old age, increased serum concentrations of homocysteine and reduced vitamin B-12 or folic acid identify individuals with cognitive impairment but do not predict cognitive decline. Hence, these concentrations are unlikely to be biological risk factors of cognitive decline in old age. We think that high serum concentrations of homocysteine and low serum concentrations of vitamin B-12 and folic acid in subjects with dementia are more likely to be a consequence of the cognitive decline, possibly caused by a change in dietary intake, and not a contributory cause.
ACKNOWLEDGMENTS
SPM was responsible for data analysis and writing of the manuscript. JG was responsible for the study design, supervision of data collection, interpretation of data, and writing of the manuscript. MF was responsible for laboratory analysis and writing of the manuscript. JJ was responsible for consultation, interpretation of data, and writing of the manuscript. DJS was responsible for consultation, interpretation of data, and writing of the manuscript. RGJW was responsible for study design, interpretation of data, and writing of the manuscript. AJMdC was responsible for study design, data analysis, interpretation of data, and writing of the manuscript. None of the authors had a conflict of interest.
REFERENCES